ambiser
/
face-api
mirror of https://github.com/vladmandic/face-api.git
9
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
face-api/dist/face-api.js

4949 lines
1.2 MiB
Raw Blame History

/*
Face-API
homepage: <https://github.com/vladmandic/face-api>
author: <https://github.com/vladmandic>'
*/
var faceapi=(()=>{var fb=Object.defineProperty;var UA=Object.getOwnPropertyDescriptor;var GA=Object.getOwnPropertyNames;var HA=Object.prototype.hasOwnProperty;var jA=(e=>typeof require!="undefined"?require:typeof Proxy!="undefined"?new Proxy(e,{get:(t,n)=>(typeof require!="undefined"?require:t)[n]}):e)(function(e){if(typeof require!="undefined")return require.apply(this,arguments);throw new Error('Dynamic require of "'+e+'" is not supported')});var yh=(e,t)=>{for(var n in t)fb(e,n,{get:t[n],enumerable:!0})},qA=(e,t,n,a)=>{if(t&&typeof t=="object"||typeof t=="function")for(let r of GA(t))!HA.call(e,r)&&r!==n&&fb(e,r,{get:()=>t[r],enumerable:!(a=UA(t,r))||a.enumerable});return e};var KA=e=>qA(fb({},"__esModule",{value:!0}),e);var nce={};yh(nce,{AgeGenderNet:()=>Nd,BoundingBox:()=>Fo,Box:()=>lt,ComposableTask:()=>wa,ComputeAllFaceDescriptorsTask:()=>Hr,ComputeFaceDescriptorsTaskBase:()=>Ed,ComputeSingleFaceDescriptorTask:()=>jr,DetectAllFaceLandmarksTask:()=>Ad,DetectAllFacesTask:()=>wp,DetectFaceLandmarksTaskBase:()=>Fd,DetectFacesTaskBase:()=>Dd,DetectSingleFaceLandmarksTask:()=>$d,DetectSingleFaceTask:()=>Rd,Dimensions:()=>yn,FACE_EXPRESSION_LABELS:()=>m1,FaceDetection:()=>vt,FaceDetectionNet:()=>w1,FaceExpressionNet:()=>Sd,FaceExpressions:()=>Vr,FaceLandmark68Net:()=>zo,FaceLandmark68TinyNet:()=>Td,FaceLandmarkNet:()=>b1,FaceLandmarks:()=>ra,FaceLandmarks5:()=>e1,FaceLandmarks68:()=>$o,FaceMatch:()=>op,FaceMatcher:()=>Md,FaceRecognitionNet:()=>Bo,Gender:()=>Cg,LabeledBox:()=>lp,LabeledFaceDescriptors:()=>xr,NetInput:()=>wr,NeuralNetwork:()=>ln,ObjectDetection:()=>zr,Point:()=>Pe,PredictedBox:()=>t1,Rect:()=>Ao,SsdMobilenetv1:()=>Fs,SsdMobilenetv1Options:()=>va,TinyFaceDetector:()=>Ho,TinyFaceDetectorOptions:()=>Cd,TinyYolov2:()=>Uo,TinyYolov2Options:()=>ar,allFaces:()=>Zpe,allFacesSsdMobilenetv1:()=>KF,allFacesTinyYolov2:()=>Jpe,awaitMediaLoaded:()=>l1,bufferToImage:()=>u1,computeFaceDescriptor:()=>Lpe,createCanvas:()=>Po,createCanvasFromMedia:()=>vd,createFaceDetectionNet:()=>Ape,createFaceRecognitionNet:()=>vpe,createSsdMobilenetv1:()=>DF,createTinyFaceDetector:()=>ece,createTinyYolov2:()=>Rpe,detectAllFaces:()=>Og,detectFaceLandmarks:()=>jF,detectFaceLandmarksTiny:()=>Ope,detectLandmarks:()=>Ype,detectSingleFace:()=>Qpe,draw:()=>g1,env:()=>et,euclideanDistance:()=>S1,extendWithAge:()=>$g,extendWithFaceDescriptor:()=>Ag,extendWithFaceDetection:()=>Do,extendWithFaceExpressions:()=>Sg,extendWithFaceLandmarks:()=>yp,extendWithGender:()=>Dg,extractFaceTensors:()=>cp,extractFaces:()=>pp,fetchImage:()=>ipe,fetchJson:()=>d1,fetchNetWeights:()=>ope,fetchOrThrow:()=>Wr,fetchVideo:()=>lpe,getContext2dOrThrow:()=>jn,getMediaDimensions:()=>Mo,imageTensorToCanvas:()=>p1,imageToSquare:()=>c1,inverseSigmoid:()=>Zue,iou:()=>Y0,isMediaElement:()=>gg,isMediaLoaded:()=>xd,isWithAge:()=>wpe,isWithFaceDetection:()=>vr,isWithFaceExpressions:()=>f1,isWithFaceLandmarks:()=>Lo,isWithGender:()=>kpe,loadAgeGenderModel:()=>qpe,loadFaceDetectionModel:()=>Kpe,loadFaceExpressionModel:()=>jpe,loadFaceLandmarkModel:()=>Upe,loadFaceLandmarkTinyModel:()=>Gpe,loadFaceRecognitionModel:()=>Hpe,loadSsdMobilenetv1Model:()=>qF,loadTinyFaceDetectorModel:()=>Wpe,loadTinyYolov2Model:()=>Vpe,loadWeightMap:()=>h1,locateFaces:()=>Xpe,matchDimensions:()=>upe,minBbox:()=>Q0,nets:()=>tt,nonMaxSuppression:()=>J0,normalize:()=>tr,padToSquare:()=>Z0,predictAgeAndGender:()=>Bpe,recognizeFaceExpressions:()=>zpe,resizeResults:()=>XF,resolveInput:()=>Ro,shuffleArray:()=>Jue,sigmoid:()=>gd,ssdMobilenetv1:()=>HF,tf:()=>Le,tinyFaceDetector:()=>Mpe,tinyYolov2:()=>Ppe,toNetInput:()=>wt,utils:()=>X0,validateConfig:()=>k1,version:()=>tce});var Le={};yh(Le,{Abs:()=>Ll,Acos:()=>zl,Acosh:()=>Bl,AdadeltaOptimizer:()=>Af,AdagradOptimizer:()=>$f,AdamOptimizer:()=>Df,AdamaxOptimizer:()=>Rf,Add:()=>bs,AddN:()=>wi,All:()=>Wl,Any:()=>Vl,ArgMax:()=>ki,ArgMin:()=>yc,Asin:()=>Ul,Asinh:()=>Gl,Atan:()=>Hl,Atan2:()=>ql,Atanh:()=>jl,AvgPool:()=>Ii,AvgPool3D:()=>bc,AvgPool3DGrad:()=>Im,AvgPoolGrad:()=>km,BackendWasm:()=>lF,BatchMatMul:()=>Si,BatchToSpaceND:()=>Kl,Bincount:()=>Sm,BroadcastArgs:()=>Nm,BroadcastTo:()=>DI,Callback:()=>BN,CallbackList:()=>G2,Cast:()=>Ni,Ceil:()=>Ti,ClipByValue:()=>xs,Complex:()=>Tm,ComplexAbs:()=>xc,Concat:()=>Xl,Conv2D:()=>_i,Conv2DBackpropFilter:()=>_m,Conv2DBackpropInput:()=>Ci,Conv3D:()=>vc,Conv3DBackpropFilterV2:()=>Cm,Conv3DBackpropInputV2:()=>Em,Cos:()=>Ei,Cosh:()=>Fi,CropAndResize:()=>Ql,Cumprod:()=>Yl,Cumsum:()=>Ai,CustomCallback:()=>j2,DataStorage:()=>xm,DenseBincount:()=>Fm,DepthToSpace:()=>Jl,DepthwiseConv2dNative:()=>$i,DepthwiseConv2dNativeBackpropFilter:()=>Am,DepthwiseConv2dNativeBackpropInput:()=>$m,Diag:()=>Dm,Dilation2D:()=>wc,Dilation2DBackpropFilter:()=>Hh,Dilation2DBackpropInput:()=>Gh,ENV:()=>Rx,EarlyStopping:()=>WN,Einsum:()=>Rm,Elu:()=>Ri,EluGrad:()=>Mm,Environment:()=>AI,Equal:()=>eu,Erf:()=>Zl,Exp:()=>Mi,ExpandDims:()=>tu,Expm1:()=>nu,FFT:()=>Pm,Fill:()=>kc,FlipLeftRight:()=>au,Floor:()=>Pi,FloorDiv:()=>Oi,FromPixels:()=>jh,FusedBatchNorm:()=>Li,FusedConv2D:()=>ri,FusedDepthwiseConv2D:()=>si,GPGPUContext:()=>zh,GatherNd:()=>su,GatherV2:()=>ru,GraphModel:()=>f0,Greater:()=>iu,GreaterEqual:()=>zi,History:()=>H2,IFFT:()=>Om,Identity:()=>Bi,Imag:()=>Lm,InputSpec:()=>Bt,IsFinite:()=>ou,IsInf:()=>lu,IsNan:()=>uu,KernelBackend:()=>gc,LRN:()=>Nc,LRNGrad:()=>Bm,LayerVariable:()=>O2,LayersModel:()=>Fr,LeakyRelu:()=>Wi,Less:()=>pu,LessEqual:()=>cu,LinSpace:()=>zm,Log:()=>Vi,Log1p:()=>du,LogSoftmax:()=>RI,LogicalAnd:()=>hu,LogicalNot:()=>Ic,LogicalOr:()=>Sc,LowerBound:()=>U$,MathBackendWebGL:()=>sg,Max:()=>Ui,MaxPool:()=>Hi,MaxPool3D:()=>Tc,MaxPool3DGrad:()=>Vm,MaxPoolGrad:()=>Wm,MaxPoolWithArgmax:()=>Um,Maximum:()=>Gi,Mean:()=>ji,Min:()=>qi,Minimum:()=>Ki,MirrorPad:()=>Xi,Mod:()=>mu,MomentumOptimizer:()=>Mf,Multinomial:()=>Gm,Multiply:()=>Yi,Neg:()=>fu,NonMaxSuppressionV3:()=>yu,NonMaxSuppressionV4:()=>bu,NonMaxSuppressionV5:()=>xu,NotEqual:()=>gu,OP_SCOPE_SUFFIX:()=>GI,OneHot:()=>Qi,OnesLike:()=>vu,Optimizer:()=>Mr,OptimizerConstructors:()=>Jr,Pack:()=>wu,PadV2:()=>Ji,Pool:()=>G$,Pow:()=>Zi,Prelu:()=>eo,Prod:()=>to,RMSPropOptimizer:()=>Pf,RNN:()=>yr,Range:()=>_c,Rank:()=>Db,Real:()=>Hm,RealDiv:()=>Di,Reciprocal:()=>ku,Reduction:()=>In,Relu:()=>no,Relu6:()=>ro,Reshape:()=>Iu,ResizeBilinear:()=>ao,ResizeBilinearGrad:()=>qm,ResizeNearestNeighbor:()=>Cc,ResizeNearestNeighborGrad:()=>jm,Reverse:()=>so,RotateWithOffset:()=>zu,Round:()=>io,Rsqrt:()=>oo,SGDOptimizer:()=>Qc,ScatterNd:()=>Su,SearchSorted:()=>Km,Select:()=>Nu,Selu:()=>Tu,Sequential:()=>Fl,Sigmoid:()=>uo,Sign:()=>Eu,Sin:()=>lo,Sinh:()=>Cu,Slice:()=>_u,Softmax:()=>ho,Softplus:()=>Fu,SpaceToBatchND:()=>Au,SparseFillEmptyRows:()=>Ec,SparseReshape:()=>Du,SparseSegmentMean:()=>Fc,SparseSegmentSum:()=>Ac,SparseToDense:()=>Xm,SplitV:()=>$u,Sqrt:()=>po,Square:()=>$c,SquaredDifference:()=>mo,Step:()=>ws,StridedSlice:()=>Ru,StringNGrams:()=>Ym,StringSplit:()=>Qm,StringToHashBucketFast:()=>Jm,Sub:()=>fo,Sum:()=>co,SymbolicTensor:()=>Ua,Tan:()=>go,Tanh:()=>yo,Tensor:()=>Fe,TensorBuffer:()=>qt,Tile:()=>vs,TopK:()=>Mu,Transform:()=>Pu,Transpose:()=>Cr,Unique:()=>Zm,Unpack:()=>Ou,UnsortedSegmentSum:()=>Dc,UpperBound:()=>H$,Variable:()=>ls,ZerosLike:()=>Lu,_FusedMatMul:()=>ai,abs:()=>zt,acos:()=>Yx,acosh:()=>Qx,add:()=>J,addN:()=>NS,all:()=>sf,any:()=>rc,argMax:()=>li,argMin:()=>Jx,asin:()=>Zx,asinh:()=>ev,atan:()=>tv,atan2:()=>nv,atanh:()=>av,avgPool:()=>ga,avgPool3d:()=>sv,backend:()=>oS,backend_util:()=>C,basicLSTMCell:()=>WM,batchNorm:()=>$r,batchNorm2d:()=>ES,batchNorm3d:()=>FS,batchNorm4d:()=>AS,batchToSpaceND:()=>Wc,bincount:()=>iv,booleanMaskAsync:()=>q3,broadcastArgs:()=>$S,broadcastTo:()=>xl,broadcast_util:()=>Bu,browser:()=>bo,buffer:()=>Ve,callbacks:()=>I6,cast:()=>oe,ceil:()=>ov,clipByValue:()=>an,clone:()=>Er,complex:()=>Ar,concat:()=>Ze,concat1d:()=>DS,concat2d:()=>RS,concat3d:()=>MS,concat4d:()=>PS,constraints:()=>B2,conv1d:()=>of,conv2d:()=>Rt,conv2dTranspose:()=>lf,conv3d:()=>uv,conv3dTranspose:()=>LS,copyRegisteredKernels:()=>X$,cos:()=>Vc,cosh:()=>uf,cosineWindow:()=>Ov,cumprod:()=>sc,cumsum:()=>pf,customGrad:()=>cr,data:()=>cT,denseBincount:()=>zS,deprecationWarn:()=>Ux,depthToSpace:()=>pv,depthwiseConv2d:()=>Ss,deregisterOp:()=>T6,device_util:()=>Oc,diag:()=>bP,dilation2d:()=>cv,disableDeprecationWarnings:()=>IR,dispose:()=>Re,disposeVariables:()=>SR,div:()=>fe,divNoNan:()=>dv,dot:()=>BS,dropout:()=>u2,einsum:()=>WS,elu:()=>Wu,enableDebugMode:()=>kR,enableProdMode:()=>wR,enclosingPowerOfTwo:()=>p2,engine:()=>sr,env:()=>X,equal:()=>Zn,erf:()=>hv,euclideanNorm:()=>gv,exp:()=>gn,expandDims:()=>mn,expm1:()=>yv,eye:()=>bv,fft:()=>Xc,fill:()=>Cn,findBackend:()=>AR,findBackendFactory:()=>$R,floor:()=>Vu,floorDiv:()=>rf,forceHalfFloat:()=>EC,fused:()=>ps,gather:()=>ci,gatherND:()=>l2,gather_util:()=>Gx,getBackend:()=>ER,getGradient:()=>Ab,getKernel:()=>qh,getKernelsForBackend:()=>Kh,getThreadsCount:()=>zue,gpgpu_util:()=>oC,grad:()=>nO,grads:()=>aO,greater:()=>Un,greaterEqual:()=>Ns,ifft:()=>_l,imag:()=>zc,image:()=>Ln,inTopKAsync:()=>aL,initializers:()=>W2,input:()=>iN,io:()=>en,irfft:()=>Sf,isFinite:()=>jS,isInf:()=>qS,isNaN:()=>xv,keep:()=>tn,kernel_impls:()=>gr,layers:()=>V2,leakyRelu:()=>Gc,less:()=>cf,lessEqual:()=>Ts,linalg:()=>v2,linspace:()=>KS,loadGraphModel:()=>EH,loadGraphModelSync:()=>FH,loadLayersModel:()=>DU,localResponseNormalization:()=>vv,log:()=>ea,log1p:()=>Hc,logSigmoid:()=>YS,logSoftmax:()=>hf,logSumExp:()=>wv,logicalAnd:()=>Ca,logicalNot:()=>jc,logicalOr:()=>mf,logicalXor:()=>QS,losses:()=>zz,lowerBound:()=>JS,matMul:()=>De,math:()=>iS,max:()=>Ta,maxPool:()=>Pt,maxPool3d:()=>Iv,maxPoolWithArgmax:()=>ZS,maximum:()=>fr,mean:()=>Et,memory:()=>Qh,meshgrid:()=>wO,metrics:()=>ON,min:()=>ic,minimum:()=>Uu,mirrorPad:()=>Sv,mod:()=>Nv,model:()=>AU,models:()=>LN,moments:()=>ff,movingAverage:()=>X3,mul:()=>B,multiRNNCell:()=>_O,multinomial:()=>e2,neg:()=>It,nextFrame:()=>Bv,norm:()=>Uc,notEqual:()=>di,oneHot:()=>Sl,ones:()=>Qn,onesLike:()=>ta,op:()=>z,outerProduct:()=>$O,pad:()=>ya,pad1d:()=>MO,pad2d:()=>OO,pad3d:()=>zO,pad4d:()=>WO,pool:()=>t2,pow:()=>Dr,prelu:()=>Kc,print:()=>eS,prod:()=>gf,profile:()=>NR,rand:()=>XO,randomGamma:()=>ZO,randomNormal:()=>n2,randomUniform:()=>Gu,range:()=>Tl,ready:()=>CR,real:()=>Nl,reciprocal:()=>Cv,registerBackend:()=>nf,registerCallbackConstructor:()=>RU,registerGradient:()=>MI,registerKernel:()=>Rc,registerOp:()=>N6,regularizers:()=>zN,relu:()=>Xe,relu6:()=>yf,removeBackend:()=>FR,reshape:()=>W,reverse:()=>na,reverse1d:()=>o3,reverse2d:()=>u3,reverse3d:()=>c3,reverse4d:()=>h3,rfft:()=>Yc,round:()=>bf,rsqrt:()=>xf,scalar:()=>we,scatterND:()=>o2,scatter_util:()=>Hx,searchSorted:()=>kv,selu:()=>vf,separableConv2d:()=>vo,sequential:()=>$U,serialization:()=>se,setBackend:()=>_R,setPlatform:()=>DR,setThreadsCount:()=>Lue,setWasmPath:()=>Pue,setWasmPaths:()=>Oue,setWebGLContext:()=>A_,setdiff1dAsync:()=>a2,sigmoid:()=>ma,sign:()=>Ev,signal:()=>Lz,sin:()=>wf,sinh:()=>kf,slice:()=>He,slice1d:()=>If,slice2d:()=>Fv,slice3d:()=>Hu,slice4d:()=>oc,slice_util:()=>Kt,softmax:()=>Qa,softplus:()=>xo,spaceToBatchND:()=>qc,sparse:()=>Wp,sparseToDense:()=>Pv,spectral:()=>Oz,split:()=>zn,sqrt:()=>pn,square:()=>ut,squaredDifference:()=>Nf,squeeze:()=>dr,stack:()=>Mt,step:()=>ju,stridedSlice:()=>Av,string:()=>$h,sub:()=>ce,sum:()=>be,sumOutType:()=>tf,tan:()=>$v,tanh:()=>ui,tensor:()=>Jn,tensor1d:()=>qe,tensor2d:()=>Ha,tensor3d:()=>af,tensor4d:()=>Ja,tensor5d:()=>L3,tensor6d:()=>z3,tensor_util:()=>Ga,test_util:()=>kS,tidy:()=>O,tile:()=>On,time:()=>TR,topk:()=>Dv,train:()=>Hs,transpose:()=>Ae,truncatedNormal:()=>Tf,unique:()=>Zh,unregisterGradient:()=>K$,unregisterKernel:()=>q$,unsortedSegmentSum:()=>Rv,unstack:()=>ht,upcastType:()=>fa,upperBound:()=>r2,util:()=>w,valueAndGrad:()=>rO,valueAndGrads:()=>sO,variable:()=>s2,variableGrads:()=>XS,version:()=>Yue,version_converter:()=>$H,version_core:()=>uM,version_layers:()=>lw,version_wasm:()=>Bue,version_webgl:()=>q9,webgl:()=>K9,webgl_util:()=>F_,where:()=>fn,whereAsync:()=>Mv,zeros:()=>St,zerosLike:()=>Ke});var XA=Object.create,Ex=Object.defineProperty,YA=Object.getOwnPropertyDescriptor,QA=Object.getOwnPropertyNames,JA=Object.getPrototypeOf,ZA=Object.prototype.hasOwnProperty,ft=(e,t)=>()=>(t||e((t={exports:{}}).exports,t),t.exports),Me=(e,t)=>{for(var n in t)Ex(e,n,{get:t[n],enumerable:!0})},e$=(e,t,n,a)=>{if(t&&typeof t=="object"||typeof t=="function")for(let r of QA(t))!ZA.call(e,r)&&r!==n&&Ex(e,r,{get:()=>t[r],enumerable:!(a=YA(t,r))||a.enumerable});return e},xi=(e,t,n)=>(n=e!=null?XA(JA(e)):{},e$(t||!e||!e.__esModule?Ex(n,"default",{value:e,enumerable:!0}):n,e)),t$=ft((e,t)=>{t.exports=a;var n=null;try{n=new WebAssembly.Instance(new WebAssembly.Module(new Uint8Array([0,97,115,109,1,0,0,0,1,13,2,96,0,1,127,96,4,127,127,127,127,1,127,3,7,6,0,1,1,1,1,1,6,6,1,127,1,65,0,11,7,50,6,3,109,117,108,0,1,5,100,105,118,95,115,0,2,5,100,105,118,95,117,0,3,5,114,101,109,95,115,0,4,5,114,101,109,95,117,0,5,8,103,101,116,95,104,105,103,104,0,0,10,191,1,6,4,0,35,0,11,36,1,1,126,32,0,173,32,1,173,66,32,134,132,32,2,173,32,3,173,66,32,134,132,126,34,4,66,32,135,167,36,0,32,4,167,11,36,1,1,126,32,0,173,32,1,173,66,32,134,132,32,2,173,32,3,173,66,32,134,132,127,34,4,66,32,135,167,36,0,32,4,167,11,36,1,1,126,32,0,173,32,1,173,66,32,134,132,32,2,173,32,3,173,66,32,134,132,128,34,4,66,32,135,167,36,0,32,4,167,11,36,1,1,126,32,0,173,32,1,173,66,32,134,132,32,2,173,32,3,173,66,32,134,132,129,34,4,66,32,135,167,36,0,32,4,167,11,36,1,1,126,32,0,173,32,1,173,66,32,134,132,32,2,173,32,3,173,66,32,134,132,130,34,4,66,32,135,167,36,0,32,4,167,11])),{}).exports}catch(S){}function a(S,P,V){this.low=S|0,this.high=P|0,this.unsigned=!!V}a.prototype.__isLong__,Object.defineProperty(a.prototype,"__isLong__",{value:!0});function r(S){return(S&&S.__isLong__)===!0}a.isLong=r;var s={},i={};function o(S,P){var V,j,q;return P?(S>>>=0,(q=0<=S&&S<256)&&(j=i[S],j)?j:(V=u(S,(S|0)<0?-1:0,!0),q&&(i[S]=V),V)):(S|=0,(q=-128<=S&&S<128)&&(j=s[S],j)?j:(V=u(S,S<0?-1:0,!1),q&&(s[S]=V),V))}a.fromInt=o;function l(S,P){if(isNaN(S))return P?v:x;if(P){if(S<0)return v;if(S>=g)return A}else{if(S<=-y)return M;if(S+1>=y)return E}return S<0?l(-S,P).neg():u(S%f|0,S/f|0,P)}a.fromNumber=l;function u(S,P,V){return new a(S,P,V)}a.fromBits=u;var p=Math.pow;function d(S,P,V){if(S.length===0)throw Error("empty string");if(S==="NaN"||S==="Infinity"||S==="+Infinity"||S==="-Infinity")return x;if(typeof P=="number"?(V=P,P=!1):P=!!P,V=V||10,V<2||36<V)throw RangeError("radix");var j;if((j=S.indexOf("-"))>0)throw Error("interior hyphen");if(j===0)return d(S.substring(1),P,V).neg();for(var q=l(p(V,8)),K=x,Z=0;Z<S.length;Z+=8){var ee=Math.min(8,S.length-Z),re=parseInt(S.substring(Z,Z+ee),V);if(ee<8){var Y=l(p(V,ee));K=K.mul(Y).add(l(re))}else K=K.mul(q),K=K.add(l(re))}return K.unsigned=P,K}a.fromString=d;function c(S,P){return typeof S=="number"?l(S,P):typeof S=="string"?d(S,P):u(S.low,S.high,typeof P=="boolean"?P:S.unsigned)}a.fromValue=c;var h=1<<16,m=1<<24,f=h*h,g=f*f,y=g/2,b=o(m),x=o(0);a.ZERO=x;var v=o(0,!0);a.UZERO=v;var k=o(1);a.ONE=k;var T=o(1,!0);a.UONE=T;var _=o(-1);a.NEG_ONE=_;var E=u(-1,2147483647,!1);a.MAX_VALUE=E;var A=u(-1,-1,!0);a.MAX_UNSIGNED_VALUE=A;var M=u(0,-2147483648,!1);a.MIN_VALUE=M;var $=a.prototype;$.toInt=function(){return this.unsigned?this.low>>>0:this.low},$.toNumber=function(){return this.unsigned?(this.high>>>0)*f+(this.low>>>0):this.high*f+(this.low>>>0)},$.toString=function(S){if(S=S||10,S<2||36<S)throw RangeError("radix");if(this.isZero())return"0";if(this.isNegative())if(this.eq(M)){var P=l(S),V=this.div(P),j=V.mul(P).sub(this);return V.toString(S)+j.toInt().toString(S)}else return"-"+this.neg().toString(S);for(var q=l(p(S,6),this.unsigned),K=this,Z="";;){var ee=K.div(q),re=K.sub(ee.mul(q)).toInt()>>>0,Y=re.toString(S);if(K=ee,K.isZero())return Y+Z;for(;Y.length<6;)Y="0"+Y;Z=""+Y+Z}},$.getHighBits=function(){return this.high},$.getHighBitsUnsigned=function(){return this.high>>>0},$.getLowBits=function(){return this.low},$.getLowBitsUnsigned=function(){return this.low>>>0},$.getNumBitsAbs=function(){if(this.isNegative())return this.eq(M)?64:this.neg().getNumBitsAbs();for(var S=this.high!=0?this.high:this.low,P=31;P>0&&(S&1<<P)==0;P--);return this.high!=0?P+33:P+1},$.isZero=function(){return this.high===0&&this.low===0},$.eqz=$.isZero,$.isNegative=function(){return!this.unsigned&&this.high<0},$.isPositive=function(){return this.unsigned||this.high>=0},$.isOdd=function(){return(this.low&1)===1},$.isEven=function(){return(this.low&1)===0},$.equals=function(S){return r(S)||(S=c(S)),this.unsigned!==S.unsigned&&this.high>>>31===1&&S.high>>>31===1?!1:this.high===S.high&&this.low===S.low},$.eq=$.equals,$.notEquals=function(S){return!this.eq(S)},$.neq=$.notEquals,$.ne=$.notEquals,$.lessThan=function(S){return this.comp(S)<0},$.lt=$.lessThan,$.lessThanOrEqual=function(S){return this.comp(S)<=0},$.lte=$.lessThanOrEqual,$.le=$.lessThanOrEqual,$.greaterThan=function(S){return this.comp(S)>0},$.gt=$.greaterThan,$.greaterThanOrEqual=function(S){return this.comp(S)>=0},$.gte=$.greaterThanOrEqual,$.ge=$.greaterThanOrEqual,$.compare=function(S){if(r(S)||(S=c(S)),this.eq(S))return 0;var P=this.isNegative(),V=S.isNegative();return P&&!V?-1:!P&&V?1:this.unsigned?S.high>>>0>this.high>>>0||S.high===this.high&&S.low>>>0>this.low>>>0?-1:1:this.sub(S).isNegative()?-1:1},$.comp=$.compare,$.negate=function(){return!this.unsigned&&this.eq(M)?M:this.not().add(k)},$.neg=$.negate,$.add=function(S){r(S)||(S=c(S));var P=this.high>>>16,V=this.high&65535,j=this.low>>>16,q=this.low&65535,K=S.high>>>16,Z=S.high&65535,ee=S.low>>>16,re=S.low&65535,Y=0,ie=0,ae=0,le=0;return le+=q+re,ae+=le>>>16,le&=65535,ae+=j+ee,ie+=ae>>>16,ae&=65535,ie+=V+Z,Y+=ie>>>16,ie&=65535,Y+=P+K,Y&=65535,u(ae<<16|le,Y<<16|ie,this.unsigned)},$.subtract=function(S){return r(S)||(S=c(S)),this.add(S.neg())},$.sub=$.subtract,$.multiply=function(S){if(this.isZero())return x;if(r(S)||(S=c(S)),n){var P=n.mul(this.low,this.high,S.low,S.high);return u(P,n.get_high(),this.unsigned)}if(S.isZero())return x;if(this.eq(M))return S.isOdd()?M:x;if(S.eq(M))return this.isOdd()?M:x;if(this.isNegative())return S.isNegative()?this.neg().mul(S.neg()):this.neg().mul(S).neg();if(S.isNegative())return this.mul(S.neg()).neg();if(this.lt(b)&&S.lt(b))return l(this.toNumber()*S.toNumber(),this.unsigned);var V=this.high>>>16,j=this.high&65535,q=this.low>>>16,K=this.low&65535,Z=S.high>>>16,ee=S.high&65535,re=S.low>>>16,Y=S.low&65535,ie=0,ae=0,le=0,ue=0;return ue+=K*Y,le+=ue>>>16,ue&=65535,le+=q*Y,ae+=le>>>16,le&=65535,le+=K*re,ae+=le>>>16,le&=65535,ae+=j*Y,ie+=ae>>>16,ae&=65535,ae+=q*re,ie+=ae>>>16,ae&=65535,ae+=K*ee,ie+=ae>>>16,ae&=65535,ie+=V*Y+j*re+q*ee+K*Z,ie&=65535,u(le<<16|ue,ie<<16|ae,this.unsigned)},$.mul=$.multiply,$.divide=function(S){if(r(S)||(S=c(S)),S.isZero())throw Error("division by zero");if(n){if(!this.unsigned&&this.high===-2147483648&&S.low===-1&&S.high===-1)return this;var P=(this.unsigned?n.div_u:n.div_s)(this.low,this.high,S.low,S.high);return u(P,n.get_high(),this.unsigned)}if(this.isZero())return this.unsigned?v:x;var V,j,q;if(this.unsigned){if(S.unsigned||(S=S.toUnsigned()),S.gt(this))return v;if(S.gt(this.shru(1)))return T;q=v}else{if(this.eq(M)){if(S.eq(k)||S.eq(_))return M;if(S.eq(M))return k;var K=this.shr(1);return V=K.div(S).shl(1),V.eq(x)?S.isNegative()?k:_:(j=this.sub(S.mul(V)),q=V.add(j.div(S)),q)}else if(S.eq(M))return this.unsigned?v:x;if(this.isNegative())return S.isNegative()?this.neg().div(S.neg()):this.neg().div(S).neg();if(S.isNegative())return this.div(S.neg()).neg();q=x}for(j=this;j.gte(S);){V=Math.max(1,Math.floor(j.toNumber()/S.toNumber()));for(var Z=Math.ceil(Math.log(V)/Math.LN2),ee=Z<=48?1:p(2,Z-48),re=l(V),Y=re.mul(S);Y.isNegative()||Y.gt(j);)V-=ee,re=l(V,this.unsigned),Y=re.mul(S);re.isZero()&&(re=k),q=q.add(re),j=j.sub(Y)}return q},$.div=$.divide,$.modulo=function(S){if(r(S)||(S=c(S)),n){var P=(this.unsigned?n.rem_u:n.rem_s)(this.low,this.high,S.low,S.high);return u(P,n.get_high(),this.unsigned)}return this.sub(this.div(S).mul(S))},$.mod=$.modulo,$.rem=$.modulo,$.not=function(){return u(~this.low,~this.high,this.unsigned)},$.and=function(S){return r(S)||(S=c(S)),u(this.low&S.low,this.high&S.high,this.unsigned)},$.or=function(S){return r(S)||(S=c(S)),u(this.low|S.low,this.high|S.high,this.unsigned)},$.xor=function(S){return r(S)||(S=c(S)),u(this.low^S.low,this.high^S.high,this.unsigned)},$.shiftLeft=function(S){return r(S)&&(S=S.toInt()),(S&=63)===0?this:S<32?u(this.low<<S,this.high<<S|this.low>>>32-S,this.unsigned):u(0,this.low<<S-32,this.unsigned)},$.shl=$.shiftLeft,$.shiftRight=function(S){return r(S)&&(S=S.toInt()),(S&=63)===0?this:S<32?u(this.low>>>S|this.high<<32-S,this.high>>S,this.unsigned):u(this.high>>S-32,this.high>=0?0:-1,this.unsigned)},$.shr=$.shiftRight,$.shiftRightUnsigned=function(S){if(r(S)&&(S=S.toInt()),S&=63,S===0)return this;var P=this.high;if(S<32){var V=this.low;return u(V>>>S|P<<32-S,P>>>S,this.unsigned)}else return S===32?u(P,0,this.unsigned):u(P>>>S-32,0,this.unsigned)},$.shru=$.shiftRightUnsigned,$.shr_u=$.shiftRightUnsigned,$.toSigned=function(){return this.unsigned?u(this.low,this.high,!1):this},$.toUnsigned=function(){return this.unsigned?this:u(this.low,this.high,!0)},$.toBytes=function(S){return S?this.toBytesLE():this.toBytesBE()},$.toBytesLE=function(){var S=this.high,P=this.low;return[P&255,P>>>8&255,P>>>16&255,P>>>24,S&255,S>>>8&255,S>>>16&255,S>>>24]},$.toBytesBE=function(){var S=this.high,P=this.low;return[S>>>24,S>>>16&255,S>>>8&255,S&255,P>>>24,P>>>16&255,P>>>8&255,P&255]},a.fromBytes=function(S,P,V){return V?a.fromBytesLE(S,P):a.fromBytesBE(S,P)},a.fromBytesLE=function(S,P){return new a(S[0]|S[1]<<8|S[2]<<16|S[3]<<24,S[4]|S[5]<<8|S[6]<<16|S[7]<<24,P)},a.fromBytesBE=function(S,P){return new a(S[4]<<24|S[5]<<16|S[6]<<8|S[7],S[0]<<24|S[1]<<16|S[2]<<8|S[3],P)}}),n$=ft(()=>{}),a$=ft(()=>{}),r$=ft((e,t)=>{(function(n,a,r){function s(u){var p=this,d=l();p.next=function(){var c=2091639*p.s0+p.c*23283064365386963e-26;return p.s0=p.s1,p.s1=p.s2,p.s2=c-(p.c=c|0)},p.c=1,p.s0=d(" "),p.s1=d(" "),p.s2=d(" "),p.s0-=d(u),p.s0<0&&(p.s0+=1),p.s1-=d(u),p.s1<0&&(p.s1+=1),p.s2-=d(u),p.s2<0&&(p.s2+=1),d=null}function i(u,p){return p.c=u.c,p.s0=u.s0,p.s1=u.s1,p.s2=u.s2,p}function o(u,p){var d=new s(u),c=p&&p.state,h=d.next;return h.int32=function(){return d.next()*4294967296|0},h.double=function(){return h()+(h()*2097152|0)*11102230246251565e-32},h.quick=h,c&&(typeof c=="object"&&i(c,d),h.state=function(){return i(d,{})}),h}function l(){var u=4022871197,p=function(d){d=d.toString();for(var c=0;c<d.length;c++){u+=d.charCodeAt(c);var h=.02519603282416938*u;u=h>>>0,h-=u,h*=u,u=h>>>0,h-=u,u+=h*4294967296}return(u>>>0)*23283064365386963e-26};return p}a&&a.exports?a.exports=o:r&&r.amd?r(function(){return o}):this.alea=o})(e,typeof t=="object"&&t,typeof define=="function"&&define)}),s$=ft((e,t)=>{(function(n,a,r){function s(l){var u=this,p="";u.x=0,u.y=0,u.z=0,u.w=0,u.next=function(){var c=u.x^u.x<<11;return u.x=u.y,u.y=u.z,u.z=u.w,u.w^=u.w>>>19^c^c>>>8},l===(l|0)?u.x=l:p+=l;for(var d=0;d<p.length+64;d++)u.x^=p.charCodeAt(d)|0,u.next()}function i(l,u){return u.x=l.x,u.y=l.y,u.z=l.z,u.w=l.w,u}function o(l,u){var p=new s(l),d=u&&u.state,c=function(){return(p.next()>>>0)/4294967296};return c.double=function(){do var h=p.next()>>>11,m=(p.next()>>>0)/4294967296,f=(h+m)/(1<<21);while(f===0);return f},c.int32=p.next,c.quick=c,d&&(typeof d=="object"&&i(d,p),c.state=function(){return i(p,{})}),c}a&&a.exports?a.exports=o:r&&r.amd?r(function(){return o}):this.xor128=o})(e,typeof t=="object"&&t,typeof define=="function"&&define)}),i$=ft((e,t)=>{(function(n,a,r){function s(l){var u=this,p="";u.next=function(){var c=u.x^u.x>>>2;return u.x=u.y,u.y=u.z,u.z=u.w,u.w=u.v,(u.d=u.d+362437|0)+(u.v=u.v^u.v<<4^(c^c<<1))|0},u.x=0,u.y=0,u.z=0,u.w=0,u.v=0,l===(l|0)?u.x=l:p+=l;for(var d=0;d<p.length+64;d++)u.x^=p.charCodeAt(d)|0,d==p.length&&(u.d=u.x<<10^u.x>>>4),u.next()}function i(l,u){return u.x=l.x,u.y=l.y,u.z=l.z,u.w=l.w,u.v=l.v,u.d=l.d,u}function o(l,u){var p=new s(l),d=u&&u.state,c=function(){return(p.next()>>>0)/4294967296};return c.double=function(){do var h=p.next()>>>11,m=(p.next()>>>0)/4294967296,f=(h+m)/(1<<21);while(f===0);return f},c.int32=p.next,c.quick=c,d&&(typeof d=="object"&&i(d,p),c.state=function(){return i(p,{})}),c}a&&a.exports?a.exports=o:r&&r.amd?r(function(){return o}):this.xorwow=o})(e,typeof t=="object"&&t,typeof define=="function"&&define)}),o$=ft((e,t)=>{(function(n,a,r){function s(l){var u=this;u.next=function(){var d=u.x,c=u.i,h,m,f;return h=d[c],h^=h>>>7,m=h^h<<24,h=d[c+1&7],m^=h^h>>>10,h=d[c+3&7],m^=h^h>>>3,h=d[c+4&7],m^=h^h<<7,h=d[c+7&7],h=h^h<<13,m^=h^h<<9,d[c]=m,u.i=c+1&7,m};function p(d,c){var h,m,f=[];if(c===(c|0))m=f[0]=c;else for(c=""+c,h=0;h<c.length;++h)f[h&7]=f[h&7]<<15^c.charCodeAt(h)+f[h+1&7]<<13;for(;f.length<8;)f.push(0);for(h=0;h<8&&f[h]===0;++h);for(h==8?m=f[7]=-1:m=f[h],d.x=f,d.i=0,h=256;h>0;--h)d.next()}p(u,l)}function i(l,u){return u.x=l.x.slice(),u.i=l.i,u}function o(l,u){l==null&&(l=+new Date);var p=new s(l),d=u&&u.state,c=function(){return(p.next()>>>0)/4294967296};return c.double=function(){do var h=p.next()>>>11,m=(p.next()>>>0)/4294967296,f=(h+m)/(1<<21);while(f===0);return f},c.int32=p.next,c.quick=c,d&&(d.x&&i(d,p),c.state=function(){return i(p,{})}),c}a&&a.exports?a.exports=o:r&&r.amd?r(function(){return o}):this.xorshift7=o})(e,typeof t=="object"&&t,typeof define=="function"&&define)}),l$=ft((e,t)=>{(function(n,a,r){function s(l){var u=this;u.next=function(){var d=u.w,c=u.X,h=u.i,m,f;return u.w=d=d+1640531527|0,f=c[h+34&127],m=c[h=h+1&127],f^=f<<13,m^=m<<17,f^=f>>>15,m^=m>>>12,f=c[h]=f^m,u.i=h,f+(d^d>>>16)|0};function p(d,c){var h,m,f,g,y,b=[],x=128;for(c===(c|0)?(m=c,c=null):(c=c+"\0",m=0,x=Math.max(x,c.length)),f=0,g=-32;g<x;++g)c&&(m^=c.charCodeAt((g+32)%c.length)),g===0&&(y=m),m^=m<<10,m^=m>>>15,m^=m<<4,m^=m>>>13,g>=0&&(y=y+1640531527|0,h=b[g&127]^=m+y,f=h==0?f+1:0);for(f>=128&&(b[(c&&c.length||0)&127]=-1),f=127,g=4*128;g>0;--g)m=b[f+34&127],h=b[f=f+1&127],m^=m<<13,h^=h<<17,m^=m>>>15,h^=h>>>12,b[f]=m^h;d.w=y,d.X=b,d.i=f}p(u,l)}function i(l,u){return u.i=l.i,u.w=l.w,u.X=l.X.slice(),u}function o(l,u){l==null&&(l=+new Date);var p=new s(l),d=u&&u.state,c=function(){return(p.next()>>>0)/4294967296};return c.double=function(){do var h=p.next()>>>11,m=(p.next()>>>0)/4294967296,f=(h+m)/(1<<21);while(f===0);return f},c.int32=p.next,c.quick=c,d&&(d.X&&i(d,p),c.state=function(){return i(p,{})}),c}a&&a.exports?a.exports=o:r&&r.amd?r(function(){return o}):this.xor4096=o})(e,typeof t=="object"&&t,typeof define=="function"&&define)}),u$=ft((e,t)=>{(function(n,a,r){function s(l){var u=this,p="";u.next=function(){var c=u.b,h=u.c,m=u.d,f=u.a;return c=c<<25^c>>>7^h,h=h-m|0,m=m<<24^m>>>8^f,f=f-c|0,u.b=c=c<<20^c>>>12^h,u.c=h=h-m|0,u.d=m<<16^h>>>16^f,u.a=f-c|0},u.a=0,u.b=0,u.c=-1640531527,u.d=1367130551,l===Math.floor(l)?(u.a=l/4294967296|0,u.b=l|0):p+=l;for(var d=0;d<p.length+20;d++)u.b^=p.charCodeAt(d)|0,u.next()}function i(l,u){return u.a=l.a,u.b=l.b,u.c=l.c,u.d=l.d,u}function o(l,u){var p=new s(l),d=u&&u.state,c=function(){return(p.next()>>>0)/4294967296};return c.double=function(){do var h=p.next()>>>11,m=(p.next()>>>0)/4294967296,f=(h+m)/(1<<21);while(f===0);return f},c.int32=p.next,c.quick=c,d&&(typeof d=="object"&&i(d,p),c.state=function(){return i(p,{})}),c}a&&a.exports?a.exports=o:r&&r.amd?r(function(){return o}):this.tychei=o})(e,typeof t=="object"&&t,typeof define=="function"&&define)}),yI=ft(()=>{}),p$=ft((e,t)=>{(function(n,a){var r=this,s=256,i=6,o=52,l="random",u=a.pow(s,i),p=a.pow(2,o),d=p*2,c=s-1,h;function m(k,T,_){var E=[];T=T==!0?{entropy:!0}:T||{};var A=b(y(T.entropy?[k,v(n)]:k==null?x():k,3),E),M=new f(E),$=function(){for(var S=M.g(i),P=u,V=0;S<p;)S=(S+V)*s,P*=s,V=M.g(1);for(;S>=d;)S/=2,P/=2,V>>>=1;return(S+V)/P};return $.int32=function(){return M.g(4)|0},$.quick=function(){return M.g(4)/4294967296},$.double=$,b(v(M.S),n),(T.pass||_||function(S,P,V,j){return j&&(j.S&&g(j,M),S.state=function(){return g(M,{})}),V?(a[l]=S,P):S})($,A,"global"in T?T.global:this==a,T.state)}a["seed"+l]=m;function f(k){var T,_=k.length,E=this,A=0,M=E.i=E.j=0,$=E.S=[];for(_||(k=[_++]);A<s;)$[A]=A++;for(A=0;A<s;A++)$[A]=$[M=c&M+k[A%_]+(T=$[A])],$[M]=T;(E.g=function(S){for(var P,V=0,j=E.i,q=E.j,K=E.S;S--;)P=K[j=c&j+1],V=V*s+K[c&(K[j]=K[q=c&q+P])+(K[q]=P)];return E.i=j,E.j=q,V})(s)}function g(k,T){return T.i=k.i,T.j=k.j,T.S=k.S.slice(),T}function y(k,T){var _=[],E=typeof k,A;if(T&&E=="object")for(A in k)try{_.push(y(k[A],T-1))}catch(M){}return _.length?_:E=="string"?k:k+"\0"}function b(k,T){for(var _=k+"",E,A=0;A<_.length;)T[c&A]=c&(E^=T[c&A]*19)+_.charCodeAt(A++);return v(T)}function x(){try{var k;return h&&(k=h.randomBytes)?k=k(s):(k=new Uint8Array(s),(r.crypto||r.msCrypto).getRandomValues(k)),v(k)}catch(E){var T=r.navigator,_=T&&T.plugins;return[+new Date,r,_,r.screen,v(n)]}}function v(k){return String.fromCharCode.apply(0,k)}if(b(a.random(),n),typeof t=="object"&&t.exports){t.exports=m;try{h=yI()}catch(k){}}else typeof define=="function"&&define.amd&&define(function(){return m})})([],Math)}),bI=ft((e,t)=>{var n=r$(),a=s$(),r=i$(),s=o$(),i=l$(),o=u$(),l=p$();l.alea=n,l.xor128=a,l.xorwow=r,l.xorshift7=s,l.xor4096=i,l.tychei=o,t.exports=l}),c$=ft((e,t)=>{(function(n,a,r){function s(u){var p=this,d=l();p.next=function(){var c=2091639*p.s0+p.c*23283064365386963e-26;return p.s0=p.s1,p.s1=p.s2,p.s2=c-(p.c=c|0)},p.c=1,p.s0=d(" "),p.s1=d(" "),p.s2=d(" "),p.s0-=d(u),p.s0<0&&(p.s0+=1),p.s1-=d(u),p.s1<0&&(p.s1+=1),p.s2-=d(u),p.s2<0&&(p.s2+=1),d=null}function i(u,p){return p.c=u.c,p.s0=u.s0,p.s1=u.s1,p.s2=u.s2,p}function o(u,p){var d=new s(u),c=p&&p.state,h=d.next;return h.int32=function(){return d.next()*4294967296|0},h.double=function(){return h()+(h()*2097152|0)*11102230246251565e-32},h.quick=h,c&&(typeof c=="object"&&i(c,d),h.state=function(){return i(d,{})}),h}function l(){var u=4022871197,p=function(d){d=String(d);for(var c=0;c<d.length;c++){u+=d.charCodeAt(c);var h=.02519603282416938*u;u=h>>>0,h-=u,h*=u,u=h>>>0,h-=u,u+=h*4294967296}return(u>>>0)*23283064365386963e-26};return p}a&&a.exports?a.exports=o:r&&r.amd?r(function(){return o}):this.alea=o})(e,typeof t=="object"&&t,typeof define=="function"&&define)}),d$=ft((e,t)=>{(function(n,a,r){function s(l){var u=this,p="";u.x=0,u.y=0,u.z=0,u.w=0,u.next=function(){var c=u.x^u.x<<11;return u.x=u.y,u.y=u.z,u.z=u.w,u.w^=u.w>>>19^c^c>>>8},l===(l|0)?u.x=l:p+=l;for(var d=0;d<p.length+64;d++)u.x^=p.charCodeAt(d)|0,u.next()}function i(l,u){return u.x=l.x,u.y=l.y,u.z=l.z,u.w=l.w,u}function o(l,u){var p=new s(l),d=u&&u.state,c=function(){return(p.next()>>>0)/4294967296};return c.double=function(){do var h=p.next()>>>11,m=(p.next()>>>0)/4294967296,f=(h+m)/(1<<21);while(f===0);return f},c.int32=p.next,c.quick=c,d&&(typeof d=="object"&&i(d,p),c.state=function(){return i(p,{})}),c}a&&a.exports?a.exports=o:r&&r.amd?r(function(){return o}):this.xor128=o})(e,typeof t=="object"&&t,typeof define=="function"&&define)}),h$=ft((e,t)=>{(function(n,a,r){function s(l){var u=this,p="";u.next=function(){var c=u.x^u.x>>>2;return u.x=u.y,u.y=u.z,u.z=u.w,u.w=u.v,(u.d=u.d+362437|0)+(u.v=u.v^u.v<<4^(c^c<<1))|0},u.x=0,u.y=0,u.z=0,u.w=0,u.v=0,l===(l|0)?u.x=l:p+=l;for(var d=0;d<p.length+64;d++)u.x^=p.charCodeAt(d)|0,d==p.length&&(u.d=u.x<<10^u.x>>>4),u.next()}function i(l,u){return u.x=l.x,u.y=l.y,u.z=l.z,u.w=l.w,u.v=l.v,u.d=l.d,u}function o(l,u){var p=new s(l),d=u&&u.state,c=function(){return(p.next()>>>0)/4294967296};return c.double=function(){do var h=p.next()>>>11,m=(p.next()>>>0)/4294967296,f=(h+m)/(1<<21);while(f===0);return f},c.int32=p.next,c.quick=c,d&&(typeof d=="object"&&i(d,p),c.state=function(){return i(p,{})}),c}a&&a.exports?a.exports=o:r&&r.amd?r(function(){return o}):this.xorwow=o})(e,typeof t=="object"&&t,typeof define=="function"&&define)}),m$=ft((e,t)=>{(function(n,a,r){function s(l){var u=this;u.next=function(){var d=u.x,c=u.i,h,m,f;return h=d[c],h^=h>>>7,m=h^h<<24,h=d[c+1&7],m^=h^h>>>10,h=d[c+3&7],m^=h^h>>>3,h=d[c+4&7],m^=h^h<<7,h=d[c+7&7],h=h^h<<13,m^=h^h<<9,d[c]=m,u.i=c+1&7,m};function p(d,c){var h,m,f=[];if(c===(c|0))m=f[0]=c;else for(c=""+c,h=0;h<c.length;++h)f[h&7]=f[h&7]<<15^c.charCodeAt(h)+f[h+1&7]<<13;for(;f.length<8;)f.push(0);for(h=0;h<8&&f[h]===0;++h);for(h==8?m=f[7]=-1:m=f[h],d.x=f,d.i=0,h=256;h>0;--h)d.next()}p(u,l)}function i(l,u){return u.x=l.x.slice(),u.i=l.i,u}function o(l,u){l==null&&(l=+new Date);var p=new s(l),d=u&&u.state,c=function(){return(p.next()>>>0)/4294967296};return c.double=function(){do var h=p.next()>>>11,m=(p.next()>>>0)/4294967296,f=(h+m)/(1<<21);while(f===0);return f},c.int32=p.next,c.quick=c,d&&(d.x&&i(d,p),c.state=function(){return i(p,{})}),c}a&&a.exports?a.exports=o:r&&r.amd?r(function(){return o}):this.xorshift7=o})(e,typeof t=="object"&&t,typeof define=="function"&&define)}),f$=ft((e,t)=>{(function(n,a,r){function s(l){var u=this;u.next=function(){var d=u.w,c=u.X,h=u.i,m,f;return u.w=d=d+1640531527|0,f=c[h+34&127],m=c[h=h+1&127],f^=f<<13,m^=m<<17,f^=f>>>15,m^=m>>>12,f=c[h]=f^m,u.i=h,f+(d^d>>>16)|0};function p(d,c){var h,m,f,g,y,b=[],x=128;for(c===(c|0)?(m=c,c=null):(c=c+"\0",m=0,x=Math.max(x,c.length)),f=0,g=-32;g<x;++g)c&&(m^=c.charCodeAt((g+32)%c.length)),g===0&&(y=m),m^=m<<10,m^=m>>>15,m^=m<<4,m^=m>>>13,g>=0&&(y=y+1640531527|0,h=b[g&127]^=m+y,f=h==0?f+1:0);for(f>=128&&(b[(c&&c.length||0)&127]=-1),f=127,g=4*128;g>0;--g)m=b[f+34&127],h=b[f=f+1&127],m^=m<<13,h^=h<<17,m^=m>>>15,h^=h>>>12,b[f]=m^h;d.w=y,d.X=b,d.i=f}p(u,l)}function i(l,u){return u.i=l.i,u.w=l.w,u.X=l.X.slice(),u}function o(l,u){l==null&&(l=+new Date);var p=new s(l),d=u&&u.state,c=function(){return(p.next()>>>0)/4294967296};return c.double=function(){do var h=p.next()>>>11,m=(p.next()>>>0)/4294967296,f=(h+m)/(1<<21);while(f===0);return f},c.int32=p.next,c.quick=c,d&&(d.X&&i(d,p),c.state=function(){return i(p,{})}),c}a&&a.exports?a.exports=o:r&&r.amd?r(function(){return o}):this.xor4096=o})(e,typeof t=="object"&&t,typeof define=="function"&&define)}),g$=ft((e,t)=>{(function(n,a,r){function s(l){var u=this,p="";u.next=function(){var c=u.b,h=u.c,m=u.d,f=u.a;return c=c<<25^c>>>7^h,h=h-m|0,m=m<<24^m>>>8^f,f=f-c|0,u.b=c=c<<20^c>>>12^h,u.c=h=h-m|0,u.d=m<<16^h>>>16^f,u.a=f-c|0},u.a=0,u.b=0,u.c=-1640531527,u.d=1367130551,l===Math.floor(l)?(u.a=l/4294967296|0,u.b=l|0):p+=l;for(var d=0;d<p.length+20;d++)u.b^=p.charCodeAt(d)|0,u.next()}function i(l,u){return u.a=l.a,u.b=l.b,u.c=l.c,u.d=l.d,u}function o(l,u){var p=new s(l),d=u&&u.state,c=function(){return(p.next()>>>0)/4294967296};return c.double=function(){do var h=p.next()>>>11,m=(p.next()>>>0)/4294967296,f=(h+m)/(1<<21);while(f===0);return f},c.int32=p.next,c.quick=c,d&&(typeof d=="object"&&i(d,p),c.state=function(){return i(p,{})}),c}a&&a.exports?a.exports=o:r&&r.amd?r(function(){return o}):this.tychei=o})(e,typeof t=="object"&&t,typeof define=="function"&&define)}),y$=ft((e,t)=>{(function(n,a,r){var s=256,i=6,o=52,l="random",u=r.pow(s,i),p=r.pow(2,o),d=p*2,c=s-1,h;function m(k,T,_){var E=[];T=T==!0?{entropy:!0}:T||{};var A=b(y(T.entropy?[k,v(a)]:k==null?x():k,3),E),M=new f(E),$=function(){for(var S=M.g(i),P=u,V=0;S<p;)S=(S+V)*s,P*=s,V=M.g(1);for(;S>=d;)S/=2,P/=2,V>>>=1;return(S+V)/P};return $.int32=function(){return M.g(4)|0},$.quick=function(){return M.g(4)/4294967296},$.double=$,b(v(M.S),a),(T.pass||_||function(S,P,V,j){return j&&(j.S&&g(j,M),S.state=function(){return g(M,{})}),V?(r[l]=S,P):S})($,A,"global"in T?T.global:this==r,T.state)}function f(k){var T,_=k.length,E=this,A=0,M=E.i=E.j=0,$=E.S=[];for(_||(k=[_++]);A<s;)$[A]=A++;for(A=0;A<s;A++)$[A]=$[M=c&M+k[A%_]+(T=$[A])],$[M]=T;(E.g=function(S){for(var P,V=0,j=E.i,q=E.j,K=E.S;S--;)P=K[j=c&j+1],V=V*s+K[c&(K[j]=K[q=c&q+P])+(K[q]=P)];return E.i=j,E.j=q,V})(s)}function g(k,T){return T.i=k.i,T.j=k.j,T.S=k.S.slice(),T}function y(k,T){var _=[],E=typeof k,A;if(T&&E=="object")for(A in k)try{_.push(y(k[A],T-1))}catch(M){}return _.length?_:E=="string"?k:k+"\0"}function b(k,T){for(var _=k+"",E,A=0;A<_.length;)T[c&A]=c&(E^=T[c&A]*19)+_.charCodeAt(A++);return v(T)}function x(){try{var k;return h&&(k=h.randomBytes)?k=k(s):(k=new Uint8Array(s),(n.crypto||n.msCrypto).getRandomValues(k)),v(k)}catch(E){var T=n.navigator,_=T&&T.plugins;return[+new Date,n,_,n.screen,v(a)]}}function v(k){return String.fromCharCode.apply(0,k)}if(b(r.random(),a),typeof t=="object"&&t.exports){t.exports=m;try{h=yI()}catch(k){}}else typeof define=="function"&&define.amd?define(function(){return m}):r["seed"+l]=m})(typeof self!="undefined"?self:e,[],Math)}),xI=ft((e,t)=>{var n=c$(),a=d$(),r=h$(),s=m$(),i=f$(),o=g$(),l=y$();l.alea=n,l.xor128=a,l.xorwow=r,l.xorshift7=s,l.xor4096=i,l.tychei=o,t.exports=l}),vI=ft(()=>{}),Fx=ft(()=>{}),Wh=ft(()=>{}),b$=ft(()=>{}),x$=ft(()=>{}),v$=ft(()=>{}),w$=ft((e,t)=>{var n=(()=>{var a=typeof document!="undefined"&&document.currentScript?document.currentScript.src:void 0;return typeof __filename!="undefined"&&(a=a||__filename),function(r){r=r||{};function s(){return Te.buffer!=xn&&Ra(Te.buffer),Od}function i(){return Te.buffer!=xn&&Ra(Te.buffer),Ld}function o(){return Te.buffer!=xn&&Ra(Te.buffer),Ip}function l(){return Te.buffer!=xn&&Ra(Te.buffer),zd}function u(){return Te.buffer!=xn&&Ra(Te.buffer),Bd}function p(){return Te.buffer!=xn&&Ra(Te.buffer),Wd}function d(){return Te.buffer!=xn&&Ra(Te.buffer),Vd}var c=typeof r!="undefined"?r:{},h,m;c.ready=new Promise(function(N,D){h=N,m=D});var f;typeof process!="undefined"&&process.listeners&&(f={uncaughtException:process.listeners("uncaughtException"),unhandledRejection:process.listeners("unhandledRejection")});var g=Object.assign({},c),y=[],b="./this.program",x=(N,D)=>{throw D},v=typeof window=="object",k=typeof importScripts=="function",T=typeof process=="object"&&typeof process.versions=="object"&&typeof process.versions.node=="string",_=c.ENVIRONMENT_IS_PTHREAD||!1,E="";function A(N){return c.locateFile?c.locateFile(N,E):E+N}var M,$,S,P;function V(N){N instanceof Dp||Y("exiting due to exception: "+N)}var j,q,K;if(T){k?E=Wh().dirname(E)+"/":E=__dirname+"/",K=()=>{q||(j=Fx(),q=Wh())},M=function(D,U){return K(),D=q.normalize(D),j.readFileSync(D,U?void 0:"utf8")},S=D=>{var U=M(D,!0);return U.buffer||(U=new Uint8Array(U)),U},$=(D,U,Q)=>{K(),D=q.normalize(D),j.readFile(D,function(pe,he){pe?Q(pe):U(he.buffer)})},process.argv.length>1&&(b=process.argv[1].replace(/\\/g,"/")),y=process.argv.slice(2),process.on("uncaughtException",function(D){if(!(D instanceof Dp))throw D}),process.on("unhandledRejection",function(D){throw D}),x=(D,U)=>{if(Os())throw process.exitCode=D,U;V(U),process.exit(D)},c.inspect=function(){return"[Emscripten Module object]"};let N;try{N=b$()}catch(D){throw console.error('The "worker_threads" module is not supported in this node.js build - perhaps a newer version is needed?'),D}global.Worker=N.Worker}else(v||k)&&(k?E=self.location.href:typeof document!="undefined"&&document.currentScript&&(E=document.currentScript.src),typeof a!="undefined"&&a&&(E=a),E.indexOf("blob:")!==0?E=E.substr(0,E.replace(/[?#].*/,"").lastIndexOf("/")+1):E="",T||(M=N=>{var D=new XMLHttpRequest;return D.open("GET",N,!1),D.send(null),D.responseText},k&&(S=N=>{var D=new XMLHttpRequest;return D.open("GET",N,!1),D.responseType="arraybuffer",D.send(null),new Uint8Array(D.response)}),$=(N,D,U)=>{var Q=new XMLHttpRequest;Q.open("GET",N,!0),Q.responseType="arraybuffer",Q.onload=()=>{if(Q.status==200||Q.status==0&&Q.response){D(Q.response);return}U()},Q.onerror=U,Q.send(null)}),P=N=>document.title=N);T&&typeof performance=="undefined"&&(global.performance=x$().performance);var Z=console.log.bind(console),ee=console.warn.bind(console);T&&(K(),Z=N=>j.writeSync(1,N+`
`),ee=N=>j.writeSync(2,N+`
`));var re=c.print||Z,Y=c.printErr||ee;Object.assign(c,g),g=null,c.arguments&&(y=c.arguments),c.thisProgram&&(b=c.thisProgram),c.quit&&(x=c.quit);var ie=4;function ae(N){ae.shown||(ae.shown={}),ae.shown[N]||(ae.shown[N]=1,Y(N))}function le(N,D){if(typeof WebAssembly.Function=="function"){for(var U={i:"i32",j:"i64",f:"f32",d:"f64"},Q={parameters:[],results:D[0]=="v"?[]:[U[D[0]]]},pe=1;pe<D.length;++pe)Q.parameters.push(U[D[pe]]);return new WebAssembly.Function(Q,N)}var he=[1,0,1,96],ve=D.slice(0,1),_e=D.slice(1),Ct={i:127,j:126,f:125,d:124};he.push(_e.length);for(var pe=0;pe<_e.length;++pe)he.push(Ct[_e[pe]]);ve=="v"?he.push(0):he=he.concat([1,Ct[ve]]),he[1]=he.length-2;var La=new Uint8Array([0,97,115,109,1,0,0,0].concat(he,[2,7,1,1,101,1,102,0,0,7,5,1,1,102,0,0])),za=new WebAssembly.Module(La),gh=new WebAssembly.Instance(za,{e:{f:N}}),Rp=gh.exports.f;return Rp}var ue=[],ke;function ye(){if(ue.length)return ue.pop();try{la.grow(1)}catch(N){throw N instanceof RangeError?"Unable to grow wasm table. Set ALLOW_TABLE_GROWTH.":N}return la.length-1}function Ie(N,D){for(var U=N;U<N+D;U++){var Q=tl(U);Q&&ke.set(Q,U)}}var Ee=0,$e=N=>{Ee=N},Be=Atomics.load,je=Atomics.store,st=Atomics.compareExchange,nt;c.wasmBinary&&(nt=c.wasmBinary);var at=c.noExitRuntime||!0;typeof WebAssembly!="object"&&Jo("no native wasm support detected");var Te,gt,ct=!1,bn;function Qt(N,D){N||Jo(D)}function Dn(N){var D=c["_"+N];return D}function Gt(N,D,U,Q,pe){var he={string:function(ua){var ll=0;if(ua!=null&&ua!==0){var H1=(ua.length<<2)+1;ll=ol(H1),Ms(ua,ll,H1)}return ll},array:function(ua){var ll=ol(ua.length);return kr(ua,ll),ll}};function ve(ua){return D==="string"?oa(ua):D==="boolean"?Boolean(ua):ua}var _e=Dn(N),Ct=[],La=0;if(Q)for(var za=0;za<Q.length;za++){var gh=he[U[za]];gh?(La===0&&(La=hb()),Ct[za]=gh(Q[za])):Ct[za]=Q[za]}var Rp=_e.apply(null,Ct);function VA(ua){return La!==0&&dh(La),ve(ua)}return Rp=VA(Rp),Rp}function Jt(N,D,U,Q){U=U||[];var pe=U.every(function(ve){return ve==="number"}),he=D!=="string";return he&&pe&&!Q?Dn(N):function(){return Gt(N,D,U,arguments,Q)}}var Da=1;function Rn(N){var D=new TextDecoder(N);this.decode=U=>(U.buffer instanceof SharedArrayBuffer&&(U=new Uint8Array(U)),D.decode.call(D,U))}var Ht=typeof TextDecoder!="undefined"?new Rn("utf8"):void 0;function ia(N,D,U){for(var Q=D+U,pe=D;N[pe]&&!(pe>=Q);)++pe;if(pe-D>16&&N.subarray&&Ht)return Ht.decode(N.subarray(D,pe));for(var he="";D<pe;){var ve=N[D++];if(!(ve&128)){he+=String.fromCharCode(ve);continue}var _e=N[D++]&63;if((ve&224)==192){he+=String.fromCharCode((ve&31)<<6|_e);continue}var Ct=N[D++]&63;if((ve&240)==224?ve=(ve&15)<<12|_e<<6|Ct:ve=(ve&7)<<18|_e<<12|Ct<<6|N[D++]&63,ve<65536)he+=String.fromCharCode(ve);else{var La=ve-65536;he+=String.fromCharCode(55296|La>>10,56320|La&1023)}}return he}function oa(N,D){return N?ia(i(),N,D):""}function qr(N,D,U,Q){if(!(Q>0))return 0;for(var pe=U,he=U+Q-1,ve=0;ve<N.length;++ve){var _e=N.charCodeAt(ve);if(_e>=55296&&_e<=57343){var Ct=N.charCodeAt(++ve);_e=65536+((_e&1023)<<10)|Ct&1023}if(_e<=127){if(U>=he)break;D[U++]=_e}else if(_e<=2047){if(U+1>=he)break;D[U++]=192|_e>>6,D[U++]=128|_e&63}else if(_e<=65535){if(U+2>=he)break;D[U++]=224|_e>>12,D[U++]=128|_e>>6&63,D[U++]=128|_e&63}else{if(U+3>=he)break;D[U++]=240|_e>>18,D[U++]=128|_e>>12&63,D[U++]=128|_e>>6&63,D[U++]=128|_e&63}}return D[U]=0,U-pe}function Ms(N,D,U){return qr(N,i(),D,U)}function Pd(N){for(var D=0,U=0;U<N.length;++U){var Q=N.charCodeAt(U);Q>=55296&&Q<=57343&&(Q=65536+((Q&1023)<<10)|N.charCodeAt(++U)&1023),Q<=127?++D:Q<=2047?D+=2:Q<=65535?D+=3:D+=4}return D}var Kr=typeof TextDecoder!="undefined"?new Rn("utf-16le"):void 0;function kr(N,D){s().set(N,D)}function kp(N,D,U){for(var Q=0;Q<N.length;++Q)s()[D++>>0]=N.charCodeAt(Q);U||(s()[D>>0]=0)}function Yo(N,D){return N%D>0&&(N+=D-N%D),N}var xn,Od,Ld,Ip,zd,Bd,N1,Wd,Vd;_&&(xn=c.buffer);function Ra(N){xn=N,c.HEAP8=Od=new Int8Array(N),c.HEAP16=Ip=new Int16Array(N),c.HEAP32=Bd=new Int32Array(N),c.HEAPU8=Ld=new Uint8Array(N),c.HEAPU16=zd=new Uint16Array(N),c.HEAPU32=N1=new Uint32Array(N),c.HEAPF32=Wd=new Float32Array(N),c.HEAPF64=Vd=new Float64Array(N)}var Ud=c.INITIAL_MEMORY||16777216;if(_)Te=c.wasmMemory,xn=c.buffer;else if(c.wasmMemory)Te=c.wasmMemory;else if(Te=new WebAssembly.Memory({initial:Ud/65536,maximum:32768,shared:!0}),!(Te.buffer instanceof SharedArrayBuffer))throw Y("requested a shared WebAssembly.Memory but the returned buffer is not a SharedArrayBuffer, indicating that while the browser has SharedArrayBuffer it does not have WebAssembly threads support - you may need to set a flag"),T&&console.log("(on node you may need: --experimental-wasm-threads --experimental-wasm-bulk-memory and also use a recent version)"),Error("bad memory");Te&&(xn=Te.buffer),Ud=xn.byteLength,Ra(xn);var la,Qo=[],Xr=[],Lg=[],Gd=[],Ps=!1,zg=!1,Hd=0;function Os(){return at||Hd>0}function vn(){if(c.preRun)for(typeof c.preRun=="function"&&(c.preRun=[c.preRun]);c.preRun.length;)T1(c.preRun.shift());Xd(Qo)}function Sp(){Ps=!0,!_&&Xd(Xr)}function Bg(){_||(Ce.terminateAllThreads(),zg=!0)}function Wg(){if(!_){if(c.postRun)for(typeof c.postRun=="function"&&(c.postRun=[c.postRun]);c.postRun.length;)Np(c.postRun.shift());Xd(Gd)}}function T1(N){Qo.unshift(N)}function _1(N){Xr.unshift(N)}function Np(N){Gd.unshift(N)}var Yr=0,jd=null,Ma=null;function Tp(N){Yr++,c.monitorRunDependencies&&c.monitorRunDependencies(Yr)}function C1(N){if(Yr--,c.monitorRunDependencies&&c.monitorRunDependencies(Yr),Yr==0&&(jd!==null&&(clearInterval(jd),jd=null),Ma)){var D=Ma;Ma=null,D()}}c.preloadedImages={},c.preloadedAudios={};function Jo(N){_?postMessage({cmd:"onAbort",arg:N}):c.onAbort&&c.onAbort(N),N="Aborted("+N+")",Y(N),ct=!0,bn=1,N+=". Build with -s ASSERTIONS=1 for more info.";var D=new WebAssembly.RuntimeError(N);throw m(D),D}var Vg="data:application/octet-stream;base64,";function _p(N){return N.startsWith(Vg)}function qd(N){return N.startsWith("file://")}var wn;wn="tfjs-backend-wasm-threaded-simd.wasm",_p(wn)||(wn=A(wn));function Kd(N){try{if(N==wn&&nt)return new Uint8Array(nt);if(S)return S(N);throw"both async and sync fetching of the wasm failed"}catch(D){Jo(D)}}function Zo(){if(!nt&&(v||k)){if(typeof fetch=="function"&&!qd(wn))return fetch(wn,{credentials:"same-origin"}).then(function(N){if(!N.ok)throw"failed to load wasm binary file at '"+wn+"'";return N.arrayBuffer()}).catch(function(){return Kd(wn)});if($)return new Promise(function(N,D){$(wn,function(U){N(new Uint8Array(U))},D)})}return Promise.resolve().then(function(){return Kd(wn)})}function Ug(){var N={env:oh,wasi_snapshot_preview1:oh};function D(ve,_e){var Ct=ve.exports;if(c.asm=Ct,Yg(c.asm.emscripten_tls_init),la=c.asm.__indirect_function_table,_1(c.asm.__wasm_call_ctors),gt=_e,!_){var La=Ce.unusedWorkers.length;Ce.unusedWorkers.forEach(function(za){Ce.loadWasmModuleToWorker(za,function(){--La||C1("wasm-instantiate")})})}}_||Tp("wasm-instantiate");function U(ve){D(ve.instance,ve.module)}function Q(ve){return Zo().then(function(_e){return WebAssembly.instantiate(_e,N)}).then(function(_e){return _e}).then(ve,function(_e){Y("failed to asynchronously prepare wasm: "+_e),Jo(_e)})}function pe(){return!nt&&typeof WebAssembly.instantiateStreaming=="function"&&!_p(wn)&&!qd(wn)&&typeof fetch=="function"?fetch(wn,{credentials:"same-origin"}).then(function(ve){var _e=WebAssembly.instantiateStreaming(ve,N);return _e.then(U,function(Ct){return Y("wasm streaming compile failed: "+Ct),Y("falling back to ArrayBuffer instantiation"),Q(U)})}):Q(U)}if(c.instantiateWasm)try{var he=c.instantiateWasm(N,D);return he}catch(ve){return Y("Module.instantiateWasm callback failed with error: "+ve),!1}return pe().catch(m),{}}var E1,F1,Gg={};function Xd(N){for(;N.length>0;){var D=N.shift();if(typeof D=="function"){D(c);continue}var U=D.func;typeof U=="number"?D.arg===void 0?tl(U)():tl(U)(D.arg):U(D.arg===void 0?null:D.arg)}}function el(N){var D=hb(),U=N();return dh(D),U}function YF(N){return N}function A1(N){var D=/\b_Z[\w\d_]+/g;return N.replace(D,function(U){var Q=U;return U===Q?U:Q+" ["+U+"]"})}function Hg(N){u()[N>>2]=0;var D=Ce.pthreads[N];delete Ce.pthreads[N],D.worker.terminate(),db(N),Ce.runningWorkers.splice(Ce.runningWorkers.indexOf(D.worker),1),D.worker.pthread=void 0}function jg(N){var D=Ce.pthreads[N];D.worker.postMessage({cmd:"cancel"})}function Yd(N){var D=Ce.pthreads[N];if(D){u()[N>>2]=0;var U=D.worker;Ce.returnWorkerToPool(U)}}function Qd(N){zA(N)}function qg(N){if(N instanceof Dp||N=="unwind")return bn;x(1,N)}var Ce={unusedWorkers:[],runningWorkers:[],tlsInitFunctions:[],init:function(){_?Ce.initWorker():Ce.initMainThread()},initMainThread:function(){for(var N=8,D=0;D<N;++D)Ce.allocateUnusedWorker()},initWorker:function(){at=!1},pthreads:{},setExitStatus:function(N){bn=N},terminateAllThreads:function(){for(var N in Ce.pthreads){var D=Ce.pthreads[N];D&&D.worker&&Ce.returnWorkerToPool(D.worker)}for(var U=0;U<Ce.unusedWorkers.length;++U){var Q=Ce.unusedWorkers[U];Q.terminate()}Ce.unusedWorkers=[]},returnWorkerToPool:function(N){Ce.runWithoutMainThreadQueuedCalls(function(){delete Ce.pthreads[N.pthread.threadInfoStruct],Ce.unusedWorkers.push(N),Ce.runningWorkers.splice(Ce.runningWorkers.indexOf(N),1),db(N.pthread.threadInfoStruct),N.pthread=void 0})},runWithoutMainThreadQueuedCalls:function(N){u()[G1>>2]=0;try{N()}finally{u()[G1>>2]=1}},receiveObjectTransfer:function(N){},threadInit:function(){for(var N in Ce.tlsInitFunctions)Ce.tlsInitFunctions[N]()},loadWasmModuleToWorker:function(N,D){N.onmessage=U=>{var Q=U.data,pe=Q.cmd;if(N.pthread&&(Ce.currentProxiedOperationCallerThread=N.pthread.threadInfoStruct),Q.targetThread&&Q.targetThread!=ch()){var he=Ce.pthreads[Q.targetThread];he?he.worker.postMessage(Q,Q.transferList):Y('Internal error! Worker sent a message "'+pe+'" to target pthread '+Q.targetThread+", but that thread no longer exists!"),Ce.currentProxiedOperationCallerThread=void 0;return}pe==="processQueuedMainThreadWork"?z1():pe==="spawnThread"?Zd(Q):pe==="cleanupThread"?Yd(Q.thread):pe==="killThread"?Hg(Q.thread):pe==="cancelThread"?jg(Q.thread):pe==="loaded"?(N.loaded=!0,D&&D(N),N.runPthread&&(N.runPthread(),delete N.runPthread)):pe==="print"?re("Thread "+Q.threadId+": "+Q.text):pe==="printErr"?Y("Thread "+Q.threadId+": "+Q.text):pe==="alert"?alert("Thread "+Q.threadId+": "+Q.text):Q.target==="setimmediate"?N.postMessage(Q):pe==="onAbort"?c.onAbort&&c.onAbort(Q.arg):Y("worker sent an unknown command "+pe),Ce.currentProxiedOperationCallerThread=void 0},N.onerror=U=>{var Q="worker sent an error!";throw Y(Q+" "+U.filename+":"+U.lineno+": "+U.message),U},T&&(N.on("message",function(U){N.onmessage({data:U})}),N.on("error",function(U){N.onerror(U)}),N.on("detachedExit",function(){})),N.postMessage({cmd:"load",urlOrBlob:c.mainScriptUrlOrBlob||a,wasmMemory:Te,wasmModule:gt})},allocateUnusedWorker:function(){var N=A("tfjs-backend-wasm-threaded-simd.worker.js");Ce.unusedWorkers.push(new Worker(N))},getNewWorker:function(){return Ce.unusedWorkers.length==0&&(Ce.allocateUnusedWorker(),Ce.loadWasmModuleToWorker(Ce.unusedWorkers[0])),Ce.unusedWorkers.pop()}};function Kg(){var N=ch(),D=u()[N+44>>2],U=u()[N+48>>2],Q=D-U;U1(D,Q),dh(D)}c.establishStackSpace=Kg;function Jd(N){if(_)return Bs(1,0,N);try{Qd(N)}catch(D){qg(D)}}var Ls=[];function tl(N){var D=Ls[N];return D||(N>=Ls.length&&(Ls.length=N+1),Ls[N]=D=la.get(N)),D}function Xg(N,D){return tl(N)(D)}c.invokeEntryPoint=Xg;function $1(){var N=new Error;if(!N.stack){try{throw new Error}catch(D){N=D}if(!N.stack)return"(no stack trace available)"}return N.stack.toString()}function Yg(N,D,U){Ce.tlsInitFunctions.push(N)}function D1(N,D){la.set(N,D),Ls[N]=D}var zs;T?zs=()=>{var N=process.hrtime();return N[0]*1e3+N[1]/1e6}:_?zs=()=>performance.now()-c.__performance_now_clock_drift:zs=()=>performance.now();var Qg=!0;function Jg(N){return u()[L1()>>2]=N,N}function Zg(N,D){var U;if(N===0)U=Date.now();else if((N===1||N===4)&&Qg)U=zs();else return Jg(28),-1;return u()[D>>2]=U/1e3|0,u()[D+4>>2]=U%1e3*1e3*1e3|0,0}function ey(N,D){return Zg(N,D)}function ty(N){B1(N,!k,1,!v),Ce.threadInit()}function ny(N){_?postMessage({cmd:"cleanupThread",thread:N}):Yd(N)}function Zd(N){var D=Ce.getNewWorker();if(!D)return 6;Ce.runningWorkers.push(D);var U=Ce.pthreads[N.pthread_ptr]={worker:D,threadInfoStruct:N.pthread_ptr};D.pthread=U;var Q={cmd:"run",start_routine:N.startRoutine,arg:N.arg,threadInfoStruct:N.pthread_ptr};return D.runPthread=()=>{Q.time=performance.now(),D.postMessage(Q,N.transferList)},D.loaded&&(D.runPthread(),delete D.runPthread),0}function ay(N,D,U,Q){if(typeof SharedArrayBuffer=="undefined")return Y("Current environment does not support SharedArrayBuffer, pthreads are not available!"),6;var pe=[],he=0;if(_&&(pe.length===0||he))return W1(687865856,N,D,U,Q);if(he)return he;var ve={startRoutine:U,pthread_ptr:N,arg:Q,transferList:pe};return _?(ve.cmd="spawnThread",postMessage(ve,pe),0):Zd(ve)}function ry(){return 2097152}function sy(N,D){if(N==D)postMessage({cmd:"processQueuedMainThreadWork"});else if(_)postMessage({targetThread:N,cmd:"processThreadQueue"});else{var U=Ce.pthreads[N],Q=U&&U.worker;if(!Q)return;Q.postMessage({cmd:"processThreadQueue"})}return 1}function iy(){Jo("")}function oy(){T||k||ae("Blocking on the main thread is very dangerous, see https://emscripten.org/docs/porting/pthreads.html#blocking-on-the-main-browser-thread")}function eh(){return 2147483648}function ly(N,D,U){i().copyWithin(N,D,D+U)}function uy(){return T?v$().cpus().length:navigator.hardwareConcurrency}function Bs(N,D){var U=arguments.length-2,Q=arguments;return el(function(){for(var pe=U,he=ol(pe*8),ve=he>>3,_e=0;_e<U;_e++){var Ct=Q[2+_e];d()[ve+_e]=Ct}return V1(N,pe,he,D)})}var Cp=[];function py(N,D,U){Cp.length=D;for(var Q=U>>3,pe=0;pe<D;pe++)Cp[pe]=d()[Q+pe];var he=N<0,ve=he?Gg[-N-1]:Ey[N];return ve.apply(null,Cp)}function cy(N){try{return Te.grow(N-xn.byteLength+65535>>>16),Ra(Te.buffer),1}catch(D){}}function dy(N){var D=i().length;if(N=N>>>0,N<=D)return!1;var U=eh();if(N>U)return!1;for(var Q=1;Q<=4;Q*=2){var pe=D*(1+.2/Q);pe=Math.min(pe,N+100663296);var he=Math.min(U,Yo(Math.max(N,pe),65536)),ve=cy(he);if(ve)return!0}return!1}var Ue={inEventHandler:0,removeAllEventListeners:function(){for(var N=Ue.eventHandlers.length-1;N>=0;--N)Ue._removeHandler(N);Ue.eventHandlers=[],Ue.deferredCalls=[]},registerRemoveEventListeners:function(){Ue.removeEventListenersRegistered||(Lg.push(Ue.removeAllEventListeners),Ue.removeEventListenersRegistered=!0)},deferredCalls:[],deferCall:function(N,D,U){function Q(ve,_e){if(ve.length!=_e.length)return!1;for(var Ct in ve)if(ve[Ct]!=_e[Ct])return!1;return!0}for(var pe in Ue.deferredCalls){var he=Ue.deferredCalls[pe];if(he.targetFunction==N&&Q(he.argsList,U))return}Ue.deferredCalls.push({targetFunction:N,precedence:D,argsList:U}),Ue.deferredCalls.sort(function(ve,_e){return ve.precedence<_e.precedence})},removeDeferredCalls:function(N){for(var D=0;D<Ue.deferredCalls.length;++D)Ue.deferredCalls[D].targetFunction==N&&(Ue.deferredCalls.splice(D,1),--D)},canPerformEventHandlerRequests:function(){return Ue.inEventHandler&&Ue.currentEventHandler.allowsDeferredCalls},runDeferredCalls:function(){if(Ue.canPerformEventHandlerRequests())for(var N=0;N<Ue.deferredCalls.length;++N){var D=Ue.deferredCalls[N];Ue.deferredCalls.splice(N,1),--N,D.targetFunction.apply(null,D.argsList)}},eventHandlers:[],removeAllHandlersOnTarget:function(N,D){for(var U=0;U<Ue.eventHandlers.length;++U)Ue.eventHandlers[U].target==N&&(!D||D==Ue.eventHandlers[U].eventTypeString)&&Ue._removeHandler(U--)},_removeHandler:function(N){var D=Ue.eventHandlers[N];D.target.removeEventListener(D.eventTypeString,D.eventListenerFunc,D.useCapture),Ue.eventHandlers.splice(N,1)},registerOrRemoveHandler:function(N){var D=function(Q){++Ue.inEventHandler,Ue.currentEventHandler=N,Ue.runDeferredCalls(),N.handlerFunc(Q),Ue.runDeferredCalls(),--Ue.inEventHandler};if(N.callbackfunc)N.eventListenerFunc=D,N.target.addEventListener(N.eventTypeString,D,N.useCapture),Ue.eventHandlers.push(N),Ue.registerRemoveEventListeners();else for(var U=0;U<Ue.eventHandlers.length;++U)Ue.eventHandlers[U].target==N.target&&Ue.eventHandlers[U].eventTypeString==N.eventTypeString&&Ue._removeHandler(U--)},queueEventHandlerOnThread_iiii:function(N,D,U,Q,pe){el(function(){var he=ol(12);u()[he>>2]=U,u()[he+4>>2]=Q,u()[he+8>>2]=pe,cb(N,637534208,D,Q,he)})},getTargetThreadForEventCallback:function(N){switch(N){case 1:return 0;case 2:return Ce.currentProxiedOperationCallerThread;default:return N}},getNodeNameForTarget:function(N){return N?N==window?"#window":N==screen?"#screen":N&&N.nodeName?N.nodeName:"":""},fullscreenEnabled:function(){return document.fullscreenEnabled||document.webkitFullscreenEnabled}};function hy(N){var D=Pd(N)+1,U=pb(D);return Ms(N,U,D),U}function my(N,D,U,Q){el(function(){var pe=ol(12),he=0;D&&(he=hy(D)),u()[pe>>2]=he,u()[pe+4>>2]=U,u()[pe+8>>2]=Q,cb(N,657457152,0,he,pe)})}function fy(N,D,U,Q){D=D?oa(D):"",my(N,D,U,Q)}function gy(N){return N>2?oa(N):N}var yy=[0,typeof document!="undefined"?document:0,typeof window!="undefined"?window:0];function by(N){N=gy(N);var D=yy[N]||(typeof document!="undefined"?document.querySelector(N):void 0);return D}function Ep(N){return by(N)}function th(N,D,U){var Q=Ep(N);if(!Q)return-4;if(Q.canvasSharedPtr&&(u()[Q.canvasSharedPtr>>2]=D,u()[Q.canvasSharedPtr+4>>2]=U),Q.offscreenCanvas||!Q.controlTransferredOffscreen){Q.offscreenCanvas&&(Q=Q.offscreenCanvas);var pe=!1;if(Q.GLctxObject&&Q.GLctxObject.GLctx){var he=Q.GLctxObject.GLctx.getParameter(2978);pe=he[0]===0&&he[1]===0&&he[2]===Q.width&&he[3]===Q.height}Q.width=D,Q.height=U,pe&&Q.GLctxObject.GLctx.viewport(0,0,D,U)}else if(Q.canvasSharedPtr){var ve=u()[Q.canvasSharedPtr+8>>2];return fy(ve,N,D,U),1}else return-4;return 0}function nh(N,D,U){return _?Bs(2,1,N,D,U):th(N,D,U)}function xy(N,D,U){var Q=Ep(N);return Q?th(N,D,U):nh(N,D,U)}function vy(){throw"unwind"}function wy(N){var D=N.getExtension("ANGLE_instanced_arrays");if(D)return N.vertexAttribDivisor=function(U,Q){D.vertexAttribDivisorANGLE(U,Q)},N.drawArraysInstanced=function(U,Q,pe,he){D.drawArraysInstancedANGLE(U,Q,pe,he)},N.drawElementsInstanced=function(U,Q,pe,he,ve){D.drawElementsInstancedANGLE(U,Q,pe,he,ve)},1}function ky(N){var D=N.getExtension("OES_vertex_array_object");if(D)return N.createVertexArray=function(){return D.createVertexArrayOES()},N.deleteVertexArray=function(U){D.deleteVertexArrayOES(U)},N.bindVertexArray=function(U){D.bindVertexArrayOES(U)},N.isVertexArray=function(U){return D.isVertexArrayOES(U)},1}function Iy(N){var D=N.getExtension("WEBGL_draw_buffers");if(D)return N.drawBuffers=function(U,Q){D.drawBuffersWEBGL(U,Q)},1}function Sy(N){return!!(N.multiDrawWebgl=N.getExtension("WEBGL_multi_draw"))}var _t={counter:1,buffers:[],programs:[],framebuffers:[],renderbuffers:[],textures:[],shaders:[],vaos:[],contexts:{},offscreenCanvases:{},queries:[],stringCache:{},unpackAlignment:4,recordError:function(N){_t.lastError||(_t.lastError=N)},getNewId:function(N){for(var D=_t.counter++,U=N.length;U<D;U++)N[U]=null;return D},getSource:function(N,D,U,Q){for(var pe="",he=0;he<D;++he){var ve=Q?u()[Q+he*4>>2]:-1;pe+=oa(u()[U+he*4>>2],ve<0?void 0:ve)}return pe},createContext:function(N,D){N.getContextSafariWebGL2Fixed||(N.getContextSafariWebGL2Fixed=N.getContext,N.getContext=function(pe,he){var ve=N.getContextSafariWebGL2Fixed(pe,he);return pe=="webgl"==ve instanceof WebGLRenderingContext?ve:null});var U=N.getContext("webgl",D);if(!U)return 0;var Q=_t.registerContext(U,D);return Q},registerContext:function(N,D){var U=pb(8);u()[U+4>>2]=ch();var Q={handle:U,attributes:D,version:D.majorVersion,GLctx:N};return N.canvas&&(N.canvas.GLctxObject=Q),_t.contexts[U]=Q,(typeof D.enableExtensionsByDefault=="undefined"||D.enableExtensionsByDefault)&&_t.initExtensions(Q),U},makeContextCurrent:function(N){return _t.currentContext=_t.contexts[N],c.ctx=ih=_t.currentContext&&_t.currentContext.GLctx,!(N&&!ih)},getContext:function(N){return _t.contexts[N]},deleteContext:function(N){_t.currentContext===_t.contexts[N]&&(_t.currentContext=null),typeof Ue=="object"&&Ue.removeAllHandlersOnTarget(_t.contexts[N].GLctx.canvas),_t.contexts[N]&&_t.contexts[N].GLctx.canvas&&(_t.contexts[N].GLctx.canvas.GLctxObject=void 0),O1(_t.contexts[N].handle),_t.contexts[N]=null},initExtensions:function(N){if(N||(N=_t.currentContext),!N.initExtensionsDone){N.initExtensionsDone=!0;var D=N.GLctx;wy(D),ky(D),Iy(D),D.disjointTimerQueryExt=D.getExtension("EXT_disjoint_timer_query"),Sy(D);var U=D.getSupportedExtensions()||[];U.forEach(function(Q){!Q.includes("lose_context")&&!Q.includes("debug")&&D.getExtension(Q)})}}},Ny=["default","low-power","high-performance"];function Ty(N,D){var U=D>>2,Q=u()[U+6],pe={alpha:!!u()[U+0],depth:!!u()[U+1],stencil:!!u()[U+2],antialias:!!u()[U+3],premultipliedAlpha:!!u()[U+4],preserveDrawingBuffer:!!u()[U+5],powerPreference:Ny[Q],failIfMajorPerformanceCaveat:!!u()[U+7],majorVersion:u()[U+8],minorVersion:u()[U+9],enableExtensionsByDefault:u()[U+10],explicitSwapControl:u()[U+11],proxyContextToMainThread:u()[U+12],renderViaOffscreenBackBuffer:u()[U+13]},he=Ep(N);if(!he||pe.explicitSwapControl)return 0;var ve=_t.createContext(he,pe);return ve}function _y(N,D){return Ty(N,D)}var nl={mappings:{},buffers:[null,[],[]],printChar:function(N,D){var U=nl.buffers[N];D===0||D===10?((N===1?re:Y)(ia(U,0)),U.length=0):U.push(D)},varargs:void 0,get:function(){nl.varargs+=4;var N=u()[nl.varargs-4>>2];return N},getStr:function(N){var D=oa(N);return D},get64:function(N,D){return N}};function ah(N){return _?Bs(3,1,N):0}function rh(N,D,U,Q,pe){if(_)return Bs(4,1,N,D,U,Q,pe)}function sh(N,D,U,Q){if(_)return Bs(5,1,N,D,U,Q);for(var pe=0,he=0;he<U;he++){var ve=u()[D>>2],_e=u()[D+4>>2];D+=8;for(var Ct=0;Ct<_e;Ct++)nl.printChar(N,i()[ve+Ct]);pe+=_e}return u()[Q>>2]=pe,0}function Cy(N){$e(N)}Ce.init();var ih,Ey=[null,Jd,nh,ah,rh,sh],R1=!1,oh={__clock_gettime:ey,__emscripten_init_main_thread_js:ty,__emscripten_thread_cleanup:ny,__pthread_create_js:ay,_emscripten_default_pthread_stack_size:ry,_emscripten_notify_thread_queue:sy,abort:iy,emscripten_check_blocking_allowed:oy,emscripten_get_heap_max:eh,emscripten_get_now:zs,emscripten_memcpy_big:ly,emscripten_num_logical_cores:uy,emscripten_receive_on_main_thread_js:py,emscripten_resize_heap:dy,emscripten_set_canvas_element_size:xy,emscripten_unwind_to_js_event_loop:vy,emscripten_webgl_create_context:_y,exit:Qd,fd_close:ah,fd_seek:rh,fd_write:sh,memory:Te||c.wasmMemory,setTempRet0:Cy},M1=Ug(),Fy=c.___wasm_call_ctors=function(){return(Fy=c.___wasm_call_ctors=c.asm.__wasm_call_ctors).apply(null,arguments)},Ay=c._init=function(){return(Ay=c._init=c.asm.init).apply(null,arguments)},$y=c._init_with_threads_count=function(){return($y=c._init_with_threads_count=c.asm.init_with_threads_count).apply(null,arguments)},Dy=c._get_threads_count=function(){return(Dy=c._get_threads_count=c.asm.get_threads_count).apply(null,arguments)},Ry=c._register_tensor=function(){return(Ry=c._register_tensor=c.asm.register_tensor).apply(null,arguments)},My=c._dispose_data=function(){return(My=c._dispose_data=c.asm.dispose_data).apply(null,arguments)},Py=c._dispose=function(){return(Py=c._dispose=c.asm.dispose).apply(null,arguments)},Oy=c._Abs=function(){return(Oy=c._Abs=c.asm.Abs).apply(null,arguments)},Ly=c._Add=function(){return(Ly=c._Add=c.asm.Add).apply(null,arguments)},zy=c._AddN=function(){return(zy=c._AddN=c.asm.AddN).apply(null,arguments)},By=c._All=function(){return(By=c._All=c.asm.All).apply(null,arguments)},Wy=c._Any=function(){return(Wy=c._Any=c.asm.Any).apply(null,arguments)},Vy=c._ArgMax=function(){return(Vy=c._ArgMax=c.asm.ArgMax).apply(null,arguments)},Uy=c._AvgPool=function(){return(Uy=c._AvgPool=c.asm.AvgPool).apply(null,arguments)},Gy=c._BatchMatMul=function(){return(Gy=c._BatchMatMul=c.asm.BatchMatMul).apply(null,arguments)},Hy=c._Ceil=function(){return(Hy=c._Ceil=c.asm.Ceil).apply(null,arguments)},jy=c._ClipByValue=function(){return(jy=c._ClipByValue=c.asm.ClipByValue).apply(null,arguments)},qy=c._Conv2D=function(){return(qy=c._Conv2D=c.asm.Conv2D).apply(null,arguments)},Ky=c._Conv2DBackpropInput=function(){return(Ky=c._Conv2DBackpropInput=c.asm.Conv2DBackpropInput).apply(null,arguments)},Xy=c._Cos=function(){return(Xy=c._Cos=c.asm.Cos).apply(null,arguments)},Yy=c._Cosh=function(){return(Yy=c._Cosh=c.asm.Cosh).apply(null,arguments)},Qy=c._CropAndResize=function(){return(Qy=c._CropAndResize=c.asm.CropAndResize).apply(null,arguments)},Jy=c._Cumprod=function(){return(Jy=c._Cumprod=c.asm.Cumprod).apply(null,arguments)},Zy=c._Cumsum=function(){return(Zy=c._Cumsum=c.asm.Cumsum).apply(null,arguments)},eb=c._DepthToSpace=function(){return(eb=c._DepthToSpace=c.asm.DepthToSpace).apply(null,arguments)},tb=c._DepthwiseConv2dNative=function(){return(tb=c._DepthwiseConv2dNative=c.asm.DepthwiseConv2dNative).apply(null,arguments)},nb=c._Elu=function(){return(nb=c._Elu=c.asm.Elu).apply(null,arguments)},ab=c._Equal=function(){return(ab=c._Equal=c.asm.Equal).apply(null,arguments)},rb=c._Exp=function(){return(rb=c._Exp=c.asm.Exp).apply(null,arguments)},sb=c._FlipLeftRight=function(){return(sb=c._FlipLeftRight=c.asm.FlipLeftRight).apply(null,arguments)},lh=c._Floor=function(){return(lh=c._Floor=c.asm.Floor).apply(null,arguments)},uh=c._FloorDiv=function(){return(uh=c._FloorDiv=c.asm.FloorDiv).apply(null,arguments)},Fp=c._FusedBatchNorm=function(){return(Fp=c._FusedBatchNorm=c.asm.FusedBatchNorm).apply(null,arguments)},ib=c._FusedConv2D=function(){return(ib=c._FusedConv2D=c.asm.FusedConv2D).apply(null,arguments)},ob=c._FusedDepthwiseConv2D=function(){return(ob=c._FusedDepthwiseConv2D=c.asm.FusedDepthwiseConv2D).apply(null,arguments)},al=c._Gather=function(){return(al=c._Gather=c.asm.Gather).apply(null,arguments)},Ap=c._GatherNd=function(){return(Ap=c._GatherNd=c.asm.GatherNd).apply(null,arguments)},$p=c._Greater=function(){return($p=c._Greater=c.asm.Greater).apply(null,arguments)},P1=c._GreaterEqual=function(){return(P1=c._GreaterEqual=c.asm.GreaterEqual).apply(null,arguments)},rl=c._LeakyRelu=function(){return(rl=c._LeakyRelu=c.asm.LeakyRelu).apply(null,arguments)},sl=c._Less=function(){return(sl=c._Less=c.asm.Less).apply(null,arguments)},lb=c._LessEqual=function(){return(lb=c._LessEqual=c.asm.LessEqual).apply(null,arguments)},G=c._Log=function(){return(G=c._Log=c.asm.Log).apply(null,arguments)},te=c._LogicalAnd=function(){return(te=c._LogicalAnd=c.asm.LogicalAnd).apply(null,arguments)},de=c._Max=function(){return(de=c._Max=c.asm.Max).apply(null,arguments)},Se=c._MaxPool=function(){return(Se=c._MaxPool=c.asm.MaxPool).apply(null,arguments)},Je=c._Maximum=function(){return(Je=c._Maximum=c.asm.Maximum).apply(null,arguments)},rt=c._Mean=function(){return(rt=c._Mean=c.asm.Mean).apply(null,arguments)},Ge=c._Min=function(){return(Ge=c._Min=c.asm.Min).apply(null,arguments)},We=c._Minimum=function(){return(We=c._Minimum=c.asm.Minimum).apply(null,arguments)},Lt=c._MirrorPad=function(){return(Lt=c._MirrorPad=c.asm.MirrorPad).apply(null,arguments)},Pa=c._Multiply=function(){return(Pa=c._Multiply=c.asm.Multiply).apply(null,arguments)},Oa=c._Neg=function(){return(Oa=c._Neg=c.asm.Neg).apply(null,arguments)},il=c._NonMaxSuppressionV3=function(){return(il=c._NonMaxSuppressionV3=c.asm.NonMaxSuppressionV3).apply(null,arguments)},Ws=c._NonMaxSuppressionV4=function(){return(Ws=c._NonMaxSuppressionV4=c.asm.NonMaxSuppressionV4).apply(null,arguments)},ub=c._NonMaxSuppressionV5=function(){return(ub=c._NonMaxSuppressionV5=c.asm.NonMaxSuppressionV5).apply(null,arguments)},Mn=c._NotEqual=function(){return(Mn=c._NotEqual=c.asm.NotEqual).apply(null,arguments)},Qr=c._OneHot=function(){return(Qr=c._OneHot=c.asm.OneHot).apply(null,arguments)},ph=c._PadV2=function(){return(ph=c._PadV2=c.asm.PadV2).apply(null,arguments)},QF=c._Pow=function(){return(QF=c._Pow=c.asm.Pow).apply(null,arguments)},JF=c._Prelu=function(){return(JF=c._Prelu=c.asm.Prelu).apply(null,arguments)},ZF=c._Prod=function(){return(ZF=c._Prod=c.asm.Prod).apply(null,arguments)},eA=c._RealDiv=function(){return(eA=c._RealDiv=c.asm.RealDiv).apply(null,arguments)},tA=c._Relu=function(){return(tA=c._Relu=c.asm.Relu).apply(null,arguments)},nA=c._Relu6=function(){return(nA=c._Relu6=c.asm.Relu6).apply(null,arguments)},aA=c._ResizeBilinear=function(){return(aA=c._ResizeBilinear=c.asm.ResizeBilinear).apply(null,arguments)},rA=c._Reverse=function(){return(rA=c._Reverse=c.asm.Reverse).apply(null,arguments)},sA=c._RotateWithOffset=function(){return(sA=c._RotateWithOffset=c.asm.RotateWithOffset).apply(null,arguments)},iA=c._Round=function(){return(iA=c._Round=c.asm.Round).apply(null,arguments)},oA=c._Rsqrt=function(){return(oA=c._Rsqrt=c.asm.Rsqrt).apply(null,arguments)},lA=c._ScatterNd=function(){return(lA=c._ScatterNd=c.asm.ScatterNd).apply(null,arguments)},uA=c._SelectV2=function(){return(uA=c._SelectV2=c.asm.SelectV2).apply(null,arguments)},pA=c._Sigmoid=function(){return(pA=c._Sigmoid=c.asm.Sigmoid).apply(null,arguments)},cA=c._Sin=function(){return(cA=c._Sin=c.asm.Sin).apply(null,arguments)},dA=c._Softmax=function(){return(dA=c._Softmax=c.asm.Softmax).apply(null,arguments)},hA=c._SparseFillEmptyRows=function(){return(hA=c._SparseFillEmptyRows=c.asm.SparseFillEmptyRows).apply(null,arguments)},mA=c._SparseReshape=function(){return(mA=c._SparseReshape=c.asm.SparseReshape).apply(null,arguments)},fA=c._SparseSegmentReduction=function(){return(fA=c._SparseSegmentReduction=c.asm.SparseSegmentReduction).apply(null,arguments)},gA=c._Sqrt=function(){return(gA=c._Sqrt=c.asm.Sqrt).apply(null,arguments)},yA=c._Square=function(){return(yA=c._Square=c.asm.Square).apply(null,arguments)},bA=c._SquaredDifference=function(){return(bA=c._SquaredDifference=c.asm.SquaredDifference).apply(null,arguments)},xA=c._Step=function(){return(xA=c._Step=c.asm.Step).apply(null,arguments)},vA=c._StridedSlice=function(){return(vA=c._StridedSlice=c.asm.StridedSlice).apply(null,arguments)},wA=c._Sub=function(){return(wA=c._Sub=c.asm.Sub).apply(null,arguments)},kA=c._Sum=function(){return(kA=c._Sum=c.asm.Sum).apply(null,arguments)},IA=c._Tan=function(){return(IA=c._Tan=c.asm.Tan).apply(null,arguments)},SA=c._Tanh=function(){return(SA=c._Tanh=c.asm.Tanh).apply(null,arguments)},NA=c._Tile=function(){return(NA=c._Tile=c.asm.Tile).apply(null,arguments)},TA=c._TopK=function(){return(TA=c._TopK=c.asm.TopK).apply(null,arguments)},_A=c._Transform=function(){return(_A=c._Transform=c.asm.Transform).apply(null,arguments)},CA=c._Transpose=function(){return(CA=c._Transpose=c.asm.Transpose).apply(null,arguments)},EA=c.__FusedMatMul=function(){return(EA=c.__FusedMatMul=c.asm._FusedMatMul).apply(null,arguments)},pb=c._malloc=function(){return(pb=c._malloc=c.asm.malloc).apply(null,arguments)},O1=c._free=function(){return(O1=c._free=c.asm.free).apply(null,arguments)},FA=c._emscripten_tls_init=function(){return(FA=c._emscripten_tls_init=c.asm.emscripten_tls_init).apply(null,arguments)},L1=c.___errno_location=function(){return(L1=c.___errno_location=c.asm.__errno_location).apply(null,arguments)},ch=c._pthread_self=function(){return(ch=c._pthread_self=c.asm.pthread_self).apply(null,arguments)},z1=c._emscripten_main_thread_process_queued_calls=function(){return(z1=c._emscripten_main_thread_process_queued_calls=c.asm.emscripten_main_thread_process_queued_calls).apply(null,arguments)},AA=c.__emscripten_thread_crashed=function(){return(AA=c.__emscripten_thread_crashed=c.asm._emscripten_thread_crashed).apply(null,arguments)},B1=c.__emscripten_thread_init=function(){return(B1=c.__emscripten_thread_init=c.asm._emscripten_thread_init).apply(null,arguments)},$A=c._emscripten_current_thread_process_queued_calls=function(){return($A=c._emscripten_current_thread_process_queued_calls=c.asm.emscripten_current_thread_process_queued_calls).apply(null,arguments)},DA=c._emscripten_main_browser_thread_id=function(){return(DA=c._emscripten_main_browser_thread_id=c.asm.emscripten_main_browser_thread_id).apply(null,arguments)},RA=c._emscripten_sync_run_in_main_thread_2=function(){return(RA=c._emscripten_sync_run_in_main_thread_2=c.asm.emscripten_sync_run_in_main_thread_2).apply(null,arguments)},W1=c._emscripten_sync_run_in_main_thread_4=function(){return(W1=c._emscripten_sync_run_in_main_thread_4=c.asm.emscripten_sync_run_in_main_thread_4).apply(null,arguments)},V1=c._emscripten_run_in_main_runtime_thread_js=function(){return(V1=c._emscripten_run_in_main_runtime_thread_js=c.asm.emscripten_run_in_main_runtime_thread_js).apply(null,arguments)},cb=c._emscripten_dispatch_to_thread_=function(){return(cb=c._emscripten_dispatch_to_thread_=c.asm.emscripten_dispatch_to_thread_).apply(null,arguments)},db=c.__emscripten_thread_free_data=function(){return(db=c.__emscripten_thread_free_data=c.asm._emscripten_thread_free_data).apply(null,arguments)},MA=c.__emscripten_thread_exit=function(){return(MA=c.__emscripten_thread_exit=c.asm._emscripten_thread_exit).apply(null,arguments)},PA=c._memalign=function(){return(PA=c._memalign=c.asm.memalign).apply(null,arguments)},U1=c._emscripten_stack_set_limits=function(){return(U1=c._emscripten_stack_set_limits=c.asm.emscripten_stack_set_limits).apply(null,arguments)},hb=c.stackSave=function(){return(hb=c.stackSave=c.asm.stackSave).apply(null,arguments)},dh=c.stackRestore=function(){return(dh=c.stackRestore=c.asm.stackRestore).apply(null,arguments)},ol=c.stackAlloc=function(){return(ol=c.stackAlloc=c.asm.stackAlloc).apply(null,arguments)},OA=c.dynCall_iijjiiii=function(){return(OA=c.dynCall_iijjiiii=c.asm.dynCall_iijjiiii).apply(null,arguments)},LA=c.dynCall_jiji=function(){return(LA=c.dynCall_jiji=c.asm.dynCall_jiji).apply(null,arguments)},G1=c.__emscripten_allow_main_runtime_queued_calls=21464;c.cwrap=Jt,c.keepRuntimeAlive=Os,c.PThread=Ce,c.PThread=Ce,c.wasmMemory=Te,c.ExitStatus=Dp;var hh;function Dp(N){this.name="ExitStatus",this.message="Program terminated with exit("+N+")",this.status=N}Ma=function N(){hh||mb(),hh||(Ma=N)};function mb(N){if(N=N||y,Yr>0)return;if(_){h(c),Sp(),postMessage({cmd:"loaded"});return}if(vn(),Yr>0)return;function D(){hh||(hh=!0,c.calledRun=!0,!ct&&(Sp(),h(c),c.onRuntimeInitialized&&c.onRuntimeInitialized(),Wg()))}c.setStatus?(c.setStatus("Running..."),setTimeout(function(){setTimeout(function(){c.setStatus("")},1),D()},1)):D()}c.run=mb;function zA(N,D){if(bn=N,!D&&_)throw Jd(N),"unwind";Os()||Bg(),BA(N)}function BA(N){bn=N,Os()||(Ce.terminateAllThreads(),c.onExit&&c.onExit(N),ct=!0),x(N,new Dp(N))}if(c.preInit)for(typeof c.preInit=="function"&&(c.preInit=[c.preInit]);c.preInit.length>0;)c.preInit.pop()();mb();var mh;f&&(mh={uncaughtException:process.listeners("uncaughtException").filter(function(N){return!f.uncaughtException.indexOf(N)>-1}),unhandledRejection:process.listeners("unhandledRejection").filter(function(N){return!f.unhandledRejection.indexOf(N)>-1})});var fh;if(typeof WasmBackendModule!="undefined")fh=WasmBackendModule;else if(typeof r!="undefined")fh=r;else throw new Error("Could not find wasm module in post.js");if(mh){var WA=fh._dispose;fh._dispose=function(){WA(),mh.uncaughtException.forEach(function(N){process.removeListener("uncaughtException",N)}),mh.unhandledRejection.forEach(function(N){process.removeListener("unhandledRejection",N)})}}return r.ready}})();typeof e=="object"&&typeof t=="object"?t.exports=n:typeof define=="function"&&define.amd?define([],function(){return n}):typeof e=="object"&&(e.WasmBackendModuleThreadedSimd=n)}),k$=ft((e,t)=>{var n=(()=>{var a=typeof document!="undefined"&&document.currentScript?document.currentScript.src:void 0;return typeof __filename!="undefined"&&(a=a||__filename),function(r){r=r||{};var s=typeof r!="undefined"?r:{},i,o;s.ready=new Promise(function(G,te){i=G,o=te});var l;typeof process!="undefined"&&process.listeners&&(l={uncaughtException:process.listeners("uncaughtException"),unhandledRejection:process.listeners("unhandledRejection")});var u=Object.assign({},s),p=[],d="./this.program",c=(G,te)=>{throw te},h=typeof window=="object",m=typeof importScripts=="function",f=typeof process=="object"&&typeof process.versions=="object"&&typeof process.versions.node=="string",g="";function y(G){return s.locateFile?s.locateFile(G,g):g+G}var b,x,v,k;function T(G){G instanceof Ap||$("exiting due to exception: "+G)}var _,E,A;f?(m?g=Wh().dirname(g)+"/":g=__dirname+"/",A=()=>{E||(_=Fx(),E=Wh())},b=function(G,te){return A(),G=E.normalize(G),_.readFileSync(G,te?void 0:"utf8")},v=G=>{var te=b(G,!0);return te.buffer||(te=new Uint8Array(te)),te},x=(G,te,de)=>{A(),G=E.normalize(G),_.readFile(G,function(Se,Je){Se?de(Se):te(Je.buffer)})},process.argv.length>1&&(d=process.argv[1].replace(/\\/g,"/")),p=process.argv.slice(2),process.on("uncaughtException",function(G){if(!(G instanceof Ap))throw G}),process.on("unhandledRejection",function(G){throw G}),c=(G,te)=>{if(Ip())throw process.exitCode=G,te;T(te),process.exit(G)},s.inspect=function(){return"[Emscripten Module object]"}):(h||m)&&(m?g=self.location.href:typeof document!="undefined"&&document.currentScript&&(g=document.currentScript.src),a&&(g=a),g.indexOf("blob:")!==0?g=g.substr(0,g.replace(/[?#].*/,"").lastIndexOf("/")+1):g="",b=G=>{var te=new XMLHttpRequest;return te.open("GET",G,!1),te.send(null),te.responseText},m&&(v=G=>{var te=new XMLHttpRequest;return te.open("GET",G,!1),te.responseType="arraybuffer",te.send(null),new Uint8Array(te.response)}),x=(G,te,de)=>{var Se=new XMLHttpRequest;Se.open("GET",G,!0),Se.responseType="arraybuffer",Se.onload=()=>{if(Se.status==200||Se.status==0&&Se.response){te(Se.response);return}de()},Se.onerror=de,Se.send(null)},k=G=>document.title=G);var M=s.print||console.log.bind(console),$=s.printErr||console.warn.bind(console);Object.assign(s,u),u=null,s.arguments&&(p=s.arguments),s.thisProgram&&(d=s.thisProgram),s.quit&&(c=s.quit);var S=4;function P(G){P.shown||(P.shown={}),P.shown[G]||(P.shown[G]=1,$(G))}function V(G,te){if(typeof WebAssembly.Function=="function"){for(var de={i:"i32",j:"i64",f:"f32",d:"f64"},Se={parameters:[],results:te[0]=="v"?[]:[de[te[0]]]},Je=1;Je<te.length;++Je)Se.parameters.push(de[te[Je]]);return new WebAssembly.Function(Se,G)}var rt=[1,0,1,96],Ge=te.slice(0,1),We=te.slice(1),Lt={i:127,j:126,f:125,d:124};rt.push(We.length);for(var Je=0;Je<We.length;++Je)rt.push(Lt[We[Je]]);Ge=="v"?rt.push(0):rt=rt.concat([1,Lt[Ge]]),rt[1]=rt.length-2;var Pa=new Uint8Array([0,97,115,109,1,0,0,0].concat(rt,[2,7,1,1,101,1,102,0,0,7,5,1,1,102,0,0])),Oa=new WebAssembly.Module(Pa),il=new WebAssembly.Instance(Oa,{e:{f:G}}),Ws=il.exports.f;return Ws}var j=[],q;function K(){if(j.length)return j.pop();try{Kr.grow(1)}catch(G){throw G instanceof RangeError?"Unable to grow wasm table. Set ALLOW_TABLE_GROWTH.":G}return Kr.length-1}function Z(G,te){for(var de=G;de<G+te;de++){var Se=Tp(de);Se&&q.set(Se,de)}}var ee=0,re=G=>{ee=G},Y;s.wasmBinary&&(Y=s.wasmBinary);var ie=s.noExitRuntime||!0;typeof WebAssembly!="object"&&Ps("no native wasm support detected");var ae,le=!1,ue;function ke(G,te){G||Ps(te)}function ye(G){var te=s["_"+G];return te}function Ie(G,te,de,Se,Je){var rt={string:function(Mn){var Qr=0;if(Mn!=null&&Mn!==0){var ph=(Mn.length<<2)+1;Qr=Fp(ph),at(Mn,Qr,ph)}return Qr},array:function(Mn){var Qr=Fp(Mn.length);return ct(Mn,Qr),Qr}};function Ge(Mn){return te==="string"?st(Mn):te==="boolean"?Boolean(Mn):Mn}var We=ye(G),Lt=[],Pa=0;if(Se)for(var Oa=0;Oa<Se.length;Oa++){var il=rt[de[Oa]];il?(Pa===0&&(Pa=lh()),Lt[Oa]=il(Se[Oa])):Lt[Oa]=Se[Oa]}var Ws=We.apply(null,Lt);function ub(Mn){return Pa!==0&&uh(Pa),Ge(Mn)}return Ws=ub(Ws),Ws}function Ee(G,te,de,Se){de=de||[];var Je=de.every(function(Ge){return Ge==="number"}),rt=te!=="string";return rt&&Je&&!Se?ye(G):function(){return Ie(G,te,de,arguments,Se)}}var $e=1,Be=typeof TextDecoder!="undefined"?new TextDecoder("utf8"):void 0;function je(G,te,de){for(var Se=te+de,Je=te;G[Je]&&!(Je>=Se);)++Je;if(Je-te>16&&G.subarray&&Be)return Be.decode(G.subarray(te,Je));for(var rt="";te<Je;){var Ge=G[te++];if(!(Ge&128)){rt+=String.fromCharCode(Ge);continue}var We=G[te++]&63;if((Ge&224)==192){rt+=String.fromCharCode((Ge&31)<<6|We);continue}var Lt=G[te++]&63;if((Ge&240)==224?Ge=(Ge&15)<<12|We<<6|Lt:Ge=(Ge&7)<<18|We<<12|Lt<<6|G[te++]&63,Ge<65536)rt+=String.fromCharCode(Ge);else{var Pa=Ge-65536;rt+=String.fromCharCode(55296|Pa>>10,56320|Pa&1023)}}return rt}function st(G,te){return G?je(Jt,G,te):""}function nt(G,te,de,Se){if(!(Se>0))return 0;for(var Je=de,rt=de+Se-1,Ge=0;Ge<G.length;++Ge){var We=G.charCodeAt(Ge);if(We>=55296&&We<=57343){var Lt=G.charCodeAt(++Ge);We=65536+((We&1023)<<10)|Lt&1023}if(We<=127){if(de>=rt)break;te[de++]=We}else if(We<=2047){if(de+1>=rt)break;te[de++]=192|We>>6,te[de++]=128|We&63}else if(We<=65535){if(de+2>=rt)break;te[de++]=224|We>>12,te[de++]=128|We>>6&63,te[de++]=128|We&63}else{if(de+3>=rt)break;te[de++]=240|We>>18,te[de++]=128|We>>12&63,te[de++]=128|We>>6&63,te[de++]=128|We&63}}return te[de]=0,de-Je}function at(G,te,de){return nt(G,Jt,te,de)}function Te(G){for(var te=0,de=0;de<G.length;++de){var Se=G.charCodeAt(de);Se>=55296&&Se<=57343&&(Se=65536+((Se&1023)<<10)|G.charCodeAt(++de)&1023),Se<=127?++te:Se<=2047?te+=2:Se<=65535?te+=3:te+=4}return te}var gt=typeof TextDecoder!="undefined"?new TextDecoder("utf-16le"):void 0;function ct(G,te){Gt.set(G,te)}function bn(G,te,de){for(var Se=0;Se<G.length;++Se)Gt[te++>>0]=G.charCodeAt(Se);de||(Gt[te>>0]=0)}function Qt(G,te){return G%te>0&&(G+=te-G%te),G}var Dn,Gt,Jt,Da,Rn,Ht,ia,oa,qr;function Ms(G){Dn=G,s.HEAP8=Gt=new Int8Array(G),s.HEAP16=Da=new Int16Array(G),s.HEAP32=Ht=new Int32Array(G),s.HEAPU8=Jt=new Uint8Array(G),s.HEAPU16=Rn=new Uint16Array(G),s.HEAPU32=ia=new Uint32Array(G),s.HEAPF32=oa=new Float32Array(G),s.HEAPF64=qr=new Float64Array(G)}var Pd=s.INITIAL_MEMORY||16777216,Kr,kr=[],kp=[],Yo=[],xn=!1,Od=!1,Ld=0;function Ip(){return ie||Ld>0}function zd(){if(s.preRun)for(typeof s.preRun=="function"&&(s.preRun=[s.preRun]);s.preRun.length;)Vd(s.preRun.shift());Np(kr)}function Bd(){xn=!0,Np(kp)}function N1(){Od=!0}function Wd(){if(s.postRun)for(typeof s.postRun=="function"&&(s.postRun=[s.postRun]);s.postRun.length;)Ud(s.postRun.shift());Np(Yo)}function Vd(G){kr.unshift(G)}function Ra(G){kp.unshift(G)}function Ud(G){Yo.unshift(G)}var la=0,Qo=null,Xr=null;function Lg(G){la++,s.monitorRunDependencies&&s.monitorRunDependencies(la)}function Gd(G){if(la--,s.monitorRunDependencies&&s.monitorRunDependencies(la),la==0&&(Qo!==null&&(clearInterval(Qo),Qo=null),Xr)){var te=Xr;Xr=null,te()}}s.preloadedImages={},s.preloadedAudios={};function Ps(G){s.onAbort&&s.onAbort(G),G="Aborted("+G+")",$(G),le=!0,ue=1,G+=". Build with -s ASSERTIONS=1 for more info.";var te=new WebAssembly.RuntimeError(G);throw o(te),te}var zg="data:application/octet-stream;base64,";function Hd(G){return G.startsWith(zg)}function Os(G){return G.startsWith("file://")}var vn;vn="tfjs-backend-wasm.wasm",Hd(vn)||(vn=y(vn));function Sp(G){try{if(G==vn&&Y)return new Uint8Array(Y);if(v)return v(G);throw"both async and sync fetching of the wasm failed"}catch(te){Ps(te)}}function Bg(){if(!Y&&(h||m)){if(typeof fetch=="function"&&!Os(vn))return fetch(vn,{credentials:"same-origin"}).then(function(G){if(!G.ok)throw"failed to load wasm binary file at '"+vn+"'";return G.arrayBuffer()}).catch(function(){return Sp(vn)});if(x)return new Promise(function(G,te){x(vn,function(de){G(new Uint8Array(de))},te)})}return Promise.resolve().then(function(){return Sp(vn)})}function Wg(){var G={env:el,wasi_snapshot_preview1:el};function te(Ge,We){var Lt=Ge.exports;s.asm=Lt,ae=s.asm.memory,Ms(ae.buffer),Kr=s.asm.__indirect_function_table,Ra(s.asm.__wasm_call_ctors),Gd("wasm-instantiate")}Lg("wasm-instantiate");function de(Ge){te(Ge.instance)}function Se(Ge){return Bg().then(function(We){return WebAssembly.instantiate(We,G)}).then(function(We){return We}).then(Ge,function(We){$("failed to asynchronously prepare wasm: "+We),Ps(We)})}function Je(){return!Y&&typeof WebAssembly.instantiateStreaming=="function"&&!Hd(vn)&&!Os(vn)&&typeof fetch=="function"?fetch(vn,{credentials:"same-origin"}).then(function(Ge){var We=WebAssembly.instantiateStreaming(Ge,G);return We.then(de,function(Lt){return $("wasm streaming compile failed: "+Lt),$("falling back to ArrayBuffer instantiation"),Se(de)})}):Se(de)}if(s.instantiateWasm)try{var rt=s.instantiateWasm(G,te);return rt}catch(Ge){return $("Module.instantiateWasm callback failed with error: "+Ge),!1}return Je().catch(o),{}}var T1,_1;function Np(G){for(;G.length>0;){var te=G.shift();if(typeof te=="function"){te(s);continue}var de=te.func;typeof de=="number"?te.arg===void 0?Tp(de)():Tp(de)(te.arg):de(te.arg===void 0?null:te.arg)}}function Yr(G){return G}function jd(G){var te=/\b_Z[\w\d_]+/g;return G.replace(te,function(de){var Se=de;return de===Se?de:Se+" ["+de+"]"})}var Ma=[];function Tp(G){var te=Ma[G];return te||(G>=Ma.length&&(Ma.length=G+1),Ma[G]=te=Kr.get(G)),te}function C1(){var G=new Error;if(!G.stack){try{throw new Error}catch(te){G=te}if(!G.stack)return"(no stack trace available)"}return G.stack.toString()}function Jo(G,te){Kr.set(G,te),Ma[G]=te}function Vg(){Ps("")}function _p(){return 2147483648}function qd(G,te,de){Jt.copyWithin(G,te,te+de)}function wn(G){try{return ae.grow(G-Dn.byteLength+65535>>>16),Ms(ae.buffer),1}catch(te){}}function Kd(G){var te=Jt.length;G=G>>>0;var de=_p();if(G>de)return!1;for(var Se=1;Se<=4;Se*=2){var Je=te*(1+.2/Se);Je=Math.min(Je,G+100663296);var rt=Math.min(de,Qt(Math.max(G,Je),65536)),Ge=wn(rt);if(Ge)return!0}return!1}var Zo={mappings:{},buffers:[null,[],[]],printChar:function(G,te){var de=Zo.buffers[G];te===0||te===10?((G===1?M:$)(je(de,0)),de.length=0):de.push(te)},varargs:void 0,get:function(){Zo.varargs+=4;var G=Ht[Zo.varargs-4>>2];return G},getStr:function(G){var te=st(G);return te},get64:function(G,te){return G}};function Ug(G){return 0}function E1(G,te,de,Se,Je){}function F1(G,te,de,Se){for(var Je=0,rt=0;rt<de;rt++){var Ge=Ht[te>>2],We=Ht[te+4>>2];te+=8;for(var Lt=0;Lt<We;Lt++)Zo.printChar(G,Jt[Ge+Lt]);Je+=We}return Ht[Se>>2]=Je,0}function Gg(G){re(G)}var Xd=!1,el={abort:Vg,emscripten_get_heap_max:_p,emscripten_memcpy_big:qd,emscripten_resize_heap:Kd,fd_close:Ug,fd_seek:E1,fd_write:F1,setTempRet0:Gg},YF=Wg(),A1=s.___wasm_call_ctors=function(){return(A1=s.___wasm_call_ctors=s.asm.__wasm_call_ctors).apply(null,arguments)},Hg=s._init=function(){return(Hg=s._init=s.asm.init).apply(null,arguments)},jg=s._init_with_threads_count=function(){return(jg=s._init_with_threads_count=s.asm.init_with_threads_count).apply(null,arguments)},Yd=s._get_threads_count=function(){return(Yd=s._get_threads_count=s.asm.get_threads_count).apply(null,arguments)},Qd=s._register_tensor=function(){return(Qd=s._register_tensor=s.asm.register_tensor).apply(null,arguments)},qg=s._dispose_data=function(){return(qg=s._dispose_data=s.asm.dispose_data).apply(null,arguments)},Ce=s._dispose=function(){return(Ce=s._dispose=s.asm.dispose).apply(null,arguments)},Kg=s._Abs=function(){return(Kg=s._Abs=s.asm.Abs).apply(null,arguments)},Jd=s._Add=function(){return(Jd=s._Add=s.asm.Add).apply(null,arguments)},Ls=s._AddN=function(){return(Ls=s._AddN=s.asm.AddN).apply(null,arguments)},tl=s._All=function(){return(tl=s._All=s.asm.All).apply(null,arguments)},Xg=s._Any=function(){return(Xg=s._Any=s.asm.Any).apply(null,arguments)},$1=s._ArgMax=function(){return($1=s._ArgMax=s.asm.ArgMax).apply(null,arguments)},Yg=s._AvgPool=function(){return(Yg=s._AvgPool=s.asm.AvgPool).apply(null,arguments)},D1=s._BatchMatMul=function(){return(D1=s._BatchMatMul=s.asm.BatchMatMul).apply(null,arguments)},zs=s._Ceil=function(){return(zs=s._Ceil=s.asm.Ceil).apply(null,arguments)},Qg=s._ClipByValue=function(){return(Qg=s._ClipByValue=s.asm.ClipByValue).apply(null,arguments)},Jg=s._Conv2D=function(){return(Jg=s._Conv2D=s.asm.Conv2D).apply(null,arguments)},Zg=s._Conv2DBackpropInput=function(){return(Zg=s._Conv2DBackpropInput=s.asm.Conv2DBackpropInput).apply(null,arguments)},ey=s._Cos=function(){return(ey=s._Cos=s.asm.Cos).apply(null,arguments)},ty=s._Cosh=function(){return(ty=s._Cosh=s.asm.Cosh).apply(null,arguments)},ny=s._CropAndResize=function(){return(ny=s._CropAndResize=s.asm.CropAndResize).apply(null,arguments)},Zd=s._Cumprod=function(){return(Zd=s._Cumprod=s.asm.Cumprod).apply(null,arguments)},ay=s._Cumsum=function(){return(ay=s._Cumsum=s.asm.Cumsum).apply(null,arguments)},ry=s._DepthToSpace=function(){return(ry=s._DepthToSpace=s.asm.DepthToSpace).apply(null,arguments)},sy=s._DepthwiseConv2dNative=function(){return(sy=s._DepthwiseConv2dNative=s.asm.DepthwiseConv2dNative).apply(null,arguments)},iy=s._Elu=function(){return(iy=s._Elu=s.asm.Elu).apply(null,arguments)},oy=s._Equal=function(){return(oy=s._Equal=s.asm.Equal).apply(null,arguments)},eh=s._Exp=function(){return(eh=s._Exp=s.asm.Exp).apply(null,arguments)},ly=s._FlipLeftRight=function(){return(ly=s._FlipLeftRight=s.asm.FlipLeftRight).apply(null,arguments)},uy=s._Floor=function(){return(uy=s._Floor=s.asm.Floor).apply(null,arguments)},Bs=s._FloorDiv=function(){return(Bs=s._FloorDiv=s.asm.FloorDiv).apply(null,arguments)},Cp=s._FusedBatchNorm=function(){return(Cp=s._FusedBatchNorm=s.asm.FusedBatchNorm).apply(null,arguments)},py=s._FusedConv2D=function(){return(py=s._FusedConv2D=s.asm.FusedConv2D).apply(null,arguments)},cy=s._FusedDepthwiseConv2D=function(){return(cy=s._FusedDepthwiseConv2D=s.asm.FusedDepthwiseConv2D).apply(null,arguments)},dy=s._Gather=function(){return(dy=s._Gather=s.asm.Gather).apply(null,arguments)},Ue=s._GatherNd=function(){return(Ue=s._GatherNd=s.asm.GatherNd).apply(null,arguments)},hy=s._Greater=function(){return(hy=s._Greater=s.asm.Greater).apply(null,arguments)},my=s._GreaterEqual=function(){return(my=s._GreaterEqual=s.asm.GreaterEqual).apply(null,arguments)},fy=s._LeakyRelu=function(){return(fy=s._LeakyRelu=s.asm.LeakyRelu).apply(null,arguments)},gy=s._Less=function(){return(gy=s._Less=s.asm.Less).apply(null,arguments)},yy=s._LessEqual=function(){return(yy=s._LessEqual=s.asm.LessEqual).apply(null,arguments)},by=s._Log=function(){return(by=s._Log=s.asm.Log).apply(null,arguments)},Ep=s._LogicalAnd=function(){return(Ep=s._LogicalAnd=s.asm.LogicalAnd).apply(null,arguments)},th=s._Max=function(){return(th=s._Max=s.asm.Max).apply(null,arguments)},nh=s._MaxPool=function(){return(nh=s._MaxPool=s.asm.MaxPool).apply(null,arguments)},xy=s._Maximum=function(){return(xy=s._Maximum=s.asm.Maximum).apply(null,arguments)},vy=s._Mean=function(){return(vy=s._Mean=s.asm.Mean).apply(null,arguments)},wy=s._Min=function(){return(wy=s._Min=s.asm.Min).apply(null,arguments)},ky=s._Minimum=function(){return(ky=s._Minimum=s.asm.Minimum).apply(null,arguments)},Iy=s._MirrorPad=function(){return(Iy=s._MirrorPad=s.asm.MirrorPad).apply(null,arguments)},Sy=s._Multiply=function(){return(Sy=s._Multiply=s.asm.Multiply).apply(null,arguments)},_t=s._Neg=function(){return(_t=s._Neg=s.asm.Neg).apply(null,arguments)},Ny=s._NonMaxSuppressionV3=function(){return(Ny=s._NonMaxSuppressionV3=s.asm.NonMaxSuppressionV3).apply(null,arguments)},Ty=s._NonMaxSuppressionV4=function(){return(Ty=s._NonMaxSuppressionV4=s.asm.NonMaxSuppressionV4).apply(null,arguments)},_y=s._NonMaxSuppressionV5=function(){return(_y=s._NonMaxSuppressionV5=s.asm.NonMaxSuppressionV5).apply(null,arguments)},nl=s._NotEqual=function(){return(nl=s._NotEqual=s.asm.NotEqual).apply(null,arguments)},ah=s._OneHot=function(){return(ah=s._OneHot=s.asm.OneHot).apply(null,arguments)},rh=s._PadV2=function(){return(rh=s._PadV2=s.asm.PadV2).apply(null,arguments)},sh=s._Pow=function(){return(sh=s._Pow=s.asm.Pow).apply(null,arguments)},Cy=s._Prelu=function(){return(Cy=s._Prelu=s.asm.Prelu).apply(null,arguments)},ih=s._Prod=function(){return(ih=s._Prod=s.asm.Prod).apply(null,arguments)},Ey=s._RealDiv=function(){return(Ey=s._RealDiv=s.asm.RealDiv).apply(null,arguments)},R1=s._Relu=function(){return(R1=s._Relu=s.asm.Relu).apply(null,arguments)},oh=s._Relu6=function(){return(oh=s._Relu6=s.asm.Relu6).apply(null,arguments)},M1=s._ResizeBilinear=function(){return(M1=s._ResizeBilinear=s.asm.ResizeBilinear).apply(null,arguments)},Fy=s._Reverse=function(){return(Fy=s._Reverse=s.asm.Reverse).apply(null,arguments)},Ay=s._RotateWithOffset=function(){return(Ay=s._RotateWithOffset=s.asm.RotateWithOffset).apply(null,arguments)},$y=s._Round=function(){return($y=s._Round=s.asm.Round).apply(null,arguments)},Dy=s._Rsqrt=function(){return(Dy=s._Rsqrt=s.asm.Rsqrt).apply(null,arguments)},Ry=s._ScatterNd=function(){return(Ry=s._ScatterNd=s.asm.ScatterNd).apply(null,arguments)},My=s._SelectV2=function(){return(My=s._SelectV2=s.asm.SelectV2).apply(null,arguments)},Py=s._Sigmoid=function(){return(Py=s._Sigmoid=s.asm.Sigmoid).apply(null,arguments)},Oy=s._Sin=function(){return(Oy=s._Sin=s.asm.Sin).apply(null,arguments)},Ly=s._Softmax=function(){return(Ly=s._Softmax=s.asm.Softmax).apply(null,arguments)},zy=s._SparseFillEmptyRows=function(){return(zy=s._SparseFillEmptyRows=s.asm.SparseFillEmptyRows).apply(null,arguments)},By=s._SparseReshape=function(){return(By=s._SparseReshape=s.asm.SparseReshape).apply(null,arguments)},Wy=s._SparseSegmentReduction=function(){return(Wy=s._SparseSegmentReduction=s.asm.SparseSegmentReduction).apply(null,arguments)},Vy=s._Sqrt=function(){return(Vy=s._Sqrt=s.asm.Sqrt).apply(null,arguments)},Uy=s._Square=function(){return(Uy=s._Square=s.asm.Square).apply(null,arguments)},Gy=s._SquaredDifference=function(){return(Gy=s._SquaredDifference=s.asm.SquaredDifference).apply(null,arguments)},Hy=s._Step=function(){return(Hy=s._Step=s.asm.Step).apply(null,arguments)},jy=s._StridedSlice=function(){return(jy=s._StridedSlice=s.asm.StridedSlice).apply(null,arguments)},qy=s._Sub=function(){return(qy=s._Sub=s.asm.Sub).apply(null,arguments)},Ky=s._Sum=function(){return(Ky=s._Sum=s.asm.Sum).apply(null,arguments)},Xy=s._Tan=function(){return(Xy=s._Tan=s.asm.Tan).apply(null,arguments)},Yy=s._Tanh=function(){return(Yy=s._Tanh=s.asm.Tanh).apply(null,arguments)},Qy=s._Tile=function(){return(Qy=s._Tile=s.asm.Tile).apply(null,arguments)},Jy=s._TopK=function(){return(Jy=s._TopK=s.asm.TopK).apply(null,arguments)},Zy=s._Transform=function(){return(Zy=s._Transform=s.asm.Transform).apply(null,arguments)},eb=s._Transpose=function(){return(eb=s._Transpose=s.asm.Transpose).apply(null,arguments)},tb=s.__FusedMatMul=function(){return(tb=s.__FusedMatMul=s.asm._FusedMatMul).apply(null,arguments)},nb=s._malloc=function(){return(nb=s._malloc=s.asm.malloc).apply(null,arguments)},ab=s._free=function(){return(ab=s._free=s.asm.free).apply(null,arguments)},rb=s.___errno_location=function(){return(rb=s.___errno_location=s.asm.__errno_location).apply(null,arguments)},sb=s._emscripten_main_thread_process_queued_calls=function(){return(sb=s._emscripten_main_thread_process_queued_calls=s.asm.emscripten_main_thread_process_queued_calls).apply(null,arguments)},lh=s.stackSave=function(){return(lh=s.stackSave=s.asm.stackSave).apply(null,arguments)},uh=s.stackRestore=function(){return(uh=s.stackRestore=s.asm.stackRestore).apply(null,arguments)},Fp=s.stackAlloc=function(){return(Fp=s.stackAlloc=s.asm.stackAlloc).apply(null,arguments)},ib=s.dynCall_iijjiiii=function(){return(ib=s.dynCall_iijjiiii=s.asm.dynCall_iijjiiii).apply(null,arguments)},ob=s.dynCall_jiji=function(){return(ob=s.dynCall_jiji=s.asm.dynCall_jiji).apply(null,arguments)};s.cwrap=Ee;var al;function Ap(G){this.name="ExitStatus",this.message="Program terminated with exit("+G+")",this.status=G}Xr=function G(){al||$p(),al||(Xr=G)};function $p(G){if(G=G||p,la>0||(zd(),la>0))return;function te(){al||(al=!0,s.calledRun=!0,!le&&(Bd(),i(s),s.onRuntimeInitialized&&s.onRuntimeInitialized(),Wd()))}s.setStatus?(s.setStatus("Running..."),setTimeout(function(){setTimeout(function(){s.setStatus("")},1),te()},1)):te()}s.run=$p;function P1(G){ue=G,Ip()||(s.onExit&&s.onExit(G),le=!0),c(G,new Ap(G))}if(s.preInit)for(typeof s.preInit=="function"&&(s.preInit=[s.preInit]);s.preInit.length>0;)s.preInit.pop()();$p();var rl;l&&(rl={uncaughtException:process.listeners("uncaughtException").filter(function(G){return!l.uncaughtException.indexOf(G)>-1}),unhandledRejection:process.listeners("unhandledRejection").filter(function(G){return!l.unhandledRejection.indexOf(G)>-1})});var sl;if(typeof r!="undefined")sl=r;else if(typeof WasmBackendModuleThreadedSimd!="undefined")sl=WasmBackendModuleThreadedSimd;else throw new Error("Could not find wasm module in post.js");if(rl){var lb=sl._dispose;sl._dispose=function(){lb(),rl.uncaughtException.forEach(function(G){process.removeListener("uncaughtException",G)}),rl.unhandledRejection.forEach(function(G){process.removeListener("unhandledRejection",G)})}}return r.ready}})();typeof e=="object"&&typeof t=="object"?t.exports=n:typeof define=="function"&&define.amd?define([],function(){return n}):typeof e=="object"&&(e.WasmBackendModule=n)}),xm=class{constructor(e,t){this.backend=e,this.dataMover=t,this.data=new WeakMap,this.dataIdsCount=0}get(e){return this.data.has(e)||this.dataMover.moveData(this.backend,e),this.data.get(e)}set(e,t){this.dataIdsCount++,this.data.set(e,t)}has(e){return this.data.has(e)}delete(e){return this.dataIdsCount--,this.data.delete(e)}numDataIds(){return this.dataIdsCount}},gc=class{refCount(e){return pa("refCount")}incRef(e){return pa("incRef")}timerAvailable(){return!0}time(e){return pa("time")}read(e){return pa("read")}readSync(e){return pa("readSync")}readToGPU(e,t){return pa("readToGPU")}numDataIds(){return pa("numDataIds")}disposeData(e,t){return pa("disposeData")}write(e,t,n){return pa("write")}move(e,t,n,a,r){return pa("move")}memory(){return pa("memory")}floatPrecision(){return pa("floatPrecision")}epsilon(){return this.floatPrecision()===32?1e-7:1e-4}dispose(){return pa("dispose")}};function pa(e){throw new Error(`'${e}' not yet implemented or not found in the registry. This kernel may not be supported by the tfjs backend you have chosen`)}function wI(e){let t=e.length,n=0;for(;t>0;)n=Math.random()*t|0,t--,Vh(e,t,n)}function I$(e,t){if(e.length!==t.length)throw new Error(`Array sizes must match to be shuffled together First array length was ${e.length}Second array length was ${t.length}`);let n=e.length,a=0;for(;n>0;)a=Math.random()*n|0,n--,Vh(e,n,a),Vh(t,n,a)}function Zp(e,t,n){return Math.max(e,Math.min(t,n))}function S$(e){return e%2===0?e:e+1}function Vh(e,t,n){let a=e[t];e[t]=e[n],e[n]=a}function N$(e){let t=0;for(let n=0;n<e.length;n++)t+=e[n];return t}function T$(e,t){let n=Math.random();return t*n+(1-n)*e}function _$(e,t){let n=0;for(let a=0;a<e.length;a++){let r=Number(e[a])-Number(t[a]);n+=r*r}return n}function R(e,t){if(!e)throw new Error(typeof t=="string"?t:t())}function Tn(e,t,n=""){R(ys(e,t),()=>n+` Shapes ${e} and ${t} must match`)}function vi(e){R(e!=null,()=>"The input to the tensor constructor must be a non-null value.")}function ni(e,t=[],n=!1){if(t==null&&(t=[]),Array.isArray(e)||hn(e)&&!n)for(let a=0;a<e.length;++a)ni(e[a],t,n);else t.push(e);return t}function bt(e){if(e.length===0)return 1;let t=e[0];for(let n=1;n<e.length;n++)t*=e[n];return t}function C$(e){return e.length===0}function ys(e,t){if(e===t)return!0;if(e==null||t==null||e.length!==t.length)return!1;for(let n=0;n<e.length;n++)if(e[n]!==t[n])return!1;return!0}function wl(e){return e%1===0}function E$(e){if(Math.tanh!=null)return Math.tanh(e);if(e===1/0)return 1;if(e===-1/0)return-1;{let t=Math.exp(2*e);return(t-1)/(t+1)}}function F$(e){let t=Math.ceil(Math.sqrt(e));return[t,Math.ceil(e/t)]}function A$(e){let t=new Uint32Array(e);for(let n=0;n<e;++n)t[n]=n;return wI(t),t}function Xp(e,t){return t<=e.length?e:e+" ".repeat(t-e.length)}function $$(e,t=a=>0,n){return new Promise((a,r)=>{let s=0,i=()=>{if(e()){a();return}s++;let o=t(s);if(n!=null&&s>=n){r();return}setTimeout(i,o)};i()})}function D$(e,t){let n=1,a=-1;for(let s=0;s<e.length;++s)if(e[s]>=0)n*=e[s];else if(e[s]===-1){if(a!==-1)throw Error(`Shapes can only have 1 implicit size. Found -1 at dim ${a} and dim ${s}`);a=s}else if(e[s]<0)throw Error(`Shapes can not be < 0. Found ${e[s]} at dim ${s}`);if(a===-1){if(t>0&&t!==n)throw Error(`Size(${t}) must match the product of shape ${e}`);return e}if(n===0)throw Error(`Cannot infer the missing size in [${e}] when there are 0 elements`);if(t%n!==0)throw Error(`The implicit shape can't be a fractional number. Got ${t} / ${n}`);let r=e.slice();return r[a]=t/n,r}function Ea(e,t){let n=t.length;return e=e==null?t.map((a,r)=>r):[].concat(e),R(e.every(a=>a>=-n&&a<n),()=>`All values in axis param must be in range [-${n}, ${n}) but got axis ${e}`),R(e.every(a=>wl(a)),()=>`All values in axis param must be integers but got axis ${e}`),e.map(a=>a<0?n+a:a)}function kI(e,t){let n=[],a=[],r=t!=null&&Array.isArray(t)&&t.length===0,s=t==null||r?null:Ea(t,e).sort(),i=0;for(let o=0;o<e.length;++o){if(s!=null){if(s[i]===o&&e[o]!==1)throw new Error(`Can't squeeze axis ${o} since its dim '${e[o]}' is not 1`);(s[i]==null||s[i]>o)&&e[o]===1&&(n.push(e[o]),a.push(o)),s[i]<=o&&i++}e[o]!==1&&(n.push(e[o]),a.push(o))}return{newShape:n,keptDims:a}}function II(e,t){let n=null;if(e==null||e==="float32")n=new Float32Array(t);else if(e==="int32")n=new Int32Array(t);else if(e==="bool")n=new Uint8Array(t);else throw new Error(`Unknown data type ${e}`);return n}function SI(e,t){let n=null;if(e==null||e==="float32")n=new Float32Array(t);else if(e==="int32")n=new Int32Array(t);else if(e==="bool")n=new Uint8Array(t);else if(e==="string")n=new Array(t);else throw new Error(`Unknown data type ${e}`);return n}function NI(e,t){for(let n=0;n<e.length;n++){let a=e[n];if(isNaN(a)||!isFinite(a))throw Error(`A tensor of type ${t} being uploaded contains ${a}.`)}}function TI(e){return e==="bool"||e==="complex64"||e==="float32"||e==="int32"||e==="string"}function R$(e,t){return!(t==="complex64"||t==="float32"&&e!=="complex64"||t==="int32"&&e!=="float32"&&e!=="complex64"||t==="bool"&&e==="bool")}function hn(e){return e instanceof Float32Array||e instanceof Int32Array||e instanceof Uint8Array||e instanceof Uint8ClampedArray}function Fb(e){if(e==="float32"||e==="int32")return 4;if(e==="complex64")return 8;if(e==="bool")return 1;throw new Error(`Unknown dtype ${e}`)}function _I(e){if(e==null)return 0;let t=0;return e.forEach(n=>t+=n.length),t}function ts(e){return typeof e=="string"||e instanceof String}function CI(e){return typeof e=="boolean"}function EI(e){return typeof e=="number"}function vm(e){return Array.isArray(e)?vm(e[0]):e instanceof Float32Array?"float32":e instanceof Int32Array||e instanceof Uint8Array||e instanceof Uint8ClampedArray?"int32":EI(e)?"float32":ts(e)?"string":CI(e)?"bool":"float32"}function os(e){return!!(e&&e.constructor&&e.call&&e.apply)}function Uh(e,t){for(let n=t;n<e;++n)if(e%n===0)return n;return e}function Ol(e){let t=e.length;if(t<2)return[];let n=new Array(t-1);n[t-2]=e[t-1];for(let a=t-3;a>=0;--a)n[a]=n[a+1]*e[a+1];return n}function FI(e,t,n,a=!1){let r=new Array;if(t.length===1){let s=t[0]*(a?2:1);for(let i=0;i<s;i++)r[i]=n[e+i]}else{let s=t[0],i=t.slice(1),o=i.reduce((l,u)=>l*u)*(a?2:1);for(let l=0;l<s;l++)r[l]=FI(e+l*o,i,n,a)}return r}function gl(e,t,n=!1){if(e.length===0)return t[0];let a=e.reduce((r,s)=>r*s)*(n?2:1);if(a===0)return[];if(a!==t.length)throw new Error(`[${e}] does not match the input size ${t.length}${n?" for a complex tensor":""}.`);return FI(0,e,t,n)}function Ax(e,t){let n=wm(e,t);for(let a=0;a<n.length;a++)n[a]=1;return n}function wm(e,t){if(t==null||t==="float32"||t==="complex64")return new Float32Array(e);if(t==="int32")return new Int32Array(e);if(t==="bool")return new Uint8Array(e);throw new Error(`Unknown data type ${t}`)}function M$(e,t){let n=e.reduce((a,r)=>a*r,1);if(t==null||t==="float32")return gl(e,new Float32Array(n));if(t==="int32")return gl(e,new Int32Array(n));if(t==="bool")return gl(e,new Uint8Array(n));throw new Error(`Unknown data type ${t}`)}function $x(e){e.forEach(t=>{R(Number.isInteger(t)&&t>=0,()=>`Tensor must have a shape comprised of positive integers but got shape [${e}].`)})}function P$(e,t,n){if(t===0)return 0;if(t===1)return e[0];let a=e[e.length-1];for(let r=0;r<e.length-1;++r)a+=n[r]*e[r];return a}function O$(e,t,n){if(t===0)return[];if(t===1)return[e];let a=new Array(t);for(let r=0;r<a.length-1;++r)a[r]=Math.floor(e/n[r]),e-=a[r]*n[r];return a[a.length-1]=e,a}function Dx(e){return e&&e.then&&typeof e.then=="function"}var j1="tfjsflags",AI=class{constructor(e){this.global=e,this.flags={},this.flagRegistry={},this.urlFlags={},this.getQueryParams=L$,this.populateURLFlags()}setPlatform(e,t){this.platform!=null&&(X().getBool("IS_TEST")||X().getBool("PROD")||console.warn(`Platform ${this.platformName} has already been set. Overwriting the platform with ${e}.`)),this.platformName=e,this.platform=t}registerFlag(e,t,n){if(this.flagRegistry[e]={evaluationFn:t,setHook:n},this.urlFlags[e]!=null){let a=this.urlFlags[e];X().getBool("IS_TEST")||X().getBool("PROD")||console.warn(`Setting feature override from URL ${e}: ${a}.`),this.set(e,a)}}async getAsync(e){return e in this.flags?this.flags[e]:(this.flags[e]=await this.evaluateFlag(e),this.flags[e])}get(e){if(e in this.flags)return this.flags[e];let t=this.evaluateFlag(e);if(Dx(t))throw new Error(`Flag ${e} cannot be synchronously evaluated. Please use getAsync() instead.`);return this.flags[e]=t,this.flags[e]}getNumber(e){return this.get(e)}getBool(e){return this.get(e)}getFlags(){return this.flags}get features(){return this.flags}set(e,t){if(this.flagRegistry[e]==null)throw new Error(`Cannot set flag ${e} as it has not been registered.`);this.flags[e]=t,this.flagRegistry[e].setHook!=null&&this.flagRegistry[e].setHook(t)}evaluateFlag(e){if(this.flagRegistry[e]==null)throw new Error(`Cannot evaluate flag '${e}': no evaluation function found.`);return this.flagRegistry[e].evaluationFn()}setFlags(e){this.flags=Object.assign({},e)}reset(){this.flags={},this.urlFlags={},this.populateURLFlags()}populateURLFlags(){if(typeof this.global=="undefined"||typeof this.global.location=="undefined"||typeof this.global.location.search=="undefined")return;let e=this.getQueryParams(this.global.location.search);j1 in e&&e[j1].split(",").forEach(t=>{let[n,a]=t.split(":");this.urlFlags[n]=B$(n,a)})}};function L$(e){let t={};return e.replace(/[?&]([^=?&]+)(?:=([^&]*))?/g,(n,...a)=>(z$(t,a[0],a[1]),a.join("="))),t}function z$(e,t,n){e[decodeURIComponent(t)]=decodeURIComponent(n||"")}function B$(e,t){if(t=t.toLowerCase(),t==="true"||t==="false")return t==="true";if(`${+t}`===t)return+t;throw new Error(`Could not parse value flag value ${t} for flag ${e}.`)}function X(){return Rx}var Rx=null;function W$(e){Rx=e}var gb;function $I(){if(gb==null){let e;if(typeof window!="undefined")e=window;else if(typeof global!="undefined")e=global;else if(typeof process!="undefined")e=process;else if(typeof self!="undefined")e=self;else throw new Error("Could not find a global object");gb=e}return gb}function V$(){let e=$I();return e._tfGlobals==null&&(e._tfGlobals=new Map),e._tfGlobals}function Mx(e,t){let n=V$();if(n.has(e))return n.get(e);{let a=t();return n.set(e,a),n.get(e)}}var Ll="Abs",zl="Acos",Bl="Acosh",bs="Add",wi="AddN",Wl="All",Vl="Any",ki="ArgMax",yc="ArgMin",Ul="Asin",Gl="Asinh",Hl="Atan",jl="Atanh",ql="Atan2",Ii="AvgPool",km="AvgPoolGrad",bc="AvgPool3D",Im="AvgPool3DGrad",Si="BatchMatMul",Kl="BatchToSpaceND",Sm="Bincount",DI="BroadcastTo",Nm="BroadcastArgs",Ni="Cast",Ti="Ceil",xs="ClipByValue",Tm="Complex",xc="ComplexAbs",Xl="Concat",_i="Conv2D",_m="Conv2DBackpropFilter",Ci="Conv2DBackpropInput",vc="Conv3D",Cm="Conv3DBackpropFilterV2",Em="Conv3DBackpropInputV2",Ei="Cos",Fi="Cosh",Yl="Cumprod",Ai="Cumsum",Ql="CropAndResize",Fm="DenseBincount",Jl="DepthToSpace",$i="DepthwiseConv2dNative",Am="DepthwiseConv2dNativeBackpropFilter",$m="DepthwiseConv2dNativeBackpropInput",Dm="Diag",wc="Dilation2D",Gh="Dilation2DBackpropInput",Hh="Dilation2DBackpropFilter",Di="RealDiv",Rm="Einsum",Ri="Elu",Mm="EluGrad",Zl="Erf",eu="Equal",Mi="Exp",tu="ExpandDims",nu="Expm1",Pm="FFT",kc="Fill",au="FlipLeftRight",Pi="Floor",Oi="FloorDiv",Li="FusedBatchNorm",ru="GatherV2",su="GatherNd",iu="Greater",zi="GreaterEqual",Bi="Identity",Om="IFFT",Lm="Imag",ou="IsFinite",lu="IsInf",uu="IsNan",Wi="LeakyRelu",pu="Less",cu="LessEqual",zm="LinSpace",Vi="Log",du="Log1p",hu="LogicalAnd",Ic="LogicalNot",Sc="LogicalOr",RI="LogSoftmax",U$="LowerBound",Nc="LRN",Bm="LRNGrad",Ui="Max",Gi="Maximum",Hi="MaxPool",Wm="MaxPoolGrad",Tc="MaxPool3D",Vm="MaxPool3DGrad",Um="MaxPoolWithArgmax",ji="Mean",qi="Min",Ki="Minimum",Xi="MirrorPad",mu="Mod",Gm="Multinomial",Yi="Multiply",fu="Neg",gu="NotEqual",yu="NonMaxSuppressionV3",bu="NonMaxSuppressionV4",xu="NonMaxSuppressionV5",vu="OnesLike",Qi="OneHot",wu="Pack",Ji="PadV2",G$="Pool",Zi="Pow",eo="Prelu",to="Prod",_c="Range",Hm="Real",ku="Reciprocal",no="Relu",Iu="Reshape",Cc="ResizeNearestNeighbor",jm="ResizeNearestNeighborGrad",ao="ResizeBilinear",qm="ResizeBilinearGrad",ro="Relu6",so="Reverse",io="Round",oo="Rsqrt",Su="ScatterNd",Km="SearchSorted",Nu="Select",Tu="Selu",_u="Slice",lo="Sin",Cu="Sinh",Eu="Sign",uo="Sigmoid",Fu="Softplus",po="Sqrt",co="Sum",Au="SpaceToBatchND",$u="SplitV",ho="Softmax",Ec="SparseFillEmptyRows",Du="SparseReshape",Fc="SparseSegmentMean",Ac="SparseSegmentSum",Xm="SparseToDense",mo="SquaredDifference",$c="Square",Ru="StridedSlice",Ym="StringNGrams",Qm="StringSplit",Jm="StringToHashBucketFast",fo="Sub",go="Tan",yo="Tanh",vs="Tile",Mu="TopK",Pu="Transform",Cr="Transpose",Zm="Unique",Ou="Unpack",Dc="UnsortedSegmentSum",H$="UpperBound",Lu="ZerosLike",ws="Step",jh="FromPixels",zu="RotateWithOffset",ai="_FusedMatMul",ri="FusedConv2D",si="FusedDepthwiseConv2D";function es(...e){X().getBool("IS_TEST")||X().getBool("PROD")||console.warn(...e)}function j$(...e){X().getBool("IS_TEST")||X().getBool("PROD")||console.log(...e)}var kl=Mx("kernelRegistry",()=>new Map),ec=Mx("gradRegistry",()=>new Map);function qh(e,t){let n=Px(e,t);return kl.get(n)}function Ab(e){return ec.get(e)}function Kh(e){let t=kl.entries(),n=[];for(;;){let{done:a,value:r}=t.next();if(a)break;let[s,i]=r,[o]=s.split("_");o===e&&n.push(i)}return n}function Rc(e){let{kernelName:t,backendName:n}=e,a=Px(t,n);kl.has(a)&&es(`The kernel '${t}' for backend '${n}' is already registered`),kl.set(a,e)}function MI(e){let{kernelName:t}=e;ec.has(t)&&X().getBool("DEBUG")&&es(`Overriding the gradient for '${t}'`),ec.set(t,e)}function q$(e,t){let n=Px(e,t);if(!kl.has(n))throw new Error(`The kernel '${e}' for backend '${t}' is not registered`);kl.delete(n)}function K$(e){if(!ec.has(e))throw new Error(`The gradient '${e}' for backend is not registered`);ec.delete(e)}function X$(e,t){Kh(e).forEach(n=>{let a=Object.assign({},n,{backendName:t});Rc(a)})}function Px(e,t){return`${t}_${e}`}var w={};Me(w,{arraysEqual:()=>ys,assert:()=>R,assertNonNegativeIntegerDimensions:()=>$x,assertNonNull:()=>vi,assertShapesMatch:()=>Tn,bytesFromStringArray:()=>_I,bytesPerElement:()=>Fb,checkConversionForErrors:()=>NI,clamp:()=>Zp,computeStrides:()=>Ol,createScalarValue:()=>tD,createShuffledIndices:()=>A$,decodeString:()=>Xh,distSquared:()=>_$,encodeString:()=>Pc,fetch:()=>aD,fingerPrint64:()=>eD,flatten:()=>ni,getArrayFromDType:()=>SI,getTypedArrayFromDType:()=>II,hasEncodingLoss:()=>R$,hexToLong:()=>Mc,indexToLoc:()=>O$,inferDtype:()=>vm,inferFromImplicitShape:()=>D$,isBoolean:()=>CI,isFunction:()=>os,isInt:()=>wl,isNumber:()=>EI,isPromise:()=>Dx,isScalarShape:()=>C$,isString:()=>ts,isTypedArray:()=>hn,isValidDtype:()=>TI,locToIndex:()=>P$,makeOnesTypedArray:()=>Ax,makeZerosNestedTypedArray:()=>M$,makeZerosTypedArray:()=>wm,nearestDivisor:()=>Uh,nearestLargerEven:()=>S$,now:()=>tc,parseAxisParam:()=>Ea,randUniform:()=>T$,repeatedTry:()=>$$,rightPad:()=>Xp,shuffle:()=>wI,shuffleCombo:()=>I$,sizeFromShape:()=>bt,sizeToSquarishShape:()=>F$,squeezeShape:()=>kI,sum:()=>N$,swap:()=>Vh,tanh:()=>E$,toNestedArray:()=>gl,toTypedArray:()=>ef});var q1=xi(t$()),js=q1.default||q1;function Mc(e){return js.fromString(e,!0,16)}var PI=Mc("c3a5c85c97cb3127"),Gs=Mc("b492b66fbe98f273"),kn=Mc("9ae16a3b2f90404f");function $b(e){return e.xor(e.shru(47))}function OI(e,t,n){let a=e.slice(t,t+n);return js.fromBytes(Array.from(a),!0,!0)}function yt(e,t){return OI(e,t,8)}function K1(e,t){return OI(e,t,4)}function Zt(e,t){return t===0?e:e.shru(t).or(e.shl(64-t))}function rs(e,t,n=Mc("9ddfea08eb382d69")){let a=e.xor(t).mul(n);a=a.xor(a.shru(47));let r=t.xor(a).mul(n);return r=r.xor(r.shru(47)),r=r.mul(n),r}function Y$(e,t,n,a,r,s){r=r.add(e),s=Zt(s.add(r).add(a),21);let i=r;return r=r.add(t),r=r.add(n),s=s.add(Zt(r,44)),[r.add(a),s.add(i)]}function bh(e,t,n,a){return Y$(yt(e,t),yt(e,t+8),yt(e,t+16),yt(e,t+24),n,a)}function Q$(e,t=e.length){if(t>=8){let n=kn.add(t*2),a=yt(e,0).add(kn),r=yt(e,t-8),s=Zt(r,37).mul(n).add(a),i=Zt(a,25).add(r).mul(n);return rs(s,i,n)}if(t>=4){let n=kn.add(t*2),a=K1(e,0);return rs(a.shl(3).add(t),K1(e,t-4),n)}if(t>0){let n=e[0],a=e[t>>1],r=e[t-1],s=n+(a<<8),i=t+(r<<2);return $b(kn.mul(s).xor(PI.mul(i))).mul(kn)}return kn}function J$(e,t=e.length){let n=kn.add(t*2),a=yt(e,0).mul(Gs),r=yt(e,8),s=yt(e,t-8).mul(n),i=yt(e,t-16).mul(kn);return rs(Zt(a.add(r),43).add(Zt(s,30)).add(i),a.add(Zt(r.add(kn),18)).add(s),n)}function Z$(e,t=e.length){let n=kn.add(t*2),a=yt(e,0).mul(kn),r=yt(e,8),s=yt(e,t-8).mul(n),i=yt(e,t-16).mul(kn),o=Zt(a.add(r),43).add(Zt(s,30)).add(i),l=rs(o,a.add(Zt(r.add(kn),18)).add(s),n),u=yt(e,16).mul(n),p=yt(e,24),d=o.add(yt(e,t-32)).mul(n),c=l.add(yt(e,t-24)).mul(n);return rs(Zt(u.add(p),43).add(Zt(d,30)).add(c),u.add(Zt(p.add(a),18)).add(d),n)}function eD(e,t=e.length){let n=js.fromNumber(81,!0);if(t<=32)return t<=16?Q$(e,t):J$(e,t);if(t<=64)return Z$(e,t);let a=n,r=n.mul(Gs).add(113),s=$b(r.mul(kn).add(113)).mul(kn),i=[js.UZERO,js.UZERO],o=[js.UZERO,js.UZERO];a=a.mul(kn).add(yt(e,0));let l=0,u=(t-1>>6)*64,p=u+(t-1&63)-63;do a=Zt(a.add(r).add(i[0]).add(yt(e,l+8)),37).mul(Gs),r=Zt(r.add(i[1]).add(yt(e,l+48)),42).mul(Gs),a=a.xor(o[1]),r=r.add(i[0]).add(yt(e,l+40)),s=Zt(s.add(o[0]),33).mul(Gs),i=bh(e,l,i[1].mul(Gs),a.add(o[0])),o=bh(e,l+32,s.add(o[1]),r.add(yt(e,l+16))),[s,a]=[a,s],l+=64;while(l!==u);let d=Gs.add(s.and(255).shl(1));return l=p,o[0]=o[0].add(t-1&63),i[0]=i[0].add(o[0]),o[0]=o[0].add(i[0]),a=Zt(a.add(r).add(i[0]).add(yt(e,l+8)),37).mul(d),r=Zt(r.add(i[1]).add(yt(e,l+48)),42).mul(d),a=a.xor(o[1].mul(9)),r=r.add(i[0].mul(9).add(yt(e,l+40))),s=Zt(s.add(o[0]),33).mul(d),i=bh(e,l,i[1].mul(d),a.add(o[0])),o=bh(e,l+32,s.add(o[1]),r.add(yt(e,l+16))),[s,a]=[a,s],rs(rs(i[0],o[0],d).add($b(r).mul(PI)).add(s),rs(i[1],o[1],d).add(a),d)}function tD(e,t){return t==="string"?Pc(e):ef([e],t)}function nD(e,t){return e instanceof Float32Array&&t==="float32"||e instanceof Int32Array&&t==="int32"||e instanceof Uint8Array&&t==="bool"}function ef(e,t){if(t==="string")throw new Error("Cannot convert a string[] to a TypedArray");if(Array.isArray(e)&&(e=ni(e)),X().getBool("DEBUG")&&NI(e,t),nD(e,t))return e;if(t==null||t==="float32"||t==="complex64")return new Float32Array(e);if(t==="int32")return new Int32Array(e);if(t==="bool"){let n=new Uint8Array(e.length);for(let a=0;a<n.length;++a)Math.round(e[a])!==0&&(n[a]=1);return n}else throw new Error(`Unknown data type ${t}`)}function tc(){return X().platform.now()}function aD(e,t){return X().platform.fetch(e,t)}function Pc(e,t="utf-8"){return t=t||"utf-8",X().platform.encode(e,t)}function Xh(e,t="utf-8"){return t=t||"utf-8",X().platform.decode(e,t)}var rD=class{constructor(e,t){this.backendTimer=e,this.logger=t,t==null&&(this.logger=new iD)}profileKernel(e,t,n){let a,r=()=>{a=n()},s,i=tc();if(this.backendTimer.timerAvailable())s=this.backendTimer.time(r);else{r();for(let o of a)o.dataSync();s=Promise.resolve({kernelMs:tc()-i})}if(X().getBool("CHECK_COMPUTATION_FOR_ERRORS"))for(let o=0;o<a.length;o++){let l=a[o];l.data().then(u=>{sD(u,l.dtype,e)})}return{kernelName:e,outputs:a,inputs:t,timeMs:s.then(o=>o.kernelMs),extraInfo:s.then(o=>o.getExtraProfileInfo!=null?o.getExtraProfileInfo():"")}}logKernelProfile(e){let{kernelName:t,outputs:n,timeMs:a,inputs:r,extraInfo:s}=e;n.forEach(i=>{Promise.all([i.data(),a,s]).then(o=>{this.logger.logKernelProfile(t,i,o[0],o[1],r,o[2])})})}};function sD(e,t,n){if(t!=="float32")return!1;for(let a=0;a<e.length;a++){let r=e[a];if(isNaN(r)||!isFinite(r))return console.warn(`Found ${r} in the result of '${n}'`),!0}return!1}var iD=class{logKernelProfile(e,t,n,a,r,s){let i=typeof a=="number"?Xp(`${a}ms`,9):a.error,o=Xp(e,25),l=t.rank,u=t.size,p=Xp(t.shape.toString(),14),d="";for(let c in r){let h=r[c];if(h!=null){let m=h.shape||t.shape,f=m.length;d+=`${c}: ${f}D ${f>0?m:""} `}}console.log(`%c${o} %c${i} %c${l}D ${p} %c${u} %c${d} %c${s}`,"font-weight:bold","color:red","color:blue","color: orange","color: green","color: steelblue")}};function oD(e,t,n){let a={},r={};for(let l=0;l<t.length;l++)a[t[l].id]=!0;for(let l=0;l<e.length;l++){let u=e[l],p=u.inputs;for(let d in p){let c=p[d],h=!1;for(let m=0;m<t.length;m++)if(a[c.id]){u.outputs.forEach(f=>a[f.id]=!0),h=!0,r[u.id]=!0;break}if(h)break}}let s={};s[n.id]=!0;let i={};for(let l=e.length-1;l>=0;l--){let u=e[l],p=u.inputs;for(let d=0;d<u.outputs.length;d++)if(s[u.outputs[d].id]){for(let c in p)s[p[c].id]=!0,i[u.id]=!0;break}}let o=[];for(let l=0;l<e.length;l++){let u=e[l];if(r[u.id]&&i[u.id]){let p={};for(let c in u.inputs){let h=u.inputs[c];a[h.id]&&(p[c]=h)}let d=Object.assign({},u);d.inputs=p,d.outputs=u.outputs,o.push(d)}}return o}function lD(e,t,n,a){for(let r=t.length-1;r>=0;r--){let s=t[r],i=[];if(s.outputs.forEach(l=>{let u=e[l.id];u!=null?i.push(u):i.push(null)}),s.gradient==null)throw new Error(`Cannot compute gradient: gradient function not found for ${s.kernelName}.`);let o=s.gradient(i);for(let l in s.inputs){if(!(l in o))throw new Error(`Cannot backprop through input ${l}. Available gradients found: ${Object.keys(o)}.`);let u=n(()=>o[l]());if(u.dtype!=="float32")throw new Error(`Error in gradient for op ${s.kernelName}. The gradient of input ${l} must have 'float32' dtype, but has '${u.dtype}'`);let p=s.inputs[l];if(!ys(u.shape,p.shape))throw new Error(`Error in gradient for op ${s.kernelName}. The gradient of input '${l}' has shape '${u.shape}', which does not match the shape of the input '${p.shape}'`);if(e[p.id]==null)e[p.id]=u;else{let d=e[p.id];e[p.id]=a(d,u),d.dispose()}}}}var X1=20,Mp=3,yb=7;function uD(e,t,n,a){let r=Ol(t),s=pD(e,t,n,r),i=t.length,o=Fh(e,t,n,r,s),l=["Tensor"];return a&&(l.push(` dtype: ${n}`),l.push(` rank: ${i}`),l.push(` shape: [${t}]`),l.push(" values:")),l.push(o.map(u=>" "+u).join(`
`)),l.join(`
`)}function pD(e,t,n,a){let r=bt(t),s=a[a.length-1],i=new Array(s).fill(0),o=t.length,l=n==="complex64"?Bp(e):e;if(o>1)for(let u=0;u<r/s;u++){let p=u*s;for(let d=0;d<s;d++)i[d]=Math.max(i[d],zp(l[p+d],0,n).length)}return i}function zp(e,t,n){let a;return Array.isArray(e)?a=`${parseFloat(e[0].toFixed(yb))} + ${parseFloat(e[1].toFixed(yb))}j`:ts(e)?a=`'${e}'`:n==="bool"?a=LI(e):a=parseFloat(e.toFixed(yb)).toString(),Xp(a,t)}function LI(e){return e===0?"false":"true"}function Fh(e,t,n,a,r,s=!0){let i=n==="complex64"?2:1,o=t[0],l=t.length;if(l===0){if(n==="complex64"){let f=Bp(e);return[zp(f[0],0,n)]}return n==="bool"?[LI(e[0])]:[e[0].toString()]}if(l===1){if(o>X1){let g=Mp*i,y=Array.from(e.slice(0,g)),b=Array.from(e.slice((o-Mp)*i,o*i));return n==="complex64"&&(y=Bp(y),b=Bp(b)),["["+y.map((x,v)=>zp(x,r[v],n)).join(", ")+", ..., "+b.map((x,v)=>zp(x,r[o-Mp+v],n)).join(", ")+"]"]}let f=n==="complex64"?Bp(e):Array.from(e);return["["+f.map((g,y)=>zp(g,r[y],n)).join(", ")+"]"]}let u=t.slice(1),p=a.slice(1),d=a[0]*i,c=[];if(o>X1){for(let f=0;f<Mp;f++){let g=f*d,y=g+d;c.push(...Fh(e.slice(g,y),u,n,p,r,!1))}c.push("...");for(let f=o-Mp;f<o;f++){let g=f*d,y=g+d;c.push(...Fh(e.slice(g,y),u,n,p,r,f===o-1))}}else for(let f=0;f<o;f++){let g=f*d,y=g+d;c.push(...Fh(e.slice(g,y),u,n,p,r,f===o-1))}let h=l===2?",":"";c[0]="["+c[0]+h;for(let f=1;f<c.length-1;f++)c[f]=" "+c[f]+h;let m=`,
`;for(let f=2;f<l;f++)m+=`
`;return c[c.length-1]=" "+c[c.length-1]+"]"+(s?"":m),c}function Bp(e){let t=[];for(let n=0;n<e.length;n+=2)t.push([e[n],e[n+1]]);return t}var qt=class{constructor(e,t,n){if(this.dtype=t,this.shape=e.slice(),this.size=bt(e),n!=null){let a=n.length;R(a===this.size,()=>`Length of values '${a}' does not match the size inferred by the shape '${this.size}'.`)}if(t==="complex64")throw new Error("complex64 dtype TensorBuffers are not supported. Please create a TensorBuffer for the real and imaginary parts separately and call tf.complex(real, imag).");this.values=n||SI(t,this.size),this.strides=Ol(e)}set(e,...t){t.length===0&&(t=[0]),R(t.length===this.rank,()=>`The number of provided coordinates (${t.length}) must match the rank (${this.rank})`);let n=this.locToIndex(t);this.values[n]=e}get(...e){e.length===0&&(e=[0]);let t=0;for(let a of e){if(a<0||a>=this.shape[t]){let r=`Requested out of range element at ${e}. Buffer shape=${this.shape}`;throw new Error(r)}t++}let n=e[e.length-1];for(let a=0;a<e.length-1;++a)n+=this.strides[a]*e[a];return this.values[n]}locToIndex(e){if(this.rank===0)return 0;if(this.rank===1)return e[0];let t=e[e.length-1];for(let n=0;n<e.length-1;++n)t+=this.strides[n]*e[n];return t}indexToLoc(e){if(this.rank===0)return[];if(this.rank===1)return[e];let t=new Array(this.shape.length);for(let n=0;n<t.length-1;++n)t[n]=Math.floor(e/this.strides[n]),e-=t[n]*this.strides[n];return t[t.length-1]=e,t}get rank(){return this.shape.length}toTensor(){return Ba().makeTensor(this.values,this.shape,this.dtype)}},Ba=null,dl=null,cD=null;function dD(e){Ba=e}function hD(e){dl=e}function mD(e){cD=e}var Fe=class{constructor(e,t,n,a){this.kept=!1,this.isDisposedInternal=!1,this.shape=e.slice(),this.dtype=t||"float32",this.size=bt(e),this.strides=Ol(e),this.dataId=n,this.id=a,this.rankType=this.rank<5?this.rank.toString():"higher"}get rank(){return this.shape.length}async buffer(){let e=await this.data();return dl.buffer(this.shape,this.dtype,e)}bufferSync(){return dl.buffer(this.shape,this.dtype,this.dataSync())}async array(){let e=await this.data();return gl(this.shape,e,this.dtype==="complex64")}arraySync(){return gl(this.shape,this.dataSync(),this.dtype==="complex64")}async data(){this.throwIfDisposed();let e=Ba().read(this.dataId);if(this.dtype==="string"){let t=await e;try{return t.map(n=>Xh(n))}catch(n){throw new Error("Failed to decode the string bytes into utf-8. To get the original bytes, call tensor.bytes().")}}return e}dataToGPU(e){return this.throwIfDisposed(),Ba().readToGPU(this.dataId,e)}dataSync(){this.throwIfDisposed();let e=Ba().readSync(this.dataId);if(this.dtype==="string")try{return e.map(t=>Xh(t))}catch(t){throw new Error("Failed to decode the string bytes into utf-8. To get the original bytes, call tensor.bytes().")}return e}async bytes(){this.throwIfDisposed();let e=await Ba().read(this.dataId);return this.dtype==="string"?e:new Uint8Array(e.buffer)}dispose(){this.isDisposed||(Ba().disposeTensor(this),this.isDisposedInternal=!0)}get isDisposed(){return this.isDisposedInternal}throwIfDisposed(){if(this.isDisposed)throw new Error("Tensor is disposed.")}print(e=!1){return dl.print(this,e)}clone(){return this.throwIfDisposed(),dl.clone(this)}toString(e=!1){let t=this.dataSync();return uD(t,this.shape,this.dtype,e)}cast(e){return this.throwIfDisposed(),dl.cast(this,e)}variable(e=!0,t,n){return this.throwIfDisposed(),Ba().makeVariable(this,e,t,n)}};Object.defineProperty(Fe,Symbol.hasInstance,{value:e=>!!e&&e.data!=null&&e.dataSync!=null&&e.throwIfDisposed!=null});function ne(){return Mx("Tensor",()=>Fe)}ne();var ls=class extends Fe{constructor(e,t,n,a){super(e.shape,e.dtype,e.dataId,a),this.trainable=t,this.name=n}assign(e){if(e.dtype!==this.dtype)throw new Error(`dtype of the new value (${e.dtype}) and previous value (${this.dtype}) must match`);if(!ys(e.shape,this.shape))throw new Error(`shape of the new value (${e.shape}) and previous value (${this.shape}) must match`);Ba().disposeTensor(this),this.dataId=e.dataId,Ba().incRef(this,null)}dispose(){Ba().disposeVariable(this),this.isDisposedInternal=!0}};Object.defineProperty(ls,Symbol.hasInstance,{value:e=>e instanceof Fe&&e.assign!=null&&e.assign instanceof Function});var Ga={};Me(Ga,{assertTypesMatch:()=>zI,getTensorsInContainer:()=>Ox,isTensorInList:()=>gD,makeTypesMatch:()=>At});var Db;(function(e){e.R0="R0",e.R1="R1",e.R2="R2",e.R3="R3",e.R4="R4",e.R5="R5",e.R6="R6"})(Db||(Db={}));var Rb;(function(e){e.float32="float32",e.int32="int32",e.bool="int32",e.complex64="complex64"})(Rb||(Rb={}));var Mb;(function(e){e.float32="float32",e.int32="int32",e.bool="bool",e.complex64="complex64"})(Mb||(Mb={}));var Pb;(function(e){e.float32="float32",e.int32="float32",e.bool="float32",e.complex64="complex64"})(Pb||(Pb={}));var Ob;(function(e){e.float32="complex64",e.int32="complex64",e.bool="complex64",e.complex64="complex64"})(Ob||(Ob={}));var fD={float32:Pb,int32:Rb,bool:Mb,complex64:Ob};function fa(e,t){if(e==="string"||t==="string"){if(e==="string"&&t==="string")return"string";throw new Error(`Can not upcast ${e} with ${t}`)}return fD[e][t]}function tf(e){return fa(e,"int32")}function At(e,t){if(e.dtype===t.dtype)return[e,t];let n=fa(e.dtype,t.dtype);return[e.cast(n),t.cast(n)]}function zI(e,t){R(e.dtype===t.dtype,()=>`The dtypes of the first(${e.dtype}) and second(${t.dtype}) input must match`)}function gD(e,t){return t.some(n=>n.id===e.id)}function Ox(e){let t=[];return BI(e,t,new Set),t}function BI(e,t,n){if(e==null)return;if(e instanceof Fe){t.push(e);return}if(!yD(e))return;let a=e;for(let r in a){let s=a[r];n.has(s)||(n.add(s),BI(s,t,n))}}function yD(e){return Array.isArray(e)||typeof e=="object"}function bb(e){return e.kernelName!=null}var Y1=class{constructor(){this.registeredVariables={},this.nextTapeNodeId=0,this.numBytes=0,this.numTensors=0,this.numStringTensors=0,this.numDataBuffers=0,this.gradientDepth=0,this.kernelDepth=0,this.scopeStack=[],this.numDataMovesStack=[],this.nextScopeId=0,this.tensorInfo=new WeakMap,this.profiling=!1,this.activeProfile={newBytes:0,newTensors:0,peakBytes:0,kernels:[],result:null,get kernelNames(){return Array.from(new Set(this.kernels.map(e=>e.name)))}}}dispose(){for(let e in this.registeredVariables)this.registeredVariables[e].dispose()}},nc=class{constructor(e){this.ENV=e,this.registry={},this.registryFactory={},this.pendingBackendInitId=0,this.state=new Y1}async ready(){if(this.pendingBackendInit!=null)return this.pendingBackendInit.then(()=>{});if(this.backendInstance!=null)return;let e=this.getSortedBackends();for(let t=0;t<e.length;t++){let n=e[t];if(await this.initializeBackend(n).success){await this.setBackend(n);return}}throw new Error("Could not initialize any backends, all backend initializations failed.")}get backend(){if(this.pendingBackendInit!=null)throw new Error(`Backend '${this.backendName}' has not yet been initialized. Make sure to await tf.ready() or await tf.setBackend() before calling other methods`);if(this.backendInstance==null){let{name:e,asyncInit:t}=this.initializeBackendsAndReturnBest();if(t)throw new Error(`The highest priority backend '${e}' has not yet been initialized. Make sure to await tf.ready() or await tf.setBackend() before calling other methods`);this.setBackend(e)}return this.backendInstance}backendNames(){return Object.keys(this.registryFactory)}findBackend(e){if(!(e in this.registry))if(e in this.registryFactory){let{asyncInit:t}=this.initializeBackend(e);if(t)return null}else return null;return this.registry[e]}findBackendFactory(e){return e in this.registryFactory?this.registryFactory[e].factory:null}registerBackend(e,t,n=1){return e in this.registryFactory?(es(`${e} backend was already registered. Reusing existing backend factory.`),!1):(this.registryFactory[e]={factory:t,priority:n},!0)}async setBackend(e){if(this.registryFactory[e]==null)throw new Error(`Backend name '${e}' not found in registry`);if(this.backendName=e,this.registry[e]==null){this.backendInstance=null;let{success:t,asyncInit:n}=this.initializeBackend(e);if(!(n?await t:t))return!1}return this.backendInstance=this.registry[e],this.setupRegisteredKernels(),this.profiler=new rD(this.backendInstance),!0}setupRegisteredKernels(){Kh(this.backendName).forEach(e=>{e.setupFunc!=null&&e.setupFunc(this.backendInstance)})}disposeRegisteredKernels(e){Kh(e).forEach(t=>{t.disposeFunc!=null&&t.disposeFunc(this.registry[e])})}initializeBackend(e){let t=this.registryFactory[e];if(t==null)throw new Error(`Cannot initialize backend ${e}, no registration found.`);try{let n=t.factory();if(n&&!(n instanceof gc)&&typeof n.then=="function"){let a=++this.pendingBackendInitId,r=n.then(s=>a<this.pendingBackendInitId?!1:(this.registry[e]=s,this.pendingBackendInit=null,!0)).catch(s=>(a<this.pendingBackendInitId||(this.pendingBackendInit=null,es(`Initialization of backend ${e} failed`),es(s.stack||s.message)),!1));return this.pendingBackendInit=r,{success:r,asyncInit:!0}}else return this.registry[e]=n,{success:!0,asyncInit:!1}}catch(n){return es(`Initialization of backend ${e} failed`),es(n.stack||n.message),{success:!1,asyncInit:!1}}}removeBackend(e){if(!(e in this.registryFactory))throw new Error(`${e} backend not found in registry`);this.backendName===e&&this.pendingBackendInit!=null&&this.pendingBackendInitId++,e in this.registry&&(this.disposeRegisteredKernels(e),this.registry[e].dispose(),delete this.registry[e]),delete this.registryFactory[e],this.backendName===e&&(this.pendingBackendInit=null,this.backendName=null,this.backendInstance=null)}getSortedBackends(){if(Object.keys(this.registryFactory).length===0)throw new Error("No backend found in registry.");return Object.keys(this.registryFactory).sort((e,t)=>this.registryFactory[t].priority-this.registryFactory[e].priority)}initializeBackendsAndReturnBest(){let e=this.getSortedBackends();for(let t=0;t<e.length;t++){let n=e[t],{success:a,asyncInit:r}=this.initializeBackend(n);if(r||a)return{name:n,asyncInit:r}}throw new Error("Could not initialize any backends, all backend initializations failed.")}moveData(e,t){let n=this.state.tensorInfo.get(t),a=n.backend,r=this.readSync(t),s=a.refCount(t);a.disposeData(t,!0),n.backend=e,e.move(t,r,n.shape,n.dtype,s),this.shouldCheckForMemLeaks()&&this.state.numDataMovesStack[this.state.numDataMovesStack.length-1]++}tidy(e,t){let n=null;if(t==null){if(typeof e!="function")throw new Error("Please provide a function to tidy()");t=e}else{if(typeof e!="string"&&!(e instanceof String))throw new Error("When calling with two arguments, the first argument to tidy() must be a string");if(typeof t!="function")throw new Error("When calling with two arguments, the 2nd argument to tidy() must be a function");n=e}let a;return this.scopedRun(()=>this.startScope(n),()=>this.endScope(a),()=>(a=t(),a instanceof Promise&&console.error("Cannot return a Promise inside of tidy."),a))}scopedRun(e,t,n){e();try{let a=n();return t(),a}catch(a){throw t(),a}}nextTensorId(){return nc.nextTensorId++}nextVariableId(){return nc.nextVariableId++}clone(e){let t=L.runKernel(Bi,{x:e}),n={x:e},a=s=>({x:()=>{let i="float32",o={x:s},l={dtype:i};return L.runKernel(Ni,o,l)}}),r=[];return this.addTapeNode(this.state.activeScope.name,n,[t],a,r,{}),t}runKernel(e,t,n){if(this.backendName==null&&this.backend,qh(e,this.backendName)==null)throw new Error(`Kernel '${e}' not registered for backend '${this.backendName}'`);return this.runKernelFunc({kernelName:e,inputs:t,attrs:n})}shouldCheckForMemLeaks(){return this.ENV.getBool("IS_TEST")}checkKernelForMemLeak(e,t,n){let a=this.backend.numDataIds(),r=0;n.forEach(o=>{r+=o.dtype==="complex64"?3:1});let s=this.state.numDataMovesStack[this.state.numDataMovesStack.length-1],i=a-t-r-s;if(i>0)throw new Error(`Backend '${this.backendName}' has an internal memory leak (${i} data ids) after running '${e}'`)}runKernelFunc(e){let t,n=[],a=this.isTapeOn(),r=this.state.numBytes,s=this.state.numTensors;this.shouldCheckForMemLeaks()&&this.state.numDataMovesStack.push(0);let i;this.backendName==null&&this.backend;let o,l=bb(e)?e.kernelName:this.state.activeScope!=null?this.state.activeScope.name:"";if(bb(e)){let{kernelName:h,inputs:m,attrs:f}=e;this.backendName==null&&this.backend;let g=qh(h,this.backendName);R(g!=null,()=>`Cannot find registered kernel '${h}' for backend '${this.backendName}'`),i=()=>{let y=this.backend.numDataIds();o=g.kernelFunc({inputs:m,attrs:f,backend:this.backend});let b=Array.isArray(o)?o:[o];this.shouldCheckForMemLeaks()&&this.checkKernelForMemLeak(h,y,b);let x=b.map(v=>v.rank!=null?v:this.makeTensorFromTensorInfo(v));if(a){let v=this.getTensorsForGradient(h,m,x);n=this.saveTensorsForBackwardMode(v)}return x}}else{let{forwardFunc:h}=e,m=f=>{!a||(n=f.map(g=>this.keep(this.clone(g))))};i=()=>{let f=this.backend.numDataIds();o=this.tidy(()=>h(this.backend,m));let g=Array.isArray(o)?o:[o];return this.shouldCheckForMemLeaks()&&this.checkKernelForMemLeak(l,f,g),g}}let{inputs:u,attrs:p}=e,d=bb(e)?null:e.backwardsFunc,c;return this.scopedRun(()=>this.state.kernelDepth++,()=>this.state.kernelDepth--,()=>{!this.ENV.getBool("DEBUG")&&!this.state.profiling?t=i():(c=this.profiler.profileKernel(l,u,()=>i()),this.ENV.getBool("DEBUG")&&this.profiler.logKernelProfile(c),t=c.outputs)}),a&&this.addTapeNode(l,u,t,d,n,p),this.state.profiling&&this.state.activeProfile.kernels.push({name:l,bytesAdded:this.state.numBytes-r,totalBytesSnapshot:this.state.numBytes,tensorsAdded:this.state.numTensors-s,totalTensorsSnapshot:this.state.numTensors,inputShapes:Object.keys(u).map(h=>u[h]!=null?u[h].shape:null),outputShapes:t.map(h=>h.shape),kernelTimeMs:c.timeMs,extraInfo:c.extraInfo}),Array.isArray(o)?t:t[0]}saveTensorsForBackwardMode(e){return e.map(t=>this.keep(this.clone(t)))}getTensorsForGradient(e,t,n){let a=Ab(e);if(a!=null){let r=a.inputsToSave||[],s=a.outputsToSave||[],i;a.saveAllInputs?(R(Array.isArray(t),()=>"saveAllInputs is true, expected inputs to be an array."),i=Object.keys(t).map(l=>t[l])):i=r.map(l=>t[l]);let o=n.filter((l,u)=>s[u]);return i.concat(o)}return[]}makeTensor(e,t,n,a){if(e==null)throw new Error("Values passed to engine.makeTensor() are null");n=n||"float32",a=a||this.backend;let r=e;n==="string"&&ts(e[0])&&(r=e.map(o=>Pc(o)));let s=a.write(r,t,n),i=new Fe(t,n,s,this.nextTensorId());if(this.trackTensor(i,a),n==="string"){let o=this.state.tensorInfo.get(s),l=_I(r);this.state.numBytes+=l-o.bytes,o.bytes=l}return i}makeTensorFromDataId(e,t,n,a){n=n||"float32";let r={dataId:e,shape:t,dtype:n};return this.makeTensorFromTensorInfo(r,a)}makeTensorFromTensorInfo(e,t){let{dataId:n,shape:a,dtype:r}=e,s=new Fe(a,r,n,this.nextTensorId());return this.trackTensor(s,t),s}makeVariable(e,t=!0,n,a){n=n||this.nextVariableId().toString(),a!=null&&a!==e.dtype&&(e=e.cast(a));let r=new ls(e,t,n,this.nextTensorId());if(this.state.registeredVariables[r.name]!=null)throw new Error(`Variable with name ${r.name} was already registered`);return this.state.registeredVariables[r.name]=r,this.incRef(r,this.backend),r}trackTensor(e,t){this.state.numTensors++,e.dtype==="string"&&this.state.numStringTensors++;let n=0;e.dtype!=="complex64"&&e.dtype!=="string"&&(n=e.size*Fb(e.dtype)),this.state.numBytes+=n,this.state.tensorInfo.has(e.dataId)||(this.state.numDataBuffers++,this.state.tensorInfo.set(e.dataId,{backend:t||this.backend,dtype:e.dtype,shape:e.shape,bytes:n})),e instanceof ls||this.track(e)}incRef(e,t){this.trackTensor(e,t),this.backend.incRef(e.dataId)}removeDataId(e,t){this.state.tensorInfo.has(e)&&this.state.tensorInfo.get(e).backend===t&&(this.state.tensorInfo.delete(e),this.state.numDataBuffers--)}disposeTensor(e){if(!this.state.tensorInfo.has(e.dataId))return;let t=this.state.tensorInfo.get(e.dataId);if(this.state.numTensors--,e.dtype==="string"&&(this.state.numStringTensors--,this.state.numBytes-=t.bytes),e.dtype!=="complex64"&&e.dtype!=="string"){let n=e.size*Fb(e.dtype);this.state.numBytes-=n}t.backend.disposeData(e.dataId)&&this.removeDataId(e.dataId,t.backend)}disposeVariables(){for(let e in this.state.registeredVariables){let t=this.state.registeredVariables[e];this.disposeVariable(t)}}disposeVariable(e){this.disposeTensor(e),this.state.registeredVariables[e.name]!=null&&delete this.state.registeredVariables[e.name]}memory(){let e=this.backend.memory();return e.numTensors=this.state.numTensors,e.numDataBuffers=this.state.numDataBuffers,e.numBytes=this.state.numBytes,this.state.numStringTensors>0&&(e.unreliable=!0,e.reasons==null&&(e.reasons=[]),e.reasons.push("Memory usage by string tensors is approximate (2 bytes per character)")),e}async profile(e){this.state.profiling=!0;let t=this.state.numBytes,n=this.state.numTensors;this.state.activeProfile.kernels=[],this.state.activeProfile.result=await e(),this.state.profiling=!1,this.state.activeProfile.peakBytes=Math.max(...this.state.activeProfile.kernels.map(a=>a.totalBytesSnapshot)),this.state.activeProfile.newBytes=this.state.numBytes-t,this.state.activeProfile.newTensors=this.state.numTensors-n;for(let a of this.state.activeProfile.kernels)a.kernelTimeMs=await a.kernelTimeMs,a.extraInfo=await a.extraInfo;return this.state.activeProfile}isTapeOn(){return this.state.gradientDepth>0&&this.state.kernelDepth===0}addTapeNode(e,t,n,a,r,s){let i={id:this.state.nextTapeNodeId++,kernelName:e,inputs:t,outputs:n,saved:r},o=Ab(e);o!=null&&(a=o.gradFunc),a!=null&&(i.gradient=l=>(l=l.map((u,p)=>{if(u==null){let d=n[p],c=wm(d.size,d.dtype);return this.makeTensor(c,d.shape,d.dtype)}return u}),a(l.length>1?l:l[0],r,s))),this.state.activeTape.push(i)}keep(e){return e.kept=!0,e}startTape(){this.state.gradientDepth===0&&(this.state.activeTape=[]),this.state.gradientDepth++}endTape(){this.state.gradientDepth--}startScope(e){let t={track:[],name:"unnamed scope",id:this.state.nextScopeId++};e&&(t.name=e),this.state.scopeStack.push(t),this.state.activeScope=t}endScope(e){let t=Ox(e),n=new Set(t.map(r=>r.id));for(let r=0;r<this.state.activeScope.track.length;r++){let s=this.state.activeScope.track[r];!s.kept&&!n.has(s.id)&&s.dispose()}let a=this.state.scopeStack.pop();this.state.activeScope=this.state.scopeStack.length===0?null:this.state.scopeStack[this.state.scopeStack.length-1],t.forEach(r=>{!r.kept&&r.scopeId===a.id&&this.track(r)})}gradients(e,t,n,a=!1){if(R(t.length>0,()=>"gradients() received an empty list of xs."),n!=null&&n.dtype!=="float32")throw new Error(`dy must have 'float32' dtype, but has '${n.dtype}'`);let r=this.scopedRun(()=>this.startTape(),()=>this.endTape(),()=>this.tidy("forward",e));R(r instanceof Fe,()=>"The result y returned by f() must be a tensor.");let s=oD(this.state.activeTape,t,r);if(!a&&s.length===0&&t.length>0)throw new Error("Cannot compute gradient of y=f(x) with respect to x. Make sure that the f you passed encloses all operations that lead from x to y.");return this.tidy("backward",()=>{let i={};i[r.id]=n==null?bD(r.shape):n,lD(i,s,l=>this.tidy(l),xD);let o=t.map(l=>i[l.id]);return this.state.gradientDepth===0&&(this.state.activeTape.forEach(l=>{for(let u of l.saved)u.dispose()}),this.state.activeTape=null),{value:r,grads:o}})}customGrad(e){return R(os(e),()=>"The f passed in customGrad(f) must be a function."),(...t)=>{R(t.every(i=>i instanceof Fe),()=>"The args passed in customGrad(f)(x1, x2,...) must all be tensors");let n,a={};t.forEach((i,o)=>{a[o]=i});let r=(i,o)=>(n=e(...t,o),R(n.value instanceof Fe,()=>"The function f passed in customGrad(f) must return an object where `obj.value` is a tensor"),R(os(n.gradFunc),()=>"The function f passed in customGrad(f) must return an object where `obj.gradFunc` is a function."),n.value),s=(i,o)=>{let l=n.gradFunc(i,o),u=Array.isArray(l)?l:[l];R(u.length===t.length,()=>"The function f passed in customGrad(f) must return an object where `obj.gradFunc` is a function that returns the same number of tensors as inputs passed to f(...)."),R(u.every(d=>d instanceof Fe),()=>"The function f passed in customGrad(f) must return an object where `obj.gradFunc` is a function that returns a list of only tensors.");let p={};return u.forEach((d,c)=>{p[c]=()=>d}),p};return this.runKernelFunc({forwardFunc:r,backwardsFunc:s,inputs:a})}}readSync(e){return this.state.tensorInfo.get(e).backend.readSync(e)}read(e){return this.state.tensorInfo.get(e).backend.read(e)}readToGPU(e,t){return this.state.tensorInfo.get(e).backend.readToGPU(e,t)}async time(e){let t=tc(),n=await this.backend.time(e);return n.wallMs=tc()-t,n}track(e){return this.state.activeScope!=null&&(e.scopeId=this.state.activeScope.id,this.state.activeScope.track.push(e)),e}get registeredVariables(){return this.state.registeredVariables}reset(){this.pendingBackendInitId++,this.state.dispose(),this.ENV.reset(),this.state=new Y1;for(let e in this.registry)this.disposeRegisteredKernels(e),this.registry[e].dispose(),delete this.registry[e];this.backendName=null,this.backendInstance=null,this.pendingBackendInit=null}};nc.nextTensorId=0;nc.nextVariableId=0;function bD(e){let t=Ax(bt(e),"float32");return L.makeTensor(t,e,"float32")}function WI(){let e=$I();if(e._tfengine==null){let t=new AI(e);e._tfengine=new nc(t)}return W$(e._tfengine.ENV),dD(()=>e._tfengine),e._tfengine}var L=WI();function xD(e,t){let n={a:e,b:t};return L.runKernel(bs,n)}var Oc={};Me(Oc,{isBrowser:()=>VI,isMobile:()=>kD,mockIsMobile:()=>wD});function vD(){return typeof navigator!="undefined"&&navigator!=null}var Lb;function wD(e){Lb=e}function kD(e){if(Lb!==void 0)return Lb;if(e||vD()){if(e||(e=navigator),e.product==="ReactNative")return!0;let t=e.userAgent||e.vendor||(typeof window!="undefined"?window.opera:"");if(!t){let n=e;return n.userAgentData&&n.userAgentData.mobile}return/(android|bb\d+|meego).+mobile|avantgo|bada\/|blackberry|blazer|compal|elaine|fennec|hiptop|iemobile|ip(hone|od)|iris|kindle|lge |maemo|midp|mmp|mobile.+firefox|netfront|opera m(ob|in)i|palm( os)?|phone|p(ixi|re)\/|plucker|pocket|psp|series(4|6)0|symbian|treo|up\.(browser|link)|vodafone|wap|windows ce|xda|xiino/i.test(t)||/1207|6310|6590|3gso|4thp|50[1-6]i|770s|802s|a wa|abac|ac(er|oo|s\-)|ai(ko|rn)|al(av|ca|co)|amoi|an(ex|ny|yw)|aptu|ar(ch|go)|as(te|us)|attw|au(di|\-m|r |s )|avan|be(ck|ll|nq)|bi(lb|rd)|bl(ac|az)|br(e|v)w|bumb|bw\-(n|u)|c55\/|capi|ccwa|cdm\-|cell|chtm|cldc|cmd\-|co(mp|nd)|craw|da(it|ll|ng)|dbte|dc\-s|devi|dica|dmob|do(c|p)o|ds(12|\-d)|el(49|ai)|em(l2|ul)|er(ic|k0)|esl8|ez([4-7]0|os|wa|ze)|fetc|fly(\-|_)|g1 u|g560|gene|gf\-5|g\-mo|go(\.w|od)|gr(ad|un)|haie|hcit|hd\-(m|p|t)|hei\-|hi(pt|ta)|hp( i|ip)|hs\-c|ht(c(\-| |_|a|g|p|s|t)|tp)|hu(aw|tc)|i\-(20|go|ma)|i230|iac( |\-|\/)|ibro|idea|ig01|ikom|im1k|inno|ipaq|iris|ja(t|v)a|jbro|jemu|jigs|kddi|keji|kgt( |\/)|klon|kpt |kwc\-|kyo(c|k)|le(no|xi)|lg( g|\/(k|l|u)|50|54|\-[a-w])|libw|lynx|m1\-w|m3ga|m50\/|ma(te|ui|xo)|mc(01|21|ca)|m\-cr|me(rc|ri)|mi(o8|oa|ts)|mmef|mo(01|02|bi|de|do|t(\-| |o|v)|zz)|mt(50|p1|v )|mwbp|mywa|n10[0-2]|n20[2-3]|n30(0|2)|n50(0|2|5)|n7(0(0|1)|10)|ne((c|m)\-|on|tf|wf|wg|wt)|nok(6|i)|nzph|o2im|op(ti|wv)|oran|owg1|p800|pan(a|d|t)|pdxg|pg(13|\-([1-8]|c))|phil|pire|pl(ay|uc)|pn\-2|po(ck|rt|se)|prox|psio|pt\-g|qa\-a|qc(07|12|21|32|60|\-[2-7]|i\-)|qtek|r380|r600|raks|rim9|ro(ve|zo)|s55\/|sa(ge|ma|mm|ms|ny|va)|sc(01|h\-|oo|p\-)|sdk\/|se(c(\-|0|1)|47|mc|nd|ri)|sgh\-|shar|sie(\-|m)|sk\-0|sl(45|id)|sm(al|ar|b3|it|t5)|so(ft|ny)|sp(01|h\-|v\-|v )|sy(01|mb)|t2(18|50)|t6(00|10|18)|ta(gt|lk)|tcl\-|tdg\-|tel(i|m)|tim\-|t\-mo|to(pl|sh)|ts(70|m\-|m3|m5)|tx\-9|up(\.b|g1|si)|utst|v400|v750|veri|vi(rg|te)|vk(40|5[0-3]|\-v)|vm40|voda|vulc|vx(52|53|60|61|70|80|81|83|85|98)|w3c(\-| )|webc|whit|wi(g |nc|nw)|wmlb|wonu|x700|yas\-|your|zeto|zte\-/i.test(t.substr(0,4))}return!1}function VI(){return typeof window!="undefined"&&window.document!=null||typeof WorkerGlobalScope!="undefined"}var _a=X();_a.registerFlag("DEBUG",()=>!1,e=>{e&&console.warn("Debugging mode is ON. The output of every math call will be downloaded to CPU and checked for NaNs. This significantly impacts performance.")});_a.registerFlag("IS_BROWSER",()=>VI());_a.registerFlag("IS_NODE",()=>typeof process!="undefined"&&typeof process.versions!="undefined"&&typeof process.versions.node!="undefined");_a.registerFlag("IS_CHROME",()=>typeof navigator!="undefined"&&navigator!=null&&navigator.userAgent!=null&&/Chrome/.test(navigator.userAgent)&&/Google Inc/.test(navigator.vendor));_a.registerFlag("PROD",()=>!1);_a.registerFlag("TENSORLIKE_CHECK_SHAPE_CONSISTENCY",()=>_a.getBool("DEBUG"));_a.registerFlag("DEPRECATION_WARNINGS_ENABLED",()=>!0);_a.registerFlag("IS_TEST",()=>!1);_a.registerFlag("CHECK_COMPUTATION_FOR_ERRORS",()=>!0);_a.registerFlag("WRAP_TO_IMAGEBITMAP",()=>!1);_a.registerFlag("ENGINE_COMPILE_ONLY",()=>!1);function pr(e,t){let n=e;if(hn(e))return t==="string"?[]:[e.length];if(!Array.isArray(e))return[];let a=[];for(;Array.isArray(n)||hn(n)&&t!=="string";)a.push(n.length),n=n[0];return Array.isArray(e)&&X().getBool("TENSORLIKE_CHECK_SHAPE_CONSISTENCY")&&UI(e,a,[]),a}function UI(e,t,n){if(n=n||[],!Array.isArray(e)&&!hn(e)){R(t.length===0,()=>`Element arr[${n.join("][")}] is a primitive, but should be an array/TypedArray of ${t[0]} elements`);return}R(t.length>0,()=>`Element arr[${n.join("][")}] should be a primitive, but is an array of ${e.length} elements`),R(e.length===t[0],()=>`Element arr[${n.join("][")}] should have ${t[0]} elements, but has ${e.length} elements`);let a=t.slice(1);for(let r=0;r<e.length;++r)UI(e[r],a,n.concat(r))}function Q1(e,t,n,a){if(e!=="string_or_numeric"){if(e==null)throw new Error("Expected dtype cannot be null.");if(e!=="numeric"&&e!==t||e==="numeric"&&t==="string")throw new Error(`Argument '${n}' passed to '${a}' must be ${e} tensor, but got ${t} tensor`)}}function F(e,t,n,a="numeric"){if(e instanceof Fe)return Q1(a,e.dtype,t,n),e;let r=vm(e);if(r!=="string"&&["bool","int32","float32"].indexOf(a)>=0&&(r=a),Q1(a,r,t,n),e==null||!hn(e)&&!Array.isArray(e)&&typeof e!="number"&&typeof e!="boolean"&&typeof e!="string"){let o=e==null?"null":e.constructor.name;throw new Error(`Argument '${t}' passed to '${n}' must be a Tensor or TensorLike, but got '${o}'`)}let s=pr(e,r);!hn(e)&&!Array.isArray(e)&&(e=[e]);let i=r!=="string"?ef(e,r):ni(e,[],!0);return L.makeTensor(i,s,r)}function ac(e,t,n,a="numeric"){if(!Array.isArray(e))throw new Error(`Argument ${t} passed to ${n} must be a \`Tensor[]\` or \`TensorLike[]\``);return e.map((r,s)=>F(r,`${t}[${s}]`,n,a))}var GI="__op";function z(e){let t=Object.keys(e);if(t.length!==1)throw new Error(`Please provide an object with a single key (operation name) mapping to a function. Got an object with ${t.length} keys.`);let n=t[0],a=e[n];n.endsWith("_")&&(n=n.substring(0,n.length-1)),n=n+GI;let r=(...s)=>{L.startScope(n);try{let i=a(...s);return Dx(i)&&console.error("Cannot return a Promise inside of tidy."),L.endScope(i),i}catch(i){throw L.endScope(null),i}};return Object.defineProperty(r,"name",{value:n,configurable:!0}),r}function ID(e,t){let n=F(e,"real","complex"),a=F(t,"imag","complex");Tn(n.shape,a.shape,`real and imag shapes, ${n.shape} and ${a.shape}, must match in call to tf.complex().`);let r={real:n,imag:a};return L.runKernel(Tm,r)}var Ar=z({complex_:ID});function ks(e,t,n,a){if(a==null&&(a=vm(e)),a==="complex64")throw new Error("Cannot construct a complex64 tensor directly. Please use tf.complex(real, imag).");if(!hn(e)&&!Array.isArray(e)&&typeof e!="number"&&typeof e!="boolean"&&typeof e!="string")throw new Error("values passed to tensor(values) must be a number/boolean/string or an array of numbers/booleans/strings, or a TypedArray");if(t!=null){$x(t);let r=bt(t),s=bt(n);R(r===s,()=>`Based on the provided shape, [${t}], the tensor should have ${r} values but has ${s}`);for(let i=0;i<n.length;++i){let o=n[i],l=i===n.length-1?o!==bt(t.slice(i)):!0;R(n[i]===t[i]||!l,()=>`Error creating a new Tensor. Inferred shape (${n}) does not match the provided shape (${t}). `)}}return!hn(e)&&!Array.isArray(e)&&(e=[e]),t=t||n,e=a!=="string"?ef(e,a):ni(e,[],!0),L.makeTensor(e,t,a)}function Jn(e,t,n){let a=pr(e,n);return ks(e,t,a,n)}var zb={float32:4,float16:2,int32:4,uint16:2,uint8:1,bool:1,complex64:8},Yh=4;async function SD(e,t){let n=[],a=[],r=Array.isArray(e)?e.map(i=>i.name):Object.keys(e);for(let i=0;i<r.length;++i){let o=r[i],l=Array.isArray(e)?e[i].tensor:e[o];if(l.dtype!=="float32"&&l.dtype!=="int32"&&l.dtype!=="bool"&&l.dtype!=="string"&&l.dtype!=="complex64")throw new Error(`Unsupported dtype in weight '${o}': ${l.dtype}`);let u={name:o,shape:l.shape,dtype:l.dtype};if(l.dtype==="string"){let p=new Promise(async d=>{let c=await l.bytes(),h=c.reduce((g,y)=>g+y.length,0)+Yh*c.length,m=new Uint8Array(h),f=0;for(let g=0;g<c.length;g++){let y=c[g],b=new Uint8Array(new Uint32Array([y.length]).buffer);m.set(b,f),f+=Yh,m.set(y,f),f+=y.length}d(m)});a.push(p)}else a.push(l.data());t!=null&&(u.group=t),n.push(u)}let s=await Promise.all(a);return{data:ND(s),specs:n}}function HI(e,t){let n={},a,r=0;for(let s of t){let i=s.name,o=s.dtype,l=s.shape,u=bt(l),p;if("quantization"in s){let d=s.quantization;if(d.dtype==="uint8"||d.dtype==="uint16"){if(!("min"in d&&"scale"in d))throw new Error(`Weight ${s.name} with quantization ${d.dtype} doesn't have corresponding metadata min and scale.`)}else if(d.dtype==="float16"){if(o!=="float32")throw new Error(`Weight ${s.name} is quantized with ${d.dtype} which only supports weights of type float32 not ${o}.`)}else throw new Error(`Weight ${s.name} has unknown quantization dtype ${d.dtype}. Supported quantization dtypes are: 'uint8', 'uint16', and 'float16'.`);let c=zb[d.dtype],h=e.slice(r,r+u*c),m=d.dtype==="uint8"?new Uint8Array(h):new Uint16Array(h);if(o==="float32")if(d.dtype==="uint8"||d.dtype==="uint16"){p=new Float32Array(m.length);for(let f=0;f<m.length;f++){let g=m[f];p[f]=g*d.scale+d.min}}else if(d.dtype==="float16")a===void 0&&(a=AD()),p=a(m);else throw new Error(`Unsupported quantization type ${d.dtype} for weight type float32.`);else if(o==="int32"){if(d.dtype!=="uint8"&&d.dtype!=="uint16")throw new Error(`Unsupported quantization type ${d.dtype} for weight type int32.`);p=new Int32Array(m.length);for(let f=0;f<m.length;f++){let g=m[f];p[f]=Math.round(g*d.scale+d.min)}}else throw new Error(`Unsupported dtype in weight '${i}': ${o}`);r+=u*c}else if(o==="string"){let d=bt(s.shape);p=[];for(let c=0;c<d;c++){let h=new Uint32Array(e.slice(r,r+Yh))[0];r+=Yh;let m=new Uint8Array(e.slice(r,r+h));p.push(m),r+=h}}else{let d=zb[o],c=e.slice(r,r+u*d);if(o==="float32")p=new Float32Array(c);else if(o==="int32")p=new Int32Array(c);else if(o==="bool")p=new Uint8Array(c);else if(o==="complex64"){p=new Float32Array(c);let h=new Float32Array(p.length/2),m=new Float32Array(p.length/2);for(let y=0;y<h.length;y++)h[y]=p[y*2],m[y]=p[y*2+1];let f=Jn(h,l,"float32"),g=Jn(m,l,"float32");n[i]=Ar(f,g),f.dispose(),g.dispose()}else throw new Error(`Unsupported dtype in weight '${i}': ${o}`);r+=u*d}o!=="complex64"&&(n[i]=Jn(p,l,o))}return n}function ND(e){if(e===null)throw new Error(`Invalid input value: ${JSON.stringify(e)}`);let t=0,n=[];e.forEach(s=>{if(t+=s.byteLength,n.push(s.byteLength===s.buffer.byteLength?s:new s.constructor(s)),!(s instanceof Float32Array||s instanceof Int32Array||s instanceof Uint8Array))throw new Error(`Unsupported TypedArray subtype: ${s.constructor.name}`)});let a=new Uint8Array(t),r=0;return n.forEach(s=>{a.set(new Uint8Array(s.buffer),r),r+=s.byteLength}),a.buffer}var Lx=typeof Buffer!="undefined"&&(typeof Blob=="undefined"||typeof atob=="undefined"||typeof btoa=="undefined");function J1(e){return Lx?Buffer.byteLength(e):new Blob([e]).size}function TD(e){if(Lx)return Buffer.from(e).toString("base64");let t=new Uint8Array(e),n="";for(let a=0,r=t.length;a<r;a++)n+=String.fromCharCode(t[a]);return btoa(n)}function _D(e){if(Lx){let a=Buffer.from(e,"base64");return a.buffer.slice(a.byteOffset,a.byteOffset+a.byteLength)}let t=atob(e),n=new Uint8Array(t.length);for(let a=0;a<t.length;++a)n.set([t.charCodeAt(a)],a);return n.buffer}function zx(e){if(e.length===1)return e[0];let t=0;e.forEach(r=>{t+=r.byteLength});let n=new Uint8Array(t),a=0;return e.forEach(r=>{n.set(new Uint8Array(r),a),a+=r.byteLength}),n.buffer}function Z1(e){let t="/";for(e=e.trim();e.endsWith(t);)e=e.slice(0,e.length-1);let n=e.split(t);return n[n.length-1]}function jI(e,t){let n={modelTopology:e.modelTopology,format:e.format,generatedBy:e.generatedBy,convertedBy:e.convertedBy,weightsManifest:t};return e.signature!=null&&(n.signature=e.signature),e.userDefinedMetadata!=null&&(n.userDefinedMetadata=e.userDefinedMetadata),e.modelInitializer!=null&&(n.modelInitializer=e.modelInitializer),e.trainingConfig!=null&&(n.trainingConfig=e.trainingConfig),n}async function Bx(e,t){let n={modelTopology:e.modelTopology,format:e.format,generatedBy:e.generatedBy,convertedBy:e.convertedBy};if(e.trainingConfig!=null&&(n.trainingConfig=e.trainingConfig),e.weightsManifest!=null){let[a,r]=await t(e.weightsManifest);n.weightSpecs=a,n.weightData=r}return e.signature!=null&&(n.signature=e.signature),e.userDefinedMetadata!=null&&(n.userDefinedMetadata=e.userDefinedMetadata),e.modelInitializer!=null&&(n.modelInitializer=e.modelInitializer),n}function Lc(e){if(e.modelTopology instanceof ArrayBuffer)throw new Error("Expected JSON model topology, received ArrayBuffer.");return{dateSaved:new Date,modelTopologyType:"JSON",modelTopologyBytes:e.modelTopology==null?0:J1(JSON.stringify(e.modelTopology)),weightSpecsBytes:e.weightSpecs==null?0:J1(JSON.stringify(e.weightSpecs)),weightDataBytes:e.weightData==null?0:e.weightData.byteLength}}function CD(){let e=n=>{let a=n<<13,r=0;for(;(a&8388608)===0;)r-=8388608,a<<=1;return a&=-8388609,r+=947912704,a|r},t=new Uint32Array(2048);t[0]=0;for(let n=1;n<1024;n++)t[n]=e(n);for(let n=1024;n<2048;n++)t[n]=939524096+(n-1024<<13);return t}function ED(){let e=new Uint32Array(64);e[0]=0,e[31]=1199570944,e[32]=2147483648,e[63]=3347054592;for(let t=1;t<31;t++)e[t]=t<<23;for(let t=33;t<63;t++)e[t]=2147483648+(t-32<<23);return e}function FD(){let e=new Uint32Array(64);for(let t=0;t<64;t++)e[t]=1024;return e[0]=e[32]=0,e}function AD(){let e=CD(),t=ED(),n=FD();return a=>{let r=new ArrayBuffer(4*a.length),s=new Uint32Array(r);for(let i=0;i<a.length;i++){let o=a[i],l=e[n[o>>10]+(o&1023)]+t[o>>10];s[i]=l}return new Float32Array(r)}}var Dt=class{constructor(){this.saveRouters=[],this.loadRouters=[]}static getInstance(){return Dt.instance==null&&(Dt.instance=new Dt),Dt.instance}static registerSaveRouter(e){Dt.getInstance().saveRouters.push(e)}static registerLoadRouter(e){Dt.getInstance().loadRouters.push(e)}static getSaveHandlers(e){return Dt.getHandlers(e,"save")}static getLoadHandlers(e,t){return Dt.getHandlers(e,"load",t)}static getHandlers(e,t,n){let a=[];return(t==="load"?Dt.getInstance().loadRouters:Dt.getInstance().saveRouters).forEach(r=>{let s=r(e,n);s!==null&&a.push(s)}),a}},$D=e=>Dt.registerSaveRouter(e),DD=e=>Dt.registerLoadRouter(e),RD=e=>Dt.getSaveHandlers(e),MD=(e,t)=>Dt.getLoadHandlers(e,t),Bb="tensorflowjs",Wb=1,Ys="models_store",ns="model_info_store";function qI(){if(!X().getBool("IS_BROWSER"))throw new Error("Failed to obtain IndexedDB factory because the current environmentis not a web browser.");let e=typeof window=="undefined"?self:window,t=e.indexedDB||e.mozIndexedDB||e.webkitIndexedDB||e.msIndexedDB||e.shimIndexedDB;if(t==null)throw new Error("The current browser does not appear to support IndexedDB.");return t}function Vb(e){let t=e.result;t.createObjectStore(Ys,{keyPath:"modelPath"}),t.createObjectStore(ns,{keyPath:"modelPath"})}var ii=class{constructor(e){if(this.indexedDB=qI(),e==null||!e)throw new Error("For IndexedDB, modelPath must not be null, undefined or empty.");this.modelPath=e}async save(e){if(e.modelTopology instanceof ArrayBuffer)throw new Error("BrowserLocalStorage.save() does not support saving model topology in binary formats yet.");return this.databaseAction(this.modelPath,e)}async load(){return this.databaseAction(this.modelPath)}databaseAction(e,t){return new Promise((n,a)=>{let r=this.indexedDB.open(Bb,Wb);r.onupgradeneeded=()=>Vb(r),r.onsuccess=()=>{let s=r.result;if(t==null){let i=s.transaction(Ys,"readonly"),o=i.objectStore(Ys).get(this.modelPath);o.onsuccess=()=>{if(o.result==null)return s.close(),a(new Error(`Cannot find model with path '${this.modelPath}' in IndexedDB.`));n(o.result.modelArtifacts)},o.onerror=l=>(s.close(),a(o.error)),i.oncomplete=()=>s.close()}else{let i=Lc(t),o=s.transaction(ns,"readwrite"),l=o.objectStore(ns),u=l.put({modelPath:this.modelPath,modelArtifactsInfo:i}),p;u.onsuccess=()=>{p=s.transaction(Ys,"readwrite");let d=p.objectStore(Ys).put({modelPath:this.modelPath,modelArtifacts:t,modelArtifactsInfo:i});d.onsuccess=()=>n({modelArtifactsInfo:i}),d.onerror=c=>{l=o.objectStore(ns);let h=l.delete(this.modelPath);h.onsuccess=()=>(s.close(),a(d.error)),h.onerror=m=>(s.close(),a(d.error))}},u.onerror=d=>(s.close(),a(u.error)),o.oncomplete=()=>{p==null?s.close():p.oncomplete=()=>s.close()}}},r.onerror=s=>a(r.error)})}};ii.URL_SCHEME="indexeddb://";var KI=e=>X().getBool("IS_BROWSER")&&!Array.isArray(e)&&e.startsWith(ii.URL_SCHEME)?PD(e.slice(ii.URL_SCHEME.length)):null;Dt.registerSaveRouter(KI);Dt.registerLoadRouter(KI);function PD(e){return new ii(e)}function OD(e){return e.startsWith(ii.URL_SCHEME)?e.slice(ii.URL_SCHEME.length):e}var LD=class{constructor(){this.indexedDB=qI()}async listModels(){return new Promise((e,t)=>{let n=this.indexedDB.open(Bb,Wb);n.onupgradeneeded=()=>Vb(n),n.onsuccess=()=>{let a=n.result,r=a.transaction(ns,"readonly"),s=r.objectStore(ns).getAll();s.onsuccess=()=>{let i={};for(let o of s.result)i[o.modelPath]=o.modelArtifactsInfo;e(i)},s.onerror=i=>(a.close(),t(s.error)),r.oncomplete=()=>a.close()},n.onerror=a=>t(n.error)})}async removeModel(e){return e=OD(e),new Promise((t,n)=>{let a=this.indexedDB.open(Bb,Wb);a.onupgradeneeded=()=>Vb(a),a.onsuccess=()=>{let r=a.result,s=r.transaction(ns,"readwrite"),i=s.objectStore(ns),o=i.get(e),l;o.onsuccess=()=>{if(o.result==null)return r.close(),n(new Error(`Cannot find model with path '${e}' in IndexedDB.`));{let u=i.delete(e),p=()=>{l=r.transaction(Ys,"readwrite");let d=l.objectStore(Ys).delete(e);d.onsuccess=()=>t(o.result.modelArtifactsInfo),d.onerror=c=>n(o.error)};u.onsuccess=p,u.onerror=d=>(p(),r.close(),n(o.error))}},o.onerror=u=>(r.close(),n(o.error)),s.oncomplete=()=>{l==null?r.close():l.oncomplete=()=>r.close()}},a.onerror=r=>n(a.error)})}},Tr="/",hl="tensorflowjs_models",XI="info",zD="model_topology",BD="weight_specs",WD="weight_data",VD="model_metadata";function YI(e){return{info:[hl,e,XI].join(Tr),topology:[hl,e,zD].join(Tr),weightSpecs:[hl,e,BD].join(Tr),weightData:[hl,e,WD].join(Tr),modelMetadata:[hl,e,VD].join(Tr)}}function QI(e){for(let t of Object.values(e))window.localStorage.removeItem(t)}function UD(e){let t=e.split(Tr);if(t.length<3)throw new Error(`Invalid key format: ${e}`);return t.slice(1,t.length-1).join(Tr)}function GD(e){return e.startsWith(oi.URL_SCHEME)?e.slice(oi.URL_SCHEME.length):e}var oi=class{constructor(e){if(!X().getBool("IS_BROWSER")||typeof window=="undefined"||typeof window.localStorage=="undefined")throw new Error("The current environment does not support local storage.");if(this.LS=window.localStorage,e==null||!e)throw new Error("For local storage, modelPath must not be null, undefined or empty.");this.modelPath=e,this.keys=YI(this.modelPath)}async save(e){if(e.modelTopology instanceof ArrayBuffer)throw new Error("BrowserLocalStorage.save() does not support saving model topology in binary formats yet.");{let t=JSON.stringify(e.modelTopology),n=JSON.stringify(e.weightSpecs),a=Lc(e);try{this.LS.setItem(this.keys.info,JSON.stringify(a)),this.LS.setItem(this.keys.topology,t),this.LS.setItem(this.keys.weightSpecs,n),this.LS.setItem(this.keys.weightData,TD(e.weightData));let r={format:e.format,generatedBy:e.generatedBy,convertedBy:e.convertedBy,signature:e.signature!=null?e.signature:void 0,userDefinedMetadata:e.userDefinedMetadata!=null?e.userDefinedMetadata:void 0,modelInitializer:e.modelInitializer!=null?e.modelInitializer:void 0,trainingConfig:e.trainingConfig!=null?e.trainingConfig:void 0};return this.LS.setItem(this.keys.modelMetadata,JSON.stringify(r)),{modelArtifactsInfo:a}}catch(r){throw QI(this.keys),new Error(`Failed to save model '${this.modelPath}' to local storage: size quota being exceeded is a possible cause of this failure: modelTopologyBytes=${a.modelTopologyBytes}, weightSpecsBytes=${a.weightSpecsBytes}, weightDataBytes=${a.weightDataBytes}.`)}}}async load(){let e=JSON.parse(this.LS.getItem(this.keys.info));if(e==null)throw new Error(`In local storage, there is no model with name '${this.modelPath}'`);if(e.modelTopologyType!=="JSON")throw new Error("BrowserLocalStorage does not support loading non-JSON model topology yet.");let t={},n=JSON.parse(this.LS.getItem(this.keys.topology));if(n==null)throw new Error(`In local storage, the topology of model '${this.modelPath}' is missing.`);t.modelTopology=n;let a=JSON.parse(this.LS.getItem(this.keys.weightSpecs));if(a==null)throw new Error(`In local storage, the weight specs of model '${this.modelPath}' are missing.`);t.weightSpecs=a;let r=this.LS.getItem(this.keys.modelMetadata);if(r!=null){let i=JSON.parse(r);t.format=i.format,t.generatedBy=i.generatedBy,t.convertedBy=i.convertedBy,i.signature!=null&&(t.signature=i.signature),i.userDefinedMetadata!=null&&(t.userDefinedMetadata=i.userDefinedMetadata),i.modelInitializer!=null&&(t.modelInitializer=i.modelInitializer),i.trainingConfig!=null&&(t.trainingConfig=i.trainingConfig)}let s=this.LS.getItem(this.keys.weightData);if(s==null)throw new Error(`In local storage, the binary weight values of model '${this.modelPath}' are missing.`);return t.weightData=_D(s),t}};oi.URL_SCHEME="localstorage://";var JI=e=>X().getBool("IS_BROWSER")&&!Array.isArray(e)&&e.startsWith(oi.URL_SCHEME)?HD(e.slice(oi.URL_SCHEME.length)):null;Dt.registerSaveRouter(JI);Dt.registerLoadRouter(JI);function HD(e){return new oi(e)}var jD=class{constructor(){R(X().getBool("IS_BROWSER"),()=>"Current environment is not a web browser"),R(typeof window=="undefined"||typeof window.localStorage!="undefined",()=>"Current browser does not appear to support localStorage"),this.LS=window.localStorage}async listModels(){let e={},t=hl+Tr,n=Tr+XI;for(let a=0;a<this.LS.length;++a){let r=this.LS.key(a);if(r.startsWith(t)&&r.endsWith(n)){let s=UD(r);e[s]=JSON.parse(this.LS.getItem(r))}}return e}async removeModel(e){e=GD(e);let t=YI(e);if(this.LS.getItem(t.info)==null)throw new Error(`Cannot find model at path '${e}'`);let n=JSON.parse(this.LS.getItem(t.info));return QI(t),n}},yl="://",ca=class{constructor(){this.managers={}}static getInstance(){return ca.instance==null&&(ca.instance=new ca),ca.instance}static registerManager(e,t){R(e!=null,()=>"scheme must not be undefined or null."),e.endsWith(yl)&&(e=e.slice(0,e.indexOf(yl))),R(e.length>0,()=>"scheme must not be an empty string.");let n=ca.getInstance();R(n.managers[e]==null,()=>`A model store manager is already registered for scheme '${e}'.`),n.managers[e]=t}static getManager(e){let t=this.getInstance().managers[e];if(t==null)throw new Error(`Cannot find model manager for scheme '${e}'`);return t}static getSchemes(){return Object.keys(this.getInstance().managers)}};function Ah(e){if(e.indexOf(yl)===-1)throw new Error(`The url string provided does not contain a scheme. Supported schemes are: ${ca.getSchemes().join(",")}`);return{scheme:e.split(yl)[0],path:e.split(yl)[1]}}async function ZI(e,t,n=!1){R(e!==t,()=>`Old path and new path are the same: '${e}'`);let a=Dt.getLoadHandlers(e);R(a.length>0,()=>`Copying failed because no load handler is found for source URL ${e}.`),R(a.length<2,()=>`Copying failed because more than one (${a.length}) load handlers for source URL ${e}.`);let r=a[0],s=Dt.getSaveHandlers(t);R(s.length>0,()=>`Copying failed because no save handler is found for destination URL ${t}.`),R(s.length<2,()=>`Copying failed because more than one (${a.length}) save handlers for destination URL ${t}.`);let i=s[0],o=Ah(e).scheme,l=Ah(e).path,u=o===Ah(e).scheme,p=await r.load();n&&u&&await ca.getManager(o).removeModel(l);let d=await i.save(p);return n&&!u&&await ca.getManager(o).removeModel(l),d.modelArtifactsInfo}async function qD(){let e=ca.getSchemes(),t={};for(let n of e){let a=await ca.getManager(n).listModels();for(let r in a){let s=n+yl+r;t[s]=a[r]}}return t}async function KD(e){let t=Ah(e);return ca.getManager(t.scheme).removeModel(t.path)}async function XD(e,t){return ZI(e,t,!1)}async function YD(e,t){return ZI(e,t,!0)}var QD=class{fetch(e,t){return fetch(e,t)}now(){return performance.now()}encode(e,t){if(t!=="utf-8"&&t!=="utf8")throw new Error(`Browser's encoder only supports utf-8, but got ${t}`);return this.textEncoder==null&&(this.textEncoder=new TextEncoder),this.textEncoder.encode(e)}decode(e,t){return new TextDecoder(t).decode(e)}};if(X().get("IS_BROWSER")){X().setPlatform("browser",new QD);try{ca.registerManager(oi.URL_SCHEME,new jD)}catch(e){}try{ca.registerManager(ii.URL_SCHEME,new LD)}catch(e){}}var JD={importFetch:()=>n$()},xb,ZD=class{constructor(){this.util=a$(),this.textEncoder=new this.util.TextEncoder}fetch(e,t){return X().global.fetch!=null?X().global.fetch(e,t):(xb==null&&(xb=JD.importFetch()),xb(e,t))}now(){let e=process.hrtime();return e[0]*1e3+e[1]/1e6}encode(e,t){if(t!=="utf-8"&&t!=="utf8")throw new Error(`Node built-in encoder only supports utf-8, but got ${t}`);return this.textEncoder.encode(e)}decode(e,t){return e.length===0?"":new this.util.TextDecoder(t).decode(e)}};X().get("IS_NODE")&&!X().get("IS_BROWSER")&&X().setPlatform("node",new ZD);function Ve(e,t="float32",n){return t=t||"float32",$x(e),new qt(e,t,n)}function eR(e,t){let n=F(e,"x","cast");if(!TI(t))throw new Error(`Failed to cast to unknown dtype ${t}`);if(t==="string"&&n.dtype!=="string"||t!=="string"&&n.dtype==="string")throw new Error("Only strings can be casted to strings");let a={x:n},r={dtype:t};return L.runKernel(Ni,a,r)}var oe=z({cast_:eR});function tR(e){let t={x:F(e,"x","clone","string_or_numeric")};return L.runKernel(Bi,t)}var Er=z({clone_:tR});function eS(e,t=!1){console.log(e.toString(t))}WI();var nR={buffer:Ve,cast:oe,clone:Er,print:eS};hD(nR);var en={};Me(en,{browserFiles:()=>uR,browserHTTPRequest:()=>mR,concatenateArrayBuffers:()=>zx,copyModel:()=>XD,decodeWeights:()=>HI,encodeWeights:()=>SD,fromMemory:()=>gR,fromMemorySync:()=>sS,getLoadHandlers:()=>MD,getModelArtifactsForJSON:()=>Bx,getModelArtifactsInfoForJSON:()=>Lc,getSaveHandlers:()=>RD,http:()=>Vx,isHTTPScheme:()=>Ub,listModels:()=>qD,loadWeights:()=>pR,moveModel:()=>YD,registerLoadRouter:()=>DD,registerSaveRouter:()=>$D,removeModel:()=>KD,weightsLoaderFactory:()=>nS,withSaveHandler:()=>yR,withSaveHandlerSync:()=>bR});var aR="model",rR=".json",sR=".weights.bin";function ek(e){return new Promise(t=>setTimeout(t)).then(e)}var Il=class{constructor(e){if(!X().getBool("IS_BROWSER"))throw new Error("browserDownloads() cannot proceed because the current environment is not a browser.");e.startsWith(Il.URL_SCHEME)&&(e=e.slice(Il.URL_SCHEME.length)),(e==null||e.length===0)&&(e=aR),this.modelJsonFileName=e+rR,this.weightDataFileName=e+sR}async save(e){if(typeof document=="undefined")throw new Error("Browser downloads are not supported in this environment since `document` is not present");let t=window.URL.createObjectURL(new Blob([e.weightData],{type:"application/octet-stream"}));if(e.modelTopology instanceof ArrayBuffer)throw new Error("BrowserDownloads.save() does not support saving model topology in binary formats yet.");{let n=[{paths:["./"+this.weightDataFileName],weights:e.weightSpecs}],a=jI(e,n),r=window.URL.createObjectURL(new Blob([JSON.stringify(a)],{type:"application/json"})),s=this.modelJsonAnchor==null?document.createElement("a"):this.modelJsonAnchor;if(s.download=this.modelJsonFileName,s.href=r,await ek(()=>s.dispatchEvent(new MouseEvent("click"))),e.weightData!=null){let i=this.weightDataAnchor==null?document.createElement("a"):this.weightDataAnchor;i.download=this.weightDataFileName,i.href=t,await ek(()=>i.dispatchEvent(new MouseEvent("click")))}return{modelArtifactsInfo:Lc(e)}}}};Il.URL_SCHEME="downloads://";var iR=class{constructor(e){if(e==null||e.length<1)throw new Error(`When calling browserFiles, at least 1 file is required, but received ${e}`);this.jsonFile=e[0],this.weightsFiles=e.slice(1)}async load(){return new Promise((e,t)=>{let n=new FileReader;n.onload=a=>{let r=JSON.parse(a.target.result),s=r.modelTopology;if(s==null){t(new Error(`modelTopology field is missing from file ${this.jsonFile.name}`));return}if(r.weightsManifest==null){t(new Error(`weightManifest field is missing from file ${this.jsonFile.name}`));return}if(this.weightsFiles.length===0){e({modelTopology:s});return}let i=Bx(r,o=>this.loadWeights(o));e(i)},n.onerror=a=>t(`Failed to read model topology and weights manifest JSON from file '${this.jsonFile.name}'. BrowserFiles supports loading Keras-style tf.Model artifacts only.`),n.readAsText(this.jsonFile)})}loadWeights(e){let t=[],n=[];for(let s of e)t.push(...s.weights),n.push(...s.paths);let a=this.checkManifestAndWeightFiles(e),r=n.map(s=>this.loadWeightsFile(s,a[s]));return Promise.all(r).then(s=>[t,zx(s)])}loadWeightsFile(e,t){return new Promise((n,a)=>{let r=new FileReader;r.onload=s=>{let i=s.target.result;n(i)},r.onerror=s=>a(`Failed to weights data from file of path '${e}'.`),r.readAsArrayBuffer(t)})}checkManifestAndWeightFiles(e){let t=[],n=this.weightsFiles.map(r=>Z1(r.name)),a={};for(let r of e)r.paths.forEach(s=>{let i=Z1(s);if(t.indexOf(i)!==-1)throw new Error(`Duplicate file basename found in weights manifest: '${i}'`);if(t.push(i),n.indexOf(i)===-1)throw new Error(`Weight file with basename '${i}' is not provided.`);a[s]=this.weightsFiles[n.indexOf(i)]});if(t.length!==this.weightsFiles.length)throw new Error(`Mismatch in the number of files in weights manifest (${t.length}) and the number of weight files provided (${this.weightsFiles.length}).`);return a}},oR=e=>X().getBool("IS_BROWSER")&&!Array.isArray(e)&&e.startsWith(Il.URL_SCHEME)?lR(e.slice(Il.URL_SCHEME.length)):null;Dt.registerSaveRouter(oR);function lR(e="model"){return new Il(e)}function uR(e){return new iR(e)}function tk(e,t,n,a){i(e),n=n==null?0:n,a=a==null?1:a,o(n,a);let r=0,s=l=>(l.then(u=>{let p=n+ ++r/e.length*(a-n);return t(p),u}),l);function i(l){R(l!=null&&Array.isArray(l)&&l.length>0,()=>"promises must be a none empty array")}function o(l,u){R(l>=0&&l<=1,()=>`Progress fraction must be in range [0, 1], but got startFraction ${l}`),R(u>=0&&u<=1,()=>`Progress fraction must be in range [0, 1], but got endFraction ${u}`),R(u>=l,()=>`startFraction must be no more than endFraction, but got startFraction ${l} and endFraction ${u}`)}return Promise.all(e.map(s))}async function tS(e,t){t==null&&(t={});let n=t.fetchFunc==null?X().platform.fetch:t.fetchFunc,a=e.map(u=>n(u,t.requestInit,{isBinary:!0})),r=0,s=.5,i=(t.onProgress==null?await Promise.all(a):await tk(a,t.onProgress,r,s)).map(u=>u.arrayBuffer()),o=.5,l=1;return t.onProgress==null?await Promise.all(i):await tk(i,t.onProgress,o,l)}async function pR(e,t="",n,a){return nS(r=>tS(r,{requestInit:a}))(e,t,n)}function nS(e){return async(t,n="",a)=>{let r=t.map(()=>!1),s={},i=a!=null?a.map(()=>!1):[],o=[];if(t.forEach((h,m)=>{let f=0;h.weights.forEach(g=>{let y="quantization"in g?g.quantization.dtype:g.dtype,b=zb[y]*bt(g.shape),x=()=>{r[m]=!0,s[m]==null&&(s[m]=[]),s[m].push({manifestEntry:g,groupOffset:f,sizeBytes:b})};a!=null?a.forEach((v,k)=>{v===g.name&&(x(),i[k]=!0)}):x(),o.push(g.name),f+=b})}),!i.every(h=>h)){let h=a.filter((m,f)=>!i[f]);throw new Error(`Could not find weights in manifest with names: ${h.join(", ")}.
Manifest JSON has weights with names: ${o.join(", ")}.`)}let l=r.reduce((h,m,f)=>(m&&h.push(f),h),[]),u=[];l.forEach(h=>{t[h].paths.forEach(m=>{let f=n+(n.endsWith("/")?"":"/")+m;u.push(f)})});let p=await e(u),d={},c=0;return l.forEach(h=>{let m=t[h].paths.length,f=0;for(let x=0;x<m;x++)f+=p[c+x].byteLength;let g=new ArrayBuffer(f),y=new Uint8Array(g),b=0;for(let x=0;x<m;x++){let v=new Uint8Array(p[c+x]);y.set(v,b),b+=v.byteLength}s[h].forEach(x=>{let v=g.slice(x.groupOffset,x.groupOffset+x.sizeBytes),k=HI(v,[x.manifestEntry]);for(let T in k)d[T]=k[T]}),c+=m}),d}}var cR="application/octet-stream",dR="application/json",Wx=class{constructor(e,t){if(this.DEFAULT_METHOD="POST",t==null&&(t={}),this.weightPathPrefix=t.weightPathPrefix,this.onProgress=t.onProgress,this.weightUrlConverter=t.weightUrlConverter,t.fetchFunc!=null?(R(typeof t.fetchFunc=="function",()=>"Must pass a function that matches the signature of `fetch` (see https://developer.mozilla.org/en-US/docs/Web/API/Fetch_API)"),this.fetch=t.fetchFunc):this.fetch=X().platform.fetch,R(e!=null&&e.length>0,()=>"URL path for http must not be null, undefined or empty."),Array.isArray(e)&&R(e.length===2,()=>`URL paths for http must have a length of 2, (actual length is ${e.length}).`),this.path=e,t.requestInit!=null&&t.requestInit.body!=null)throw new Error("requestInit is expected to have no pre-existing body, but has one.");this.requestInit=t.requestInit||{}}async save(e){if(e.modelTopology instanceof ArrayBuffer)throw new Error("BrowserHTTPRequest.save() does not support saving model topology in binary formats yet.");let t=Object.assign({method:this.DEFAULT_METHOD},this.requestInit);t.body=new FormData;let n=[{paths:["./model.weights.bin"],weights:e.weightSpecs}],a=jI(e,n);t.body.append("model.json",new Blob([JSON.stringify(a)],{type:dR}),"model.json"),e.weightData!=null&&t.body.append("model.weights.bin",new Blob([e.weightData],{type:cR}),"model.weights.bin");let r=await this.fetch(this.path,t);if(r.ok)return{modelArtifactsInfo:Lc(e),responses:[r]};throw new Error(`BrowserHTTPRequest.save() failed due to HTTP response status ${r.status}.`)}async load(){let e=await this.fetch(this.path,this.requestInit);if(!e.ok)throw new Error(`Request to ${this.path} failed with status code ${e.status}. Please verify this URL points to the model JSON of the model to load.`);let t;try{t=await e.json()}catch(r){let s=`Failed to parse model JSON of response from ${this.path}.`;throw this.path.endsWith(".pb")?s+=" Your path contains a .pb file extension. Support for .pb models have been removed in TensorFlow.js 1.0 in favor of .json models. You can re-convert your Python TensorFlow model using the TensorFlow.js 1.0 conversion scripts or you can convert your.pb models with the 'pb2json'NPM script in the tensorflow/tfjs-converter repository.":s+=" Please make sure the server is serving valid JSON for this request.",new Error(s)}let n=t.modelTopology,a=t.weightsManifest;if(n==null&&a==null)throw new Error(`The JSON from HTTP path ${this.path} contains neither model topology or manifest for weights.`);return Bx(t,r=>this.loadWeights(r))}async loadWeights(e){let t=Array.isArray(this.path)?this.path[1]:this.path,[n,a]=hR(t),r=this.weightPathPrefix||n,s=[];for(let u of e)s.push(...u.weights);let i=[],o=[];for(let u of e)for(let p of u.paths)this.weightUrlConverter!=null?o.push(this.weightUrlConverter(p)):i.push(r+p+a);this.weightUrlConverter&&i.push(...await Promise.all(o));let l=await tS(i,{requestInit:this.requestInit,fetchFunc:this.fetch,onProgress:this.onProgress});return[s,zx(l)]}};Wx.URL_SCHEME_REGEX=/^https?:\/\//;function hR(e){let t=e.lastIndexOf("/"),n=e.lastIndexOf("?"),a=e.substring(0,t),r=n>t?e.substring(n):"";return[a+"/",r]}function Ub(e){return e.match(Wx.URL_SCHEME_REGEX)!=null}var aS=(e,t)=>{if(typeof fetch=="undefined"&&(t==null||t.fetchFunc==null))return null;{let n=!0;if(Array.isArray(e)?n=e.every(a=>Ub(a)):n=Ub(e),n)return Vx(e,t)}return null};Dt.registerSaveRouter(aS);Dt.registerLoadRouter(aS);function Vx(e,t){return new Wx(e,t)}function mR(e,t){return Vx(e,t)}var vb=class{constructor(e){this.modelArtifacts=e}load(){return this.modelArtifacts}},rS=class{constructor(e){this.saveHandler=e}save(e){return this.saveHandler(e)}},fR=class{constructor(e){e.load&&(this.load=()=>Promise.resolve(e.load())),e.save&&(this.save=t=>Promise.resolve(e.save(t)))}};function gR(e,t,n,a){let r=arguments;return new fR(sS(...r))}function sS(e,t,n,a){return arguments.length===1?e.modelTopology!=null||e.weightSpecs!=null?new vb(e):(console.warn("Please call tf.io.fromMemory() with only one argument. The argument should be of type ModelArtifacts. The multi-argument signature of tf.io.fromMemory() has been deprecated and will be removed in a future release."),new vb({modelTopology:e})):(console.warn("Please call tf.io.fromMemory() with only one argument. The argument should be of type ModelArtifacts. The multi-argument signature of tf.io.fromMemory() has been deprecated and will be removed in a future release."),new vb({modelTopology:e,weightSpecs:t,weightData:n,trainingConfig:a}))}function yR(e){return new rS(e)}function bR(e){return new rS(e)}var iS={};Me(iS,{confusionMatrix:()=>zR});function xR(e,t,n=!1,a=!1){let r=F(e,"a","matMul"),s=F(t,"b","matMul");[r,s]=At(r,s);let i={a:r,b:s},o={transposeA:n,transposeB:a};return L.runKernel(Si,i,o)}var De=z({matMul_:xR});function vR(e,t,n=1,a=0){if(t<2)throw new Error(`Error in oneHot: depth must be >=2, but it is ${t}`);let r={indices:F(e,"indices","oneHot","int32")},s={depth:t,onValue:n,offValue:a};return L.runKernel(Qi,r,s)}var Sl=z({oneHot_:vR});function wR(){X().set("PROD",!0)}function kR(){X().set("DEBUG",!0)}function IR(){X().set("DEPRECATION_WARNINGS_ENABLED",!1),console.warn("TensorFlow.js deprecation warnings have been disabled.")}function Ux(e){X().getBool("DEPRECATION_WARNINGS_ENABLED")&&console.warn(e+" You can disable deprecation warnings with tf.disableDeprecationWarnings().")}mD(Ux);function SR(){L.disposeVariables()}function sr(){return L}function Qh(){return L.memory()}function NR(e){return L.profile(e)}function O(e,t){return L.tidy(e,t)}function Re(e){Ox(e).forEach(t=>t.dispose())}function tn(e){return L.keep(e)}function TR(e){return L.time(e)}function _R(e){return L.setBackend(e)}function CR(){return L.ready()}function ER(){return L.backendName}function FR(e){L.removeBackend(e)}function AR(e){return L.findBackend(e)}function $R(e){return L.findBackendFactory(e)}function nf(e,t,n=1){return L.registerBackend(e,t,n)}function oS(){return L.backend}function DR(e,t){X().setPlatform(e,t)}function RR(e){let t={input:F(e,"input","imag")};return L.runKernel(Lm,t)}var zc=z({imag_:RR});function MR(e){let t={x:F(e,"x","neg")};return L.runKernel(fu,t)}var It=z({neg_:MR});function PR(e){let t={input:F(e,"input","real")};return L.runKernel(Hm,t)}var Nl=z({real_:PR});function OR(e,t,n){let a=F(e,"x","transpose");if(t==null&&(t=a.shape.map((i,o)=>o).reverse()),R(a.rank===t.length,()=>`Error in transpose: rank of input ${a.rank} must match length of perm ${t}.`),t.forEach(i=>{R(i>=0&&i<a.rank,()=>`All entries in 'perm' must be between 0 and ${a.rank-1} but got ${t}`)}),a.rank<=1)return a.clone();let r={x:a},s={perm:t};return a.dtype==="complex64"?O(()=>{let i=Nl(a),o=zc(a);return i=L.runKernel(Cr,{x:i},s),o=L.runKernel(Cr,{x:o},s),n&&(o=It(o)),Ar(i,o)}):L.runKernel(Cr,r,s)}var Ae=z({transpose_:OR});function LR(e,t,n){let a=F(e,"labels","confusionMatrix"),r=F(t,"predictions","confusionMatrix");R(n==null||n>0&&Number.isInteger(n),()=>`If provided, numClasses must be a positive integer, but got ${n}`),R(a.rank===1,()=>`Expected the rank of labels to be 1, but got ${a.rank}`),R(r.rank===1,()=>`Expected the rank of predictions to be 1, but got ${r.rank}`),R(a.shape[0]===r.shape[0],()=>`Mismatch in the number of examples: ${a.shape[0]} vs. ${r.shape[0]}. Labels and predictions should have the same number of elements.`),R(n>0&&Number.isInteger(n),()=>`numClasses is required to be a positive integer, but got ${n}`);let s=Sl(oe(a,"int32"),n),i=Sl(oe(r,"int32"),n),o=Ae(s),l=De(o,i);return oe(l,"int32")}var zR=z({confusionMatrix_:LR}),Bu={};Me(Bu,{assertAndGetBroadcastShape:()=>pt,getBroadcastDims:()=>lS,getReductionAxes:()=>Vt});function lS(e,t){let n=e.length,a=[];for(let r=0;r<n;r++){let s=n-1-r,i=e[s]||1;(t[t.length-1-r]||1)>1&&i===1&&a.unshift(s)}return a}function Vt(e,t){let n=[];for(let a=0;a<t.length;a++){let r=e[e.length-a-1],s=t.length-a-1,i=t[s];(r==null||r===1&&i>1)&&n.unshift(s)}return n}function pt(e,t){let n=[],a=Math.max(e.length,t.length);for(let r=0;r<a;r++){let s=e[e.length-r-1];s==null&&(s=1);let i=t[t.length-r-1];if(i==null&&(i=1),s===1)n.unshift(i);else if(i===1)n.unshift(s);else if(s!==i){let o=`Operands could not be broadcast together with shapes ${e} and ${t}.`;throw Error(o)}else n.unshift(s)}return n}var bo={};Me(bo,{fromPixels:()=>jR,fromPixelsAsync:()=>GR,toPixels:()=>HR});function af(e,t,n){if(vi(e),t!=null&&t.length!==3)throw new Error("tensor3d() requires shape to have three numbers");let a=pr(e,n);if(a.length!==3&&a.length!==1)throw new Error("tensor3d() requires values to be number[][][] or flat/TypedArray");if(a.length===1&&t==null)throw new Error("tensor3d() requires shape to be provided when `values` are a flat array");return ks(e,t,a,n)}var Vs;function uS(e,t=3){if(t>4)throw new Error("Cannot construct Tensor with more than 4 channels from pixels.");if(e==null)throw new Error("pixels passed to tf.browser.fromPixels() can not be null");let n=!1,a=!1,r=!1,s=!1,i=!1,o=!1;if(e.data instanceof Uint8Array)n=!0;else if(typeof ImageData!="undefined"&&e instanceof ImageData)a=!0;else if(typeof HTMLVideoElement!="undefined"&&e instanceof HTMLVideoElement)r=!0;else if(typeof HTMLImageElement!="undefined"&&e instanceof HTMLImageElement)s=!0;else if(e.getContext!=null)i=!0;else if(typeof ImageBitmap!="undefined"&&e instanceof ImageBitmap)o=!0;else throw new Error(`pixels passed to tf.browser.fromPixels() must be either an HTMLVideoElement, HTMLImageElement, HTMLCanvasElement, ImageData in browser, or OffscreenCanvas, ImageData in webworker or {data: Uint32Array, width: number, height: number}, but was ${e.constructor.name}`);if(r&&r&&e.readyState<2)throw new Error("The video element has not loaded data yet. Please wait for `loadeddata` event on the <video> element.");if(qh(jh,L.backendName)!=null){let c={pixels:e},h={numChannels:t};return L.runKernel(jh,c,h)}let[l,u]=r?[e.videoWidth,e.videoHeight]:[e.width,e.height],p;if(i)p=e.getContext("2d").getImageData(0,0,l,u).data;else if(a||n)p=e.data;else if(s||r||o){if(Vs==null)if(typeof document=="undefined")if(typeof OffscreenCanvas!="undefined"&&typeof OffscreenCanvasRenderingContext2D!="undefined")Vs=new OffscreenCanvas(1,1).getContext("2d");else throw new Error("Cannot parse input in current context. Reason: OffscreenCanvas Context2D rendering is not supported.");else Vs=document.createElement("canvas").getContext("2d");Vs.canvas.width=l,Vs.canvas.height=u,Vs.drawImage(e,0,0,l,u),p=Vs.getImageData(0,0,l,u).data}let d;if(t===4)d=new Int32Array(p);else{let c=l*u;d=new Int32Array(c*t);for(let h=0;h<c;h++)for(let m=0;m<t;++m)d[h*t+m]=p[h*4+m]}return af(d,[u,l,t],"int32")}function BR(e){return e!=null&&e.data instanceof Uint8Array}function WR(){return typeof window!="undefined"&&typeof ImageBitmap!="undefined"&&window.hasOwnProperty("createImageBitmap")}function VR(e){return e!=null&&e.width!==0&&e.height!==0}function UR(e){return WR()&&!(e instanceof ImageBitmap)&&VR(e)&&!BR(e)}async function GR(e,t=3){let n=null;if(X().getBool("WRAP_TO_IMAGEBITMAP")&&UR(e)){let a;try{a=await createImageBitmap(e,{premultiplyAlpha:"none"})}catch(r){a=null}a!=null&&a.width===e.width&&a.height===e.height?n=a:n=e}else n=e;return uS(n,t)}async function HR(e,t){let n=F(e,"img","toPixels");if(!(e instanceof Fe)){let u=n;n=oe(u,"int32"),u.dispose()}if(n.rank!==2&&n.rank!==3)throw new Error(`toPixels only supports rank 2 or 3 tensors, got rank ${n.rank}.`);let[a,r]=n.shape.slice(0,2),s=n.rank===2?1:n.shape[2];if(s>4||s===2)throw new Error(`toPixels only supports depth of size 1, 3 or 4 but got ${s}`);if(n.dtype!=="float32"&&n.dtype!=="int32")throw new Error(`Unsupported type for toPixels: ${n.dtype}. Please use float32 or int32 tensors.`);let i=await n.data(),o=n.dtype==="float32"?255:1,l=new Uint8ClampedArray(r*a*4);for(let u=0;u<a*r;++u){let p=[0,0,0,255];for(let c=0;c<s;c++){let h=i[u*s+c];if(n.dtype==="float32"){if(h<0||h>1)throw new Error(`Tensor values for a float32 Tensor must be in the range [0 - 1] but encountered ${h}.`)}else if(n.dtype==="int32"&&(h<0||h>255))throw new Error(`Tensor values for a int32 Tensor must be in the range [0 - 255] but encountered ${h}.`);s===1?(p[0]=h*o,p[1]=h*o,p[2]=h*o):p[c]=h*o}let d=u*4;l[d+0]=Math.round(p[0]),l[d+1]=Math.round(p[1]),l[d+2]=Math.round(p[2]),l[d+3]=Math.round(p[3])}if(t!=null){t.width=r,t.height=a;let u=t.getContext("2d"),p=new ImageData(l,r,a);u.putImageData(p,0,0)}return n!==e&&n.dispose(),l}var jR=z({fromPixels_:uS}),Gx={};Me(Gx,{prepareAndValidate:()=>pS});function pS(e,t){let n=e.shape.length,a=t.shape.length;if(n<1)throw new Error(`tf.gatherND() expects the input to be rank 1 or higher, but the rank was ${n}.`);if(a<1)throw new Error(`tf.gatherND() expects the indices to be rank 1 or higher, but the rank was ${a}.`);if(t.dtype!=="int32")throw new Error(`tf.gatherND() expects the indices to be int32 type, but the dtype was ${t.dtype}.`);if(t.shape[a-1]>n)throw new Error(`index innermost dimension length must be <= tensor rank; saw: ${t.shape[a-1]} vs. ${n}`);if(bt(e.shape)===0)throw new Error(`Requested more than 0 entries, but input is empty. Input shape: ${e.shape}.`);let r=t.shape,s=r[r.length-1],i=1;for(let d=0;d<r.length-1;++d)i*=r[d];let o=e.shape,l=r.slice();l.pop();let u=1;for(let d=s;d<n;++d)u*=o[d],l.push(o[d]);let p=[...Ol(e.shape).map(d=>d/u),1].slice(0,s);return[l,i,u,p]}var Hx={};Me(Hx,{calculateShapes:()=>cS,validateInput:()=>qx,validateUpdateShape:()=>jx});function jx(e,t,n){let a=t.rank>1?t.shape[t.rank-1]:1,r=t.rank>1?t.rank-1:1,s=`Must have updates.shape = indices.shape[:batchDim] + shape[sliceDim:], got updates.shape: ${n.shape}, indices.shape: ${t.shape}, shape: ${e}, sliceDim: ${a}, and batchDim: ${r}.`;if(n.rank<r)throw new Error(s+` update.rank < ${r}. `);if(e.length<a+(n.rank-r))throw new Error(s+` Output shape length < ${a+(n.rank-r)}`);if(n.rank!==r+e.length-a)throw new Error(s+` update.rank != ${r+e.length-a}`);for(let i=0;i<r;++i)if(n.shape[i]!==t.shape[i])throw new Error(s+` updates.shape[${i}] (${n.shape[i]}) != indices.shape[${i}] (${t.shape[i]}).`);for(let i=0;i<n.rank-r;++i)if(n.shape[i+r]!==e[i+a])throw new Error(s+` updates.shape[${i+r}] (${n.shape[i+r]}) != shape[${i+r}] (${e[i+r]})`)}function qx(e,t,n){if(t.rank<1)throw new Error(`tf.scatterND() expects the indices to be rank 1 or higher, but the rank was ${t.rank}.`);if(e.rank<1)throw new Error(`tf.scatterND() expects the updates to be rank 1 or higher, but the rank was ${e.rank}.`);if(t.dtype!=="int32")throw new Error(`The dtype of 'indices' should be int32, but got dtype: ${t.dtype}`);if(n.length<1)throw new Error(`Output rank must be greater or equal to 1, but got shape: ${n}`);if(n.length===0){if(t.size===0)throw new Error(`Indices specified for empty output. indices shape: ${t.shape}`);if(e.size===0)throw new Error(`Updates specified for empty output. updates shape: ${e.shape}`)}jx(n,t,e)}function cS(e,t,n){let a=t.shape.length,r=a>1?t.shape[a-1]:1,s=n.length,i=1;for(let d=r;d<s;++d)i*=n[d];let o=r<1?1:r,l=bt(t.shape)/o,u=[...Ol(n.slice(0,r)),1],p=bt(n);return{sliceRank:r,numUpdates:l,sliceSize:i,strides:u,outputSize:p}}var Kt={};Me(Kt,{assertParamsValid:()=>KR,computeFlatOffset:()=>ZR,computeOutShape:()=>YR,getNormalizedAxes:()=>QR,isSliceContinous:()=>JR,maskToAxes:()=>XR,parseSliceParams:()=>vS,sliceInfo:()=>eM,startForAxis:()=>bS,startIndicesWithElidedDims:()=>fS,stopForAxis:()=>xS,stopIndicesWithElidedDims:()=>gS,stridesForAxis:()=>yS,stridesWithElidedDims:()=>dS});var Gb=-2,qR=-1;function KR(e,t,n){let a=e.shape.length;R(a===t.length,()=>`Error in slice${a}D: Length of begin ${t} must match the rank of the array (${a}).`),R(a===n.length,()=>`Error in slice${a}D: Length of size ${n} must match the rank of the array (${a}).`);for(let r=0;r<a;++r)R(t[r]+n[r]<=e.shape[r],()=>`Error in slice${a}D: begin[${r}] + size[${r}] (${t[r]+n[r]}) would overflow input.shape[${r}] (${e.shape[r]})`)}function XR(e){let t=[],n=0;for(;e>0;)e&1&&t.push(n),e/=2,n++;return t}function YR(e,t,n){let a=[];for(let r=0;r<e.length;r++)a[r]=Math.ceil((t[r]-e[r])/n[r]);return a}function dS(e,t,n,a){let r=[...e];for(let s=r.length;s<a.length;s++)r.push(1);for(let s=0;s<n;s++)s===0?r[t]=1:(r.splice(t,0,1),r.pop());return r}function hS(e,t,n){return n<=e?n:n-(t-1)}function mS(e,t){let n=[];for(let a=0;a<e;a++)n.push(t+a);return n}function QR(e,t,n,a,r,s,i,o,l){let u=e.length,p=new Array(u),d=new Array(u),c=new Array(u);if(t.length&&n>0){let h=t[0],m=n+1;p=fS(i,h,m,a,e),d=gS(o,h,m,r,e),c=dS(s,h,m,e)}else for(let h=0;h<u;h++)p[h]=bS(i,a,s,e,h,l),d[h]=xS(o,r,s,e,h,l),c[h]=yS(s,h,l);return{begin:p,end:d,strides:c}}function fS(e,t,n,a,r){let s=[...r],i=mS(n,t);for(let o=0;o<s.length;o++)if(i.indexOf(o)>-1)s[o]=0;else{let l=hS(t,n,o),u=a[l];e&1<<l&&(u=0),s[o]=u}return s}function gS(e,t,n,a,r){let s=[...r],i=mS(n,t);for(let o=0;o<s.length;o++)if(i.indexOf(o)>-1)s[o]=Number.MAX_SAFE_INTEGER;else{let l=hS(t,n,o),u=a[l];e&1<<l&&(u=Number.MAX_SAFE_INTEGER),s[o]=u}for(let o=0;o<s.length;o++){let l=r[o];s[o]<0&&(s[o]+=l),s[o]=Zp(0,s[o],r[o])}return s}function yS(e,t,n){let a=e[t];return(n&1<<t||a==null)&&(a=1),a}function bS(e,t,n,a,r,s){let i=t[r],o=n[r]||1;(e&1<<r||s&1<<r||i==null)&&(o>0?i=Number.MIN_SAFE_INTEGER:i=Number.MAX_SAFE_INTEGER);let l=a[r];return i<0&&(i+=l),i=Zp(0,i,l-1),i}function xS(e,t,n,a,r,s){let i=t[r],o=n[r]||1;(e&1<<r||s&1<<r||i==null)&&(o>0?i=Number.MAX_SAFE_INTEGER:i=Number.MIN_SAFE_INTEGER);let l=a[r];return i<0&&(i+=l),o>0?i=Zp(0,i,l):i=Zp(-1,i,l-1),i}function JR(e,t,n){let a=n.length;for(let r=0;r<n.length;r++)if(n[r]>1){a=r;break}for(let r=a+1;r<n.length;r++)if(t[r]>0||n[r]!==e[r])return!1;return!0}function ZR(e,t){let n=e.length>0?e[e.length-1]:1;for(let a=0;a<e.length-1;a++)n+=e[a]*t[a];return n}function vS(e,t,n){let a,r=e.shape.length;typeof t=="number"?a=[t,...new Array(r-1).fill(0)]:t.length<r?a=t.concat(new Array(r-t.length).fill(0)):a=t.slice(),a.forEach(i=>{R(i!==-1,()=>"slice() does not support negative begin indexing.")});let s;return n==null?s=new Array(r).fill(-1):typeof n=="number"?s=[n,...new Array(r-1).fill(-1)]:n.length<r?s=n.concat(new Array(r-n.length).fill(-1)):s=n,s=s.map((i,o)=>i>=0?i:(R(i===-1,()=>`Negative size values should be exactly -1 but got ${i} for the slice() size at index ${o}.`),e.shape[o]-a[o])),[a,s]}function eM(e,t,n,a,r,s,i,o,l){let u;if(a==null?(u=new Array(t.length),u.fill(1)):u=a,i!=null&&(i&i-1)!==0)throw new Error("Multiple ellipses in slice is not allowed.");let p=!1,d={dims:u.length,numAddAxisAfterEllipsis:0,begin:t.slice(),end:n.slice(),strides:u.slice(),beginMask:r,endMask:s,ellipsisMask:i,newAxisMask:o,shrinkAxisMask:l};for(let b=0;b<d.dims;b++)p&&(1<<b&o)!==0&&d.numAddAxisAfterEllipsis++,1<<b&i&&(p=!0);p||(d.ellipsisMask|=1<<d.dims,d.dims++);let c={dims:e.length,beginMask:0,endMask:0,beginValid:!1,endValid:!1};tM(d,c);let h=!0,m=!0,f=!0,g=[],y=[];for(let b=0;b<e.length;++b){if(c.strides[b]===0)throw Error(`strides[${b}] must be non-zero`);let x=!!(c.shrinkAxisMask&1<<b),v=e[b];if(v===-1){g.push(x?1:-1);continue}let k=[c.beginMask&1<<b,c.endMask&1<<b],T=[c.strides[b]>0?0:-1,c.strides[b]>0?v:v-1];if(x&&c.strides[b]<=0)throw Error("only stride 1 allowed on non-range indexing.");f=f&&c.strides[b]===1;let _=!!(c.beginMask&1<<b&&c.endMask&1<<b);if(c.beginValid&&c.endValid){if(x){let $=c.begin[b]<0?v+c.begin[b]:c.begin[b];if(c.begin[b]=$,c.end[b]=c.begin[b]+1,$<0||$>=v)throw Error(`slice index ${c.begin[b]} of dimension ${b} out of bounds.`)}else c.begin[b]=nk(c.begin[b],0,c.strides[b],v,k,T),c.end[b]=nk(c.end[b],1,c.strides[b],v,k,T);let M=c.strides[b]===1&&c.begin[b]===0&&c.end[b]===v;h=h&&M,m=m&&(b===0&&c.strides[b]===1||M)}else h=h&&c.strides[b]===1&&_,m=m&&(b===0&&c.strides[b]===1||_);let E,A=!1;if(c.beginValid&&c.endValid?(E=c.end[b]-c.begin[b],A=!0):x?(E=1,A=!0):_&&v>=0&&(c.strides[b]<0?E=-v:E=v,A=!0),A){let M;E===0||E<0!=c.strides[b]<0?M=0:M=Math.trunc(E/c.strides[b])+(E%c.strides[b]!==0?1:0),g.push(M)}else g.push(-1)}for(let b=0;b<c.finalShapeGatherIndices.length;++b){let x=c.finalShapeGatherIndices[b];x>=0?y.push(g[x]):x===Gb&&y.push(1)}return{finalShapeSparse:y.filter((b,x)=>c.finalShapeGatherIndices[x]!==Gb),finalShape:y,isIdentity:h,sliceDim0:m,isSimpleSlice:f,begin:c.begin,end:c.end,strides:c.strides}}function tM(e,t){t.beginMask=0,t.endMask=0,t.shrinkAxisMask=0;let n=0;t.beginValid=e.begin!=null,t.endValid=e.end!=null,t.begin=new Array(t.dims),t.end=new Array(t.dims),t.strides=new Array(t.dims),t.finalShapeGatherIndices=[],t.finalShapeGatherIndicesSparse=[],t.inputShapeGatherIndicesSparse=new Array(t.dims);for(let a=0;a<e.dims;a++)if(1<<a&e.ellipsisMask){let r=Math.min(t.dims-(e.dims-a)+1+e.numAddAxisAfterEllipsis,t.dims);for(;n<r;n++)t.begin[n]=0,t.end[n]=0,t.strides[n]=1,t.beginMask|=1<<n,t.endMask|=1<<n,t.finalShapeGatherIndices.push(n),t.finalShapeGatherIndicesSparse.push(-1),t.inputShapeGatherIndicesSparse[n]=a}else if(1<<a&e.newAxisMask)t.finalShapeGatherIndices.push(Gb),t.finalShapeGatherIndicesSparse.push(-1);else{if(n===t.begin.length)throw Error(`Index out of range using input dim ${n}; input has only ${t.dims} dims, ${t.begin.length}.`);e.begin!=null&&(t.begin[n]=e.begin[a]),e.end!=null&&(t.end[n]=e.end[a]),t.strides[n]=e.strides[a],e.beginMask&1<<a&&(t.beginMask|=1<<n),e.endMask&1<<a&&(t.endMask|=1<<n),e.shrinkAxisMask&1<<a?(t.finalShapeGatherIndices.push(qR),t.finalShapeGatherIndicesSparse.push(-1),t.shrinkAxisMask|=1<<n):(t.finalShapeGatherIndices.push(n),t.finalShapeGatherIndicesSparse.push(a)),t.inputShapeGatherIndicesSparse[n]=a,n++}}function nk(e,t,n,a,r,s){if(r[t])return n>0?s[t]:s[t+1&1];{let i=e<0?a+e:e;return i<s[0]?s[0]:i>s[1]?s[1]:i}}var se={};Me(se,{Serializable:()=>wS,SerializationMap:()=>qs,registerClass:()=>Is});var wS=class{getClassName(){return this.constructor.className}static fromConfig(e,t){return new e(t)}},qs=class{constructor(){this.classNameMap={}}static getMap(){return qs.instance==null&&(qs.instance=new qs),qs.instance}static register(e){qs.getMap().classNameMap[e.className]=[e,e.fromConfig]}};function Is(e){R(e.className!=null,()=>"Class being registered does not have the static className property defined."),R(typeof e.className=="string",()=>"className is required to be a string, but got type "+typeof e.className),R(e.className.length>0,()=>"Class being registered has an empty-string as its className, which is disallowed."),qs.register(e)}var kS={};Me(kS,{TEST_EPSILON_FLOAT16:()=>IS,encodeStrings:()=>SS,expectArrayBuffersEqual:()=>lM,expectArraysClose:()=>aM,expectArraysEqual:()=>sM,expectNumbersClose:()=>iM,expectPromiseToFail:()=>rM,expectValuesInRange:()=>oM,testEpsilon:()=>Kx});var nM=.001,IS=.1;function aM(e,t,n){return n==null&&(n=Kx()),Hb(e,t,(a,r)=>Xx(a,r,n))}function Kx(){return L.backend.floatPrecision()===32?nM:IS}function Hb(e,t,n){let a=!0;if((hn(e)||hn(t))&&(a=!1),hn(e)&&hn(t)&&(a=!0),a){let i=e.constructor.name,o=t.constructor.name;if(i!==o)throw new Error(`Arrays are of different type. Actual: ${i}. Expected: ${o}`)}if(Array.isArray(e)&&Array.isArray(t)){let i=pr(e),o=pr(t);if(!ys(i,o))throw new Error(`Arrays have different shapes. Actual: [${i}]. Expected: [${o}]`)}let r=hn(e)?e:ni(e),s=hn(t)?t:ni(t);if(r.length!==s.length)throw new Error(`Arrays have different lengths actual: ${r.length} vs expected: ${s.length}.
Actual: ${r}.
Expected: ${s}.`);for(let i=0;i<s.length;++i){let o=r[i],l=s[i];if(!n(o,l))throw new Error(`Arrays differ: actual[${i}] = ${o}, expected[${i}] = ${l}.
Actual: ${r}.
Expected: ${s}.`)}}function rM(e,t){e().then(()=>t.fail(),()=>t())}function sM(e,t){let n=typeof t=="string"||typeof t=="number"||typeof t=="boolean"?[t]:t;return ts(e)||ts(e[0])||ts(t)||ts(t[0])?Hb(e,n,(a,r)=>a==r):Hb(e,t,(a,r)=>Xx(a,r,0))}function iM(e,t,n){if(n==null&&(n=Kx()),!Xx(e,t,n))throw new Error(`Numbers differ: actual === ${e}, expected === ${t}`)}function Xx(e,t,n){return!isFinite(e)&&!isFinite(t)?!0:!(isNaN(e)||isNaN(t)||Math.abs(e-t)>n)}function oM(e,t,n){for(let a=0;a<e.length;a++)if(e[a]<t||e[a]>n)throw new Error(`Value out of range:${e[a]} low: ${t}, high: ${n}`)}function lM(e,t){let n=new Float32Array(e),a=new Float32Array(t);if(n.length!==a.length)throw new Error(`Expected ArrayBuffer to be of length ${a.length}, but it was ${n.length}`);for(let r=0;r<a.length;r++)if(n[r]!==a[r])throw new Error(`Expected ArrayBuffer value at ${r} to be ${a[r]} but got ${n[r]} instead`)}function SS(e){for(let t=0;t<e.length;t++){let n=e[t];Array.isArray(n)?SS(n):e[t]=Pc(n)}return e}var uM="3.18.0";function pM(e,t){let n=F(e,"a","add"),a=F(t,"b","add");[n,a]=At(n,a);let r={a:n,b:a};return L.runKernel(bs,r)}var J=z({add_:pM});function cM(e,t){let n=F(e,"a","floorDiv"),a=F(t,"b","floorDiv");[n,a]=At(n,a);let r={a:n,b:a};return L.runKernel(Oi,r)}var rf=z({floorDiv_:cM});function dM(e,t){let n=F(e,"a","div"),a=F(t,"b","div");if([n,a]=At(n,a),n.dtype==="int32"&&a.dtype==="int32")return rf(n,a);let r={a:n,b:a},s={};return L.runKernel(Di,r,s)}var fe=z({div_:dM});function hM(e,t){let n=F(e,"a","mul"),a=F(t,"b","mul");[n,a]=At(n,a);let r={a:n,b:a};return L.runKernel(Yi,r)}var B=z({mul_:hM});function mM(e){let t=F(e,"x","abs");if(t.dtype==="complex64"){let n={x:t};return L.runKernel(xc,n)}else{let n={x:t};return L.runKernel(Ll,n)}}var zt=z({abs_:mM});function fM(e){let t={x:F(e,"x","acos")};return L.runKernel(zl,t)}var Yx=z({acos_:fM});function gM(e){let t={x:F(e,"x","acosh")};return L.runKernel(Bl,t)}var Qx=z({acosh_:gM});function yM(e){R(Array.isArray(e),()=>"The argument passed to tf.addN() must be a list of tensors"),R(e.length>=1,()=>`Must pass at least one tensor to tf.addN(), but got ${e.length}`);let t=e.map((r,s)=>F(r,`tensors${s}`,"addN")),n=t[0];t.forEach(r=>{if(r.dtype!==n.dtype)throw new Error("All tensors passed to tf.addN() must have the same dtype")}),t.forEach(r=>{if(!ys(r.shape,n.shape))throw new Error("All tensors passed to tf.addN() must have the same shape")});let a=t;return L.runKernel(wi,a)}var NS=z({addN_:yM});function bM(e,t=null,n=!1){let a={x:F(e,"x","all","bool")},r={axis:t,keepDims:n};return L.runKernel(Wl,a,r)}var sf=z({all_:bM});function xM(e,t=null,n=!1){let a={x:F(e,"x","any","bool")},r={axis:t,keepDims:n};return L.runKernel(Vl,a,r)}var rc=z({any_:xM});function vM(e,t=0){let n={x:F(e,"x","argMax")},a={axis:t};return L.runKernel(ki,n,a)}var li=z({argMax_:vM});function wM(e,t=0){let n={x:F(e,"x","argMin")},a={axis:t};return L.runKernel(yc,n,a)}var Jx=z({argMin_:wM});function kM(e){let t={x:F(e,"x","asin")};return L.runKernel(Ul,t)}var Zx=z({asin_:kM});function IM(e){let t={x:F(e,"x","asinh")};return L.runKernel(Gl,t)}var ev=z({asinh_:IM});function SM(e){let t={x:F(e,"x","atan")};return L.runKernel(Hl,t)}var tv=z({atan_:SM});function NM(e,t){let n=F(e,"a","atan2"),a=F(t,"b","atan2");[n,a]=At(n,a);let r={a:n,b:a};return L.runKernel(ql,r)}var nv=z({atan2_:NM});function TM(e){let t={x:F(e,"x","atanh")};return L.runKernel(jl,t)}var av=z({atanh_:TM});function _M(e,t,n,a,r="NHWC",s){let i=e[3],o=[...t,i],l=CS(r);return Bc(e,o,n,s,a,null,null,l)}function TS(e,t,n,a,r,s,i="channelsLast"){let[o,l]=Jh(t),u;if(i==="channelsLast")u=[o,l,e[3],e[3]];else if(i==="channelsFirst")u=[o,l,e[1],e[1]];else throw new Error(`Unknown dataFormat ${i}`);return Bc(e,u,n,a,r,s,!1,i)}function CM(e,t,n,a,r,s,i="NDHWC"){let[o,l,u]=jb(t),p,d;if(i==="NDHWC")d="channelsLast",p=[o,l,u,e[4],e[4]];else if(i==="NCDHW")d="channelsFirst",p=[o,l,u,e[1],e[1]];else throw new Error(`Unknown dataFormat ${i}`);return _S(e,p,n,a,r,!1,d,s)}function Bc(e,t,n,a,r,s,i=!1,o="channelsLast"){let[l,u,p,d]=[-1,-1,-1,-1];if(o==="channelsLast")[l,u,p,d]=e;else if(o==="channelsFirst")[l,d,u,p]=e;else throw new Error(`Unknown dataFormat ${o}`);let[c,h,,m]=t,[f,g]=Jh(n),[y,b]=Jh(a),x=bl(c,y),v=bl(h,b),{padInfo:k,outHeight:T,outWidth:_}=AM(r,u,p,f,g,x,v,s,o),E=i?m*d:m,A;return o==="channelsFirst"?A=[l,E,T,_]:o==="channelsLast"&&(A=[l,T,_,E]),{batchSize:l,dataFormat:o,inHeight:u,inWidth:p,inChannels:d,outHeight:T,outWidth:_,outChannels:E,padInfo:k,strideHeight:f,strideWidth:g,filterHeight:c,filterWidth:h,effectiveFilterHeight:x,effectiveFilterWidth:v,dilationHeight:y,dilationWidth:b,inShape:e,outShape:A,filterShape:t}}function _S(e,t,n,a,r,s=!1,i="channelsLast",o){let[l,u,p,d,c]=[-1,-1,-1,-1,-1];if(i==="channelsLast")[l,u,p,d,c]=e;else if(i==="channelsFirst")[l,c,u,p,d]=e;else throw new Error(`Unknown dataFormat ${i}`);let[h,m,f,,g]=t,[y,b,x]=jb(n),[v,k,T]=jb(a),_=bl(h,v),E=bl(m,k),A=bl(f,T),{padInfo:M,outDepth:$,outHeight:S,outWidth:P}=$M(r,u,p,d,y,b,x,_,E,A,o),V=s?g*c:g,j;return i==="channelsFirst"?j=[l,V,$,S,P]:i==="channelsLast"&&(j=[l,$,S,P,V]),{batchSize:l,dataFormat:i,inDepth:u,inHeight:p,inWidth:d,inChannels:c,outDepth:$,outHeight:S,outWidth:P,outChannels:V,padInfo:M,strideDepth:y,strideHeight:b,strideWidth:x,filterDepth:h,filterHeight:m,filterWidth:f,effectiveFilterDepth:_,effectiveFilterHeight:E,effectiveFilterWidth:A,dilationDepth:v,dilationHeight:k,dilationWidth:T,inShape:e,outShape:j,filterShape:t}}function EM(e,t,n,a,r){a==null&&(a=rv(e,t,n));let s=e[0],i=e[1],o=Zs((s-t+2*a)/n+1,r),l=Zs((i-t+2*a)/n+1,r);return[o,l]}function FM(e,t,n,a,r,s){r==null&&(r=rv(e,t,a));let i=e[0],o=e[1],l=e[2],u=Zs((i-t+2*r)/a+1,s),p=Zs((o-t+2*r)/a+1,s),d=Zs((l-t+2*r)/a+1,s);return[u,p,d,n]}function rv(e,t,n,a=1){let r=bl(t,a);return Math.floor((e[0]*(n-1)-n+r)/2)}function Jh(e){return typeof e=="number"?[e,e,e]:e.length===2?[e[0],e[1],1]:e}function jb(e){return typeof e=="number"?[e,e,e]:e}function bl(e,t){return t<=1?e:e+(e-1)*(t-1)}function AM(e,t,n,a,r,s,i,o,l){let u,p,d;if(typeof e=="number"){u={top:e,bottom:e,left:e,right:e,type:e===0?"VALID":"NUMBER"};let c=EM([t,n],s,a,e,o);p=c[0],d=c[1]}else if(e==="same"){p=Math.ceil(t/a),d=Math.ceil(n/r);let c=Math.max(0,(p-1)*a+s-t),h=Math.max(0,(d-1)*r+i-n),m=Math.floor(c/2),f=c-m,g=Math.floor(h/2),y=h-g;u={top:m,bottom:f,left:g,right:y,type:"SAME"}}else if(e==="valid")u={top:0,bottom:0,left:0,right:0,type:"VALID"},p=Math.ceil((t-s+1)/a),d=Math.ceil((n-i+1)/r);else if(typeof e=="object"){let c=l==="channelsLast"?e[1][0]:e[2][0],h=l==="channelsLast"?e[1][1]:e[2][1],m=l==="channelsLast"?e[2][0]:e[3][0],f=l==="channelsLast"?e[2][1]:e[3][1];u={top:c,bottom:h,left:m,right:f,type:c===0&&h===0&&m===0&&f===0?"VALID":"EXPLICIT"},p=Zs((t-s+c+h)/a+1,o),d=Zs((n-i+m+f)/r+1,o)}else throw Error(`Unknown padding parameter: ${e}`);return{padInfo:u,outHeight:p,outWidth:d}}function $M(e,t,n,a,r,s,i,o,l,u,p){let d,c,h,m;if(typeof e=="number"){d={top:e,bottom:e,left:e,right:e,front:e,back:e,type:e===0?"VALID":"NUMBER"};let f=FM([t,n,a,1],o,1,r,e,p);c=f[0],h=f[1],m=f[2]}else if(e==="same"){c=Math.ceil(t/r),h=Math.ceil(n/s),m=Math.ceil(a/i);let f=(c-1)*r+o-t,g=(h-1)*s+l-n,y=(m-1)*i+u-a,b=Math.floor(f/2),x=f-b,v=Math.floor(g/2),k=g-v,T=Math.floor(y/2),_=y-T;d={top:v,bottom:k,left:T,right:_,front:b,back:x,type:"SAME"}}else if(e==="valid")d={top:0,bottom:0,left:0,right:0,front:0,back:0,type:"VALID"},c=Math.ceil((t-o+1)/r),h=Math.ceil((n-l+1)/s),m=Math.ceil((a-u+1)/i);else throw Error(`Unknown padding parameter: ${e}`);return{padInfo:d,outDepth:c,outHeight:h,outWidth:m}}function Zs(e,t){if(!t)return Math.trunc(e);switch(t){case"round":return Math.round(e);case"ceil":return Math.ceil(e);case"floor":return Math.floor(e);default:throw new Error(`Unknown roundingMode ${t}`)}}function us(e){let[t,n,a]=Jh(e);return t===1&&n===1&&a===1}function mr(e,t){return us(e)||us(t)}function CS(e){if(e==="NHWC")return"channelsLast";if(e==="NCHW")return"channelsFirst";throw new Error(`Unknown dataFormat ${e}`)}function _n(e,t,n){if(n!=null){if(typeof t=="string")throw Error(`Error in ${e}: pad must be an integer when using dimRoundingMode ${n} but got pad ${t}.`);if(typeof t=="number")R(wl(t),()=>`Error in ${e}: pad must be an integer when using dimRoundingMode ${n} but got pad ${t}.`);else if(typeof t=="object")t.forEach(a=>{a.forEach(r=>{R(wl(r),()=>`Error in ${e}: pad must be an integer when using dimRoundingMode ${n} but got pad ${r}.`)})});else throw Error(`Error in ${e}: Unknown padding parameter: ${t}`)}}function DM(e,t){let n={x:F(e,"x","reshape","string_or_numeric")},a={shape:t};return L.runKernel(Iu,n,a)}var W=z({reshape_:DM});function RM(e,t,n,a,r){let s=F(e,"x","avgPool","float32"),i=1;R(mr(n,i),()=>`Error in avgPool: Either strides or dilations must be 1. Got strides ${n} and dilations '${i}'`);let o=s,l=!1;s.rank===3&&(l=!0,o=W(s,[1,s.shape[0],s.shape[1],s.shape[2]])),R(o.rank===4,()=>`Error in avgPool: x must be rank 4 but got rank ${o.rank}.`),_n("avgPool",a,r);let u={x:o},p={filterSize:t,strides:n,pad:a,dimRoundingMode:r},d=L.runKernel(Ii,u,p);return d=oe(d,s.dtype),l?W(d,[d.shape[1],d.shape[2],d.shape[3]]):d}var ga=z({avgPool_:RM});function MM(e,t,n,a,r,s="NDHWC"){let i=F(e,"x","avgPool3d","float32"),o=i,l=!1;i.rank===4&&(l=!0,o=W(i,[1,i.shape[0],i.shape[1],i.shape[2],i.shape[3]])),R(o.rank===5,()=>`Error in avgPool3d: x must be rank 5 but got rank ${o.rank}.`),R(s==="NDHWC",()=>`Error in avgPool3d: Only NDHWC is currently supported, but got dataFormat of ${s}`),_n("avgPool3d",a,r);let u={x:o},p={filterSize:t,strides:n,pad:a,dimRoundingMode:r,dataFormat:s},d=L.runKernel(bc,u,p);return d=oe(d,o.dtype),l?W(d,[d.shape[1],d.shape[2],d.shape[3],d.shape[4]]):d}var sv=z({avgPool3d_:MM});function PM(e,t=0){R(e.length>=1,()=>"Pass at least one tensor to concat");let n=ac(e,"tensors","concat","string_or_numeric");if(n[0].dtype==="complex64"&&n.forEach(s=>{if(s.dtype!=="complex64")throw new Error(`Cannot concatenate complex64 tensors with a tensor
with dtype ${s.dtype}. `)}),n.length===1)return Er(n[0]);let a=n,r={axis:t};return L.runKernel(Xl,a,r)}var Ze=z({concat_:PM});function OM(e){let t={x:F(e,"x","sigmoid","float32")};return L.runKernel(uo,t)}var ma=z({sigmoid_:OM});function LM(e,t,n){let a=F(e,"x","slice","string_or_numeric");if(a.rank===0)throw new Error("Slicing scalar is not possible");let r={x:a},s={begin:t,size:n};return L.runKernel(_u,r,s)}var He=z({slice_:LM});function zM(e){let t={x:F(e,"x","tanh","float32")};return L.runKernel(yo,t)}var ui=z({tanh_:zM});function BM(e,t,n,a,r,s){let i=F(e,"forgetBias","basicLSTMCell"),o=F(t,"lstmKernel","basicLSTMCell"),l=F(n,"lstmBias","basicLSTMCell"),u=F(a,"data","basicLSTMCell"),p=F(r,"c","basicLSTMCell"),d=F(s,"h","basicLSTMCell"),c=Ze([u,d],1),h=De(c,o),m=J(h,l),f=m.shape[0],g=m.shape[1]/4,y=[f,g],b=He(m,[0,0],y),x=He(m,[0,g],y),v=He(m,[0,g*2],y),k=He(m,[0,g*3],y),T=J(B(ma(b),ui(x)),B(p,ma(J(i,v)))),_=B(ui(T),ma(k));return[T,_]}var WM=z({basicLSTMCell_:BM});function VM(e,t,n){let a=F(e,"x","batchToSpaceND"),r=t.reduce((o,l)=>o*l);R(a.rank>=1+t.length,()=>`input rank is ${a.rank} but should be > than blockShape.length ${t.length}`),R(n.length===t.length,()=>`crops.length is ${n.length} but should be equal to blockShape.length ${t.length}`),R(a.shape[0]%r===0,()=>`input tensor batch is ${a.shape[0]} but is not divisible by the product of the elements of blockShape ${t.join(" * ")} === ${r}`);let s={x:a},i={blockShape:t,crops:n};return L.runKernel(Kl,s,i)}var Wc=z({batchToSpaceND_:VM});function UM(e){let t;return e.rank===0||e.rank===1?t=W(e,[1,1,1,e.size]):e.rank===2?t=W(e,[1,1,e.shape[0],e.shape[1]]):e.rank===3?t=W(e,[1,e.shape[0],e.shape[1],e.shape[2]]):t=e,t}function GM(e,t,n,a,r,s){s==null&&(s=.001);let i=F(e,"x","batchNorm"),o=F(t,"mean","batchNorm"),l=F(n,"variance","batchNorm"),u;r!=null&&(u=F(r,"scale","batchNorm"));let p;a!=null&&(p=F(a,"offset","batchNorm")),R(o.rank===l.rank,()=>"Batch normalization gradient requires mean and variance to have equal ranks."),R(p==null||o.rank===p.rank,()=>"Batch normalization gradient requires mean and offset to have equal ranks."),R(u==null||o.rank===u.rank,()=>"Batch normalization gradient requires mean and scale to have equal ranks.");let d={x:UM(i),scale:u,offset:p,mean:o,variance:l},c={varianceEpsilon:s},h=L.runKernel(Li,d,c);return W(h,i.shape)}var $r=z({batchNorm_:GM});function HM(e,t,n,a,r,s){let i=F(e,"x","batchNorm"),o=F(t,"mean","batchNorm"),l=F(n,"variance","batchNorm"),u;r!=null&&(u=F(r,"scale","batchNorm"));let p;return a!=null&&(p=F(a,"offset","batchNorm")),R(i.rank===2,()=>`Error in batchNorm2D: x must be rank 2 but got rank ${i.rank}.`),R(o.rank===2||o.rank===1,()=>`Error in batchNorm2D: mean must be rank 2 or rank 1 but got rank ${o.rank}.`),R(l.rank===2||l.rank===1,()=>`Error in batchNorm2D: variance must be rank 2 or rank 1 but got rank ${l.rank}.`),u!=null&&R(u.rank===2||u.rank===1,()=>`Error in batchNorm2D: scale must be rank 2 or rank 1 but got rank ${u.rank}.`),p!=null&&R(p.rank===2||p.rank===1,()=>`Error in batchNorm2D: offset must be rank 2 or rank 1 but got rank ${p.rank}.`),$r(i,o,l,p,u,s)}var ES=z({batchNorm2d_:HM});function jM(e,t,n,a,r,s){let i=F(e,"x","batchNorm"),o=F(t,"mean","batchNorm"),l=F(n,"variance","batchNorm"),u;r!=null&&(u=F(r,"scale","batchNorm"));let p;return a!=null&&(p=F(a,"offset","batchNorm")),R(i.rank===3,()=>`Error in batchNorm3D: x must be rank 3 but got rank ${i.rank}.`),R(o.rank===3||o.rank===1,()=>`Error in batchNorm3D: mean must be rank 3 or rank 1 but got rank ${o.rank}.`),R(l.rank===3||l.rank===1,()=>`Error in batchNorm3D: variance must be rank 3 or rank 1 but got rank ${l.rank}.`),u!=null&&R(u.rank===3||u.rank===1,()=>`Error in batchNorm3D: scale must be rank 3 or rank 1 but got rank ${u.rank}.`),p!=null&&R(p.rank===3||p.rank===1,()=>`Error in batchNorm3D: offset must be rank 3 or rank 1 but got rank ${p.rank}.`),$r(i,o,l,p,u,s)}var FS=z({batchNorm3d_:jM});function qM(e,t,n,a,r,s){let i=F(e,"x","batchNorm"),o=F(t,"mean","batchNorm"),l=F(n,"variance","batchNorm"),u;r!=null&&(u=F(r,"scale","batchNorm"));let p;return a!=null&&(p=F(a,"offset","batchNorm")),R(i.rank===4,()=>`Error in batchNorm4D: x must be rank 4 but got rank ${i.rank}.`),R(o.rank===4||o.rank===1,()=>`Error in batchNorm4D: mean must be rank 4 or rank 1 but got rank ${o.rank}.`),R(l.rank===4||l.rank===1,()=>`Error in batchNorm4D: variance must be rank 4 or rank 1 but got rank ${l.rank}.`),u!=null&&R(u.rank===4||u.rank===1,()=>`Error in batchNorm4D: scale must be rank 4 or rank 1 but got rank ${u.rank}.`),p!=null&&R(p.rank===4||p.rank===1,()=>`Error in batchNorm4D: offset must be rank 4 or rank 1 but got rank ${p.rank}.`),$r(i,o,l,p,u,s)}var AS=z({batchNorm4d_:qM});function KM(e,t,n){let a=F(e,"x","bincount"),r=F(t,"weights","bincount");R(a.dtype==="int32",()=>`Error in bincount: input dtype must be int32, but got ${a.dtype}`),R(n>=0,()=>`size must be non-negative, but got ${n}.`),R(r.size===a.size||r.size===0,()=>`Error in bincount: weights must have the same size as input or0-length, but got input shape: ${a.shape}, weights shape: ${r.shape}.`);let s={x:a,weights:r},i={size:n};return L.runKernel(Sm,s,i)}var iv=z({bincount_:KM});function XM(e,t){let n=F(e,"s0","broadcastArgs","int32"),a=F(t,"s1","broadcastArgs","int32");if(n.rank!==1)throw new Error(`broadcastArgs(): first input must be a vector (rank=1). Has rank ${n.rank}`);if(a.rank!==1)throw new Error(`broadcastArgs(): second input must be a vector (rank=1). Has rank ${a.rank}`);let r={s0:n,s1:a};return L.runKernel(Nm,r)}var $S=z({broadcastArgs_:XM});function YM(e,t){let n=F(e,"broadcastTo","x"),a=n.shape;if(t.some(l=>!(l>0)||l%1!==0))throw new Error(`broadcastTo(): Invalid broadcast shape [${t}].`);if(t.length<n.rank)throw new Error(`broadcastTo(): shape.length=${t.length} < input.rank=${n.rank}.`);if(t.length>n.rank){let l=n.shape.slice();for(;l.length<t.length;)l.unshift(1);n=W(n,l)}let r=n.shape,s=Array.from(t);for(let l=t.length-1;l>=0;l--)if(r[l]===t[l])s[l]=1;else if(n.shape[l]!==1)throw new Error(`broadcastTo(): [${a}] cannot be broadcast to [${t}].`);if(s.map((l,u)=>l>1?u:-1).filter(l=>l>=0).length===0)return Er(n);let i={x:n},o={reps:s};return L.runKernel(vs,i,o)}var xl=z({broadcastTo_:YM});function QM(e){let t={x:F(e,"x","ceil","float32")};return L.runKernel(Ti,t)}var ov=z({ceil_:QM});function JM(e,t,n){let a=F(e,"x","clipByValue");R(t<=n,()=>`Error in clip: min (${t}) must be less than or equal to max (${n}).`);let r={x:a},s={clipValueMin:t,clipValueMax:n};return L.runKernel(xs,r,s)}var an=z({clipByValue_:JM});function ZM(e){return Ze(e,0)}var DS=z({concat1d_:ZM});function eP(e,t){return Ze(e,t)}var RS=z({concat2d_:eP});function tP(e,t){return Ze(e,t)}var MS=z({concat3d_:tP});function nP(e,t){return Ze(e,t)}var PS=z({concat4d_:nP});function aP(e,t,n,a,r="NHWC",s=[1,1],i){let o=F(e,"x","conv2d","float32"),l=F(t,"filter","conv2d","float32"),u=o,p=!1;o.rank===3&&(p=!0,u=W(o,[1,o.shape[0],o.shape[1],o.shape[2]])),R(u.rank===4,()=>`Error in conv2d: input must be rank 4, but got rank ${u.rank}.`),R(l.rank===4,()=>`Error in conv2d: filter must be rank 4, but got rank ${l.rank}.`),_n("conv2d",a,i);let d=r==="NHWC"?u.shape[3]:u.shape[1];R(d===l.shape[2],()=>`Error in conv2d: depth of input (${d}) must match input depth for filter ${l.shape[2]}.`),R(mr(n,s),()=>`Error in conv2D: Either strides or dilations must be 1. Got strides ${n} and dilations '${s}'`);let c={x:u,filter:l},h={strides:n,pad:a,dataFormat:r,dilations:s,dimRoundingMode:i},m=L.runKernel(_i,c,h);return p?W(m,[m.shape[1],m.shape[2],m.shape[3]]):m}var Rt=z({conv2d_:aP});function rP(e,t,n,a,r="NWC",s=1,i){let o=F(e,"x","conv1d"),l=F(t,"filter","conv1d"),u=o,p=!1;o.rank===2&&(p=!0,u=W(o,[1,o.shape[0],o.shape[1]])),R(u.rank===3,()=>`Error in conv1d: input must be rank 3, but got rank ${u.rank}.`),R(l.rank===3,()=>`Error in conv1d: filter must be rank 3, but got rank ${l.rank}.`),_n("conv1d",a,i),R(u.shape[2]===l.shape[1],()=>`Error in conv1d: depth of input (${u.shape[2]}) must match input depth for filter ${l.shape[1]}.`),R(mr(n,s),()=>`Error in conv1D: Either stride or dilation must be 1. Got stride ${n} and dilation '${s}'`),R(r==="NWC",()=>`Error in conv1d: got dataFormat of ${r} but only NWC is currently supported.`);let d=W(l,[1,l.shape[0],l.shape[1],l.shape[2]]),c=W(u,[u.shape[0],1,u.shape[1],u.shape[2]]),h=Rt(c,d,[1,n],a,"NHWC",[1,s],i);return p?W(h,[h.shape[2],h.shape[3]]):W(h,[h.shape[0],h.shape[2],h.shape[3]])}var of=z({conv1d_:rP});function sP(e,t,n,a,r,s="NHWC",i){R(e.length===t.rank,()=>`Length of inShape (${e.length}) and rank of dy (${t.rank}) must match`);let o=e,l=t,u=!1;t.rank===3&&(u=!0,l=W(t,[1,t.shape[0],t.shape[1],t.shape[2]]),o=[1,e[0],e[1],e[2]]),R(o.length===4,()=>`Error in conv2dDerInput: inShape must be length 4, but got length ${o.length}.`),R(l.rank===4,()=>`Error in conv2dDerInput: dy must be rank 4, but got rank ${l.rank}`),R(n.rank===4,()=>`Error in conv2dDerInput: filter must be rank 4, but got rank ${n.rank}`);let p=s==="NHWC"?o[3]:o[1],d=s==="NHWC"?l.shape[3]:l.shape[1];R(p===n.shape[2],()=>`Error in conv2dDerInput: depth of input (${p}) must match input depth for filter ${n.shape[2]}.`),R(d===n.shape[3],()=>`Error in conv2dDerInput: depth of output (${d}) must match output depth for filter ${n.shape[3]}.`),_n("conv2dDerInput",r,i);let c={dy:l,filter:n},h={strides:a,pad:r,dataFormat:s,dimRoundingMode:i,inputShape:o},m=L.runKernel(Ci,c,h);return u?W(m,[m.shape[1],m.shape[2],m.shape[3]]):m}var lv=z({conv2DBackpropInput_:sP});function iP(e,t,n,a,r,s){let i=F(e,"x","conv2dTranspose"),o=F(t,"filter","conv2dTranspose");return lv(n,i,o,a,r,"NHWC",s)}var lf=z({conv2dTranspose_:iP});function oP(e,t,n,a,r="NDHWC",s=[1,1,1]){let i=F(e,"x","conv3d"),o=F(t,"filter","conv3d"),l=i,u=!1;i.rank===4&&(u=!0,l=W(i,[1,i.shape[0],i.shape[1],i.shape[2],i.shape[3]])),R(l.rank===5,()=>`Error in conv3d: input must be rank 5, but got rank ${l.rank}.`),R(o.rank===5,()=>`Error in conv3d: filter must be rank 5, but got rank ${o.rank}.`),R(l.shape[4]===o.shape[3],()=>`Error in conv3d: depth of input (${l.shape[4]}) must match input depth for filter ${o.shape[3]}.`),R(mr(n,s),()=>`Error in conv3D: Either strides or dilations must be 1. Got strides ${n} and dilations '${s}'`),R(r==="NDHWC",()=>`Error in conv3d: got dataFormat of ${r} but only NDHWC is currently supported.`);let p={x:l,filter:o},d={strides:n,pad:a,dataFormat:r,dilations:s},c=L.runKernel(vc,p,d);return u?W(c,[c.shape[1],c.shape[2],c.shape[3],c.shape[4]]):c}var uv=z({conv3d_:oP});function lP(e,t,n,a,r){R(e.length===t.rank,()=>`Length of inShape (${e.length}) and rank of dy (${t.rank}) must match`);let s=e,i=t,o=!1;t.rank===4&&(o=!0,i=W(t,[1,t.shape[0],t.shape[1],t.shape[2],t.shape[3]]),s=[1,e[0],e[1],e[2],e[3]]);let l=s[4],u=i.shape[4];R(s.length===5,()=>`Error in conv3dDerInput: inShape must be length 5, but got length ${s.length}.`),R(i.rank===5,()=>`Error in conv3dDerInput: dy must be rank 5, but got rank ${i.rank}`),R(n.rank===5,()=>`Error in conv3dDerInput: filter must be rank 5, but got rank ${n.rank}`),R(l===n.shape[3],()=>`Error in conv3dDerInput: depth of input (${l}) must match input depth for filter ${n.shape[3]}.`),R(u===n.shape[4],()=>`Error in conv3dDerInput: depth of output (${u}) must match output depth for filter ${n.shape[4]}.`);let p={dy:i,filter:n},d={pad:r,strides:a,inputShape:s},c=L.runKernel(Em,p,d);return o?W(c,[c.shape[1],c.shape[2],c.shape[3],c.shape[4]]):c}var OS=z({conv3DBackpropInput_:lP});function uP(e,t,n,a,r){let s=F(e,"x","conv3dTranspose"),i=F(t,"filter","conv3dTranspose");return OS(n,s,i,a,r)}var LS=z({conv3dTranspose_:uP});function pP(e){let t={x:F(e,"x","cos","float32")};return L.runKernel(Ei,t)}var Vc=z({cos_:pP});function cP(e){let t={x:F(e,"x","cosh","float32")};return L.runKernel(Fi,t)}var uf=z({cosh_:cP});function dP(e,t=0,n=!1,a=!1){let r={x:F(e,"x","cumprod")},s={axis:t,exclusive:n,reverse:a};return L.runKernel(Yl,r,s)}var sc=z({cumprod_:dP});function hP(e,t=0,n=!1,a=!1){let r={x:F(e,"x","cumsum")},s={axis:t,exclusive:n,reverse:a};return L.runKernel(Ai,r,s)}var pf=z({cumsum_:hP});function mP(e,t,n,a=!1){let r=F(e,"x","denseBincount"),s=F(t,"weights","denseBincount");R(r.dtype==="int32",()=>`Error in denseBincount: input dtype must be int32, but got ${r.dtype}`),R(r.rank<=2,()=>`Error in denseBincount: input must be at most rank 2, but got rank ${r.rank}.`),R(n>=0,()=>`size must be non-negative, but got ${n}.`),R(s.size===r.size||s.size===0,()=>`Error in denseBincount: weights must have the same shape as x or 0-length, but got x shape: ${r.shape}, weights shape: ${s.shape}.`);let i={x:r,weights:s},o={size:n,binaryOutput:a};return L.runKernel(Fm,i,o)}var zS=z({denseBincount_:mP});function fP(e,t,n="NHWC"){let a=F(e,"x","depthToSpace","float32"),r=n==="NHWC"?a.shape[1]:a.shape[2],s=n==="NHWC"?a.shape[2]:a.shape[3],i=n==="NHWC"?a.shape[3]:a.shape[1];R(t>1,()=>`blockSize should be > 1 for depthToSpace, but was: ${t}`),R(r*t>=0,()=>`Negative dimension size caused by overflow when multiplying
${r} and ${t} for depthToSpace with input shape
${a.shape}`),R(s*t>=0,()=>`Negative dimension size caused by overflow when multiplying
${s} and ${t} for depthToSpace with input shape
${a.shape}`),R(i%(t*t)===0,()=>`Dimension size must be evenly divisible by ${t*t} but is ${i} for depthToSpace with input shape ${a.shape}`);let o={x:a},l={blockSize:t,dataFormat:n};return L.runKernel(Jl,o,l)}var pv=z({depthToSpace_:fP});function gP(e,t,n,a,r="NHWC",s=[1,1],i){let o=F(e,"x","depthwiseConv2d","float32"),l=F(t,"filter","depthwiseConv2d","float32"),u=o,p=!1;o.rank===3&&(p=!0,u=W(o,[1,o.shape[0],o.shape[1],o.shape[2]])),R(u.rank===4,()=>`Error in depthwiseConv2d: input must be rank 4, but got rank ${u.rank}.`),R(l.rank===4,()=>`Error in depthwiseConv2d: filter must be rank 4, but got rank ${l.rank}.`),R(u.shape[3]===l.shape[2],()=>`Error in depthwiseConv2d: number of input channels (${u.shape[3]}) must match the inChannels dimension in filter ${l.shape[2]}.`),_n("depthwiseConv2d",a,i);let d={x:u,filter:l},c={strides:n,pad:a,dataFormat:r,dilations:s,dimRoundingMode:i},h=L.runKernel($i,d,c);return p?W(h,[h.shape[1],h.shape[2],h.shape[3]]):h}var Ss=z({depthwiseConv2d_:gP});function yP(e){let t={x:F(e,"x","diag")};return L.runKernel(Dm,t)}var bP=z({diag_:yP});function xP(e,t,n,a,r=[1,1],s="NHWC"){let i=F(e,"x","dilation2d"),o=F(t,"filter","dilation2d");R(i.rank===3||i.rank===4,()=>`Error in dilation2d: input must be rank 3 or 4, but got rank ${i.rank}.`),R(o.rank===3,()=>`Error in dilation2d: filter must be rank 3, but got rank ${o.rank}.`),R(s==="NHWC",()=>`Error in dilation2d: Only NHWC is currently supported, but got dataFormat of ${s}`);let l=i,u=!1;i.rank===3&&(l=W(i,[1,i.shape[0],i.shape[1],i.shape[2]]),u=!0);let p={x:l,filter:o},d={strides:n,pad:a,dilations:r},c=L.runKernel(wc,p,d);return u?W(c,[c.shape[1],c.shape[2],c.shape[3]]):c}var cv=z({dilation2d_:xP});function vP(e,t){let n=F(e,"a","equal","string_or_numeric"),a=F(t,"b","equal","string_or_numeric");[n,a]=At(n,a),pt(n.shape,a.shape);let r={a:n,b:a};return L.runKernel(eu,r)}var Zn=z({equal_:vP});function wP(e,t,n){let a=F(t,"a","where"),r=F(n,"b","where"),s=F(e,"condition","where","bool"),i=pt(pt(s.shape,a.shape),r.shape),o=xl(s,i),l=xl(a,i),u=xl(r,i),p={condition:o,t:l,e:u};return L.runKernel(Nu,p)}var fn=z({where_:wP});function kP(e){let t={x:F(e,"x","zerosLike")};return L.runKernel(Lu,t)}var Ke=z({zerosLike_:kP});function IP(e,t){let n=F(e,"a","div"),a=F(t,"b","div");[n,a]=At(n,a);let r=fe(n,a),s=Ke(r),i=Zn(a,s);return fn(i,s,r)}var dv=z({divNoNan_:IP});function SP(e,t){let n=F(e,"t1","dot"),a=F(t,"t2","dot");R((n.rank===1||n.rank===2)&&(a.rank===1||a.rank===2),()=>`Error in dot: inputs must all be rank 1 or 2, but got ranks ${n.rank} and ${a.rank}.`);let r=n.rank===1?n.size:n.shape[1],s=a.rank===1?a.size:a.shape[0];if(R(r===s,()=>`Error in dot: inner dimensions of inputs must match, but got ${r} and ${s}.`),n.rank===1&&a.rank===1){let i=W(n,[1,-1]),o=W(a,[-1,1]),l=De(i,o);return W(l,[])}else if(n.rank===1&&a.rank===2){let i=W(n,[1,-1]),o=W(a,[a.shape[0],a.shape[1]]),l=De(i,o);return W(l,[l.size])}else if(n.rank===2&&a.rank===1){let i=W(a,[-1,1]),o=De(n,i);return W(o,[o.size])}else{let i=W(a,[a.shape[0],a.shape[1]]);return De(n,i)}}var BS=z({dot_:SP});function NP(e,...t){let n=t.map((r,s)=>F(r,`tensors${s}`,"einsum")),a={equation:e};return L.runKernel(Rm,n,a)}var WS=z({einsum_:NP});function TP(e){let t={x:F(e,"x","elu","float32")};return L.runKernel(Ri,t)}var Wu=z({elu_:TP});function _P(e){let t=F(e,"x","erf");R(t.dtype==="int32"||t.dtype==="float32",()=>"Input dtype must be `int32` or `float32`."),t.dtype==="int32"&&(t=oe(t,"float32"));let n={x:t};return L.runKernel(Zl,n)}var hv=z({erf_:_P});function mv(e,t){for(let n=0;n<e.length;++n)if(e[e.length-n-1]!==t-1-n)return!1;return!0}function VS(e,t,n){let a=e.length+t.length,r=[],s=0,i=0;for(let o=0;o<a;o++)n.indexOf(o)===-1?r.push(e[s++]):r.push(t[i++]);return r}function US(e,t){let n=[],a=e.length;for(let s=0;s<a;s++)t.indexOf(s)===-1&&n.push(e[s]);let r=t.map(s=>e[s]);return[n,r]}function pi(e,t){let n=t.map(a=>1);return VS(e,n,t)}function CP(e,t,n){R(mv(t,n),()=>`${e} supports only inner-most axes for now. Got axes ${t} and rank-${n} input.`)}function GS(e,t){if(mv(e,t))return null;let n=[];for(let a=0;a<t;++a)e.indexOf(a)===-1&&n.push(a);return e.forEach(a=>n.push(a)),n}function fv(e){return e.map((t,n)=>[n,t]).sort((t,n)=>t[1]-n[1]).map(t=>t[0])}function EP(e,t){let n=[];for(let a=t-e;a<t;++a)n.push(a);return n}function FP(e,t=null,n=!1){let a={x:F(e,"x","max")},r={reductionIndices:t,keepDims:n};return L.runKernel(Ui,a,r)}var Ta=z({max_:FP});function AP(e,t=null,n=!1){let a={x:F(e,"x","min")},r={axis:t,keepDims:n};return L.runKernel(qi,a,r)}var ic=z({min_:AP});function $P(e,t){let n=F(e,"base","pow"),a=F(t,"exp","pow");[n,a]=At(n,a);let r={a:n,b:a};return L.runKernel(Zi,r)}var Dr=z({pow_:$P});function we(e,t){if((hn(e)&&t!=="string"||Array.isArray(e))&&t!=="complex64")throw new Error("Error creating a new Scalar: value must be a primitive (number|boolean|string)");if(t==="string"&&hn(e)&&!(e instanceof Uint8Array))throw new Error("When making a scalar from encoded string, the value must be `Uint8Array`.");return ks(e,[],[],t)}function DP(e){let t={x:F(e,"x","sqrt","float32")};return L.runKernel(po,t)}var pn=z({sqrt_:DP});function RP(e){let t=F(e,"x","square"),n={};return L.runKernel("Square",{x:t},n)}var ut=z({square_:RP});function MP(e,t=null,n=!1){let a=F(e,"x","sum");a.dtype==="bool"&&(a=oe(a,"int32"));let r={x:a},s={axis:t,keepDims:n};return L.runKernel(co,r,s)}var be=z({sum_:MP});function PP(e,t="euclidean",n=null,a=!1){e=F(e,"x","norm");let r=HS(e,t,n),s=r.shape;if(a){let i=Ea(n,e.shape);s=pi(r.shape,i)}return W(r,s)}function HS(e,t,n=null){if(e.rank===0)return zt(e);if(e.rank!==1&&n===null)return HS(W(e,[-1]),t,n);if(e.rank===1||typeof n=="number"||Array.isArray(n)&&n.length===1){if(t===1)return be(zt(e),n);if(t===1/0)return Ta(zt(e),n);if(t===-1/0)return ic(zt(e),n);if(t==="euclidean"||t===2)return pn(be(Dr(zt(e),we(2,"int32")),n));throw new Error(`Error in norm: invalid ord value: ${t}`)}if(Array.isArray(n)&&n.length===2){if(t===1)return Ta(be(zt(e),n[0]),n[1]-1);if(t===1/0)return Ta(be(zt(e),n[1]),n[0]);if(t===-1/0)return ic(be(zt(e),n[1]),n[0]);if(t==="fro"||t==="euclidean")return pn(be(ut(e),n));throw new Error(`Error in norm: invalid ord value: ${t}`)}throw new Error(`Error in norm: invalid axis: ${n}`)}var Uc=z({norm_:PP});function OP(e,t=null,n=!1){return Uc(e,"euclidean",t,n)}var gv=z({euclideanNorm_:OP});function LP(e){let t={x:F(e,"x","exp")};return L.runKernel(Mi,t)}var gn=z({exp_:LP});function zP(e,t=0){let n=F(e,"x","expandDims","string_or_numeric");R(t<=n.rank,()=>"Axis must be <= rank of the tensor");let a={input:n},r={dim:t};return L.runKernel(tu,a,r)}var mn=z({expandDims_:zP});function BP(e){let t={x:F(e,"x","expm1")};return L.runKernel(nu,t)}var yv=z({expm1_:BP});function WP(e,t){let n=F(e,"x","tile","string_or_numeric");R(n.rank===t.length,()=>`Error in transpose: rank of input ${n.rank} must match length of reps ${t}.`);let a={x:n},r={reps:t};return L.runKernel(vs,a,r)}var On=z({tile_:WP});function VP(e,t,n,a="float32"){t==null&&(t=e);let r=Ve([e,t],a),s=e<=t?e:t;for(let o=0;o<s;++o)r.set(1,o,o);let i=W(r.toTensor(),[e,t]);if(n==null)return i;if(n.length===1)return On(mn(i,0),[n[0],1,1]);if(n.length===2)return On(mn(mn(i,0),0),[n[0],n[1],1,1]);if(n.length===3)return On(mn(mn(mn(i,0),0),0),[n[0],n[1],n[2],1,1]);throw new Error(`eye() currently supports only 1D and 2D batchShapes, but received ${n.length}D.`)}var bv=z({eye_:VP});function Cn(e,t,n){let a={shape:e,value:t,dtype:n};return L.runKernel(kc,{},a)}function UP(e){let t={x:F(e,"x","floor","float32")};return L.runKernel(Pi,t)}var Vu=z({floor_:UP});function GP(e,t,n=0,a=0){let r=F(e,"x","gather"),s=F(t,"indices","gather","int32"),i={x:r,indices:s},o={axis:n,batchDims:a};return L.runKernel(ru,i,o)}var ci=z({gather_:GP});function HP(e,t){let n=F(e,"a","greater","string_or_numeric"),a=F(t,"b","greater","string_or_numeric");[n,a]=At(n,a),pt(n.shape,a.shape);let r={a:n,b:a};return L.runKernel(iu,r)}var Un=z({greater_:HP});function jP(e,t){let n=F(e,"a","greaterEqual","string_or_numeric"),a=F(t,"b","greaterEqual","string_or_numeric");[n,a]=At(n,a),pt(n.shape,a.shape);let r={a:n,b:a};return L.runKernel(zi,r)}var Ns=z({greaterEqual_:jP});function qP(e){let t={x:F(e,"x","isFinite")};return L.runKernel(ou,t)}var jS=z({isFinite_:qP});function KP(e){let t={x:F(e,"x","isInf")};return L.runKernel(lu,t)}var qS=z({isInf_:KP});function XP(e){let t={x:F(e,"x","isNaN")};return L.runKernel(uu,t)}var xv=z({isNaN_:XP});function YP(e,t=.2){let n={x:F(e,"x","leakyRelu")},a={alpha:t};return L.runKernel(Wi,n,a)}var Gc=z({leakyRelu_:YP});function QP(e,t){let n=F(e,"a","less","string_or_numeric"),a=F(t,"b","less","string_or_numeric");[n,a]=At(n,a),pt(n.shape,a.shape);let r={a:n,b:a};return L.runKernel(pu,r)}var cf=z({less_:QP});function JP(e,t){let n=F(e,"a","lessEqual","string_or_numeric"),a=F(t,"b","lessEqual","string_or_numeric");[n,a]=At(n,a),pt(n.shape,a.shape);let r={a:n,b:a};return L.runKernel(cu,r)}var Ts=z({lessEqual_:JP});function KS(e,t,n){if(n<=0)throw new Error("The number of values should be positive.");let a={start:e,stop:t,num:n};return L.runKernel(zm,{},a)}function ZP(e,t=5,n=1,a=1,r=.5){let s=F(e,"x","localResponseNormalization");R(s.rank===4||s.rank===3,()=>`Error in localResponseNormalization: x must be rank 3 or 4 but got
rank ${s.rank}.`),R(wl(t),()=>`Error in localResponseNormalization: depthRadius must be an integer but got depthRadius ${t}.`);let i=s,o=!1;s.rank===3&&(o=!0,i=W(s,[1,s.shape[0],s.shape[1],s.shape[2]]));let l={x:i},u={depthRadius:t,bias:n,alpha:a,beta:r},p=L.runKernel(Nc,l,u);return o?W(p,[p.shape[1],p.shape[2],p.shape[3]]):p}var vv=z({localResponseNormalization_:ZP});function eO(e){let t={x:F(e,"x","log","float32")};return L.runKernel(Vi,t)}var ea=z({log_:eO});function tO(e){let t={x:F(e,"x","log1p")};return L.runKernel(du,t)}var Hc=z({log1p_:tO});function nO(e){return R(os(e),()=>"The f passed in grad(f) must be a function"),(t,n)=>{let a=F(t,"x","tf.grad","string_or_numeric"),r=n!=null?F(n,"dy","tf.grad"):null;return L.tidy(()=>{let{value:s,grads:i}=L.gradients(()=>e(a),[a],r);return r!=null&&Tn(s.shape,r.shape,"The shape of dy passed in grad(f)(x, dy) must match the shape returned by f(x)"),df(i),i[0]})}}function aO(e){return R(os(e),()=>"The f passed in grads(f) must be a function"),(t,n)=>{R(Array.isArray(t),()=>"The args passed in grads(f)(args) must be an array of `Tensor`s or `TensorLike`s");let a=ac(t,"args","tf.grads","string_or_numeric"),r=n!=null?F(n,"dy","tf.grads"):null;return L.tidy(()=>{let{value:s,grads:i}=L.gradients(()=>e(...a),a,r);return r!=null&&Tn(s.shape,r.shape,"The shape of dy passed in grads(f)([x1,...], dy) must match the shape returned by f([x1,...])"),df(i),i})}}function rO(e){return R(os(e),()=>"The f passed in valueAndGrad(f) must be a function"),(t,n)=>{R(t instanceof Fe,()=>"The x passed in valueAndGrad(f)(x) must be a tensor"),R(n==null||n instanceof Fe,()=>"The dy passed in valueAndGrad(f)(x, dy) must be a tensor");let{grads:a,value:r}=L.gradients(()=>e(t),[t],n);return df(a),{grad:a[0],value:r}}}function sO(e){return R(os(e),()=>"The f passed in valueAndGrads(f) must be a function"),(t,n)=>{R(Array.isArray(t)&&t.every(r=>r instanceof Fe),()=>"The args passed in valueAndGrads(f)(args) must be array of tensors"),R(n==null||n instanceof Fe,()=>"The dy passed in valueAndGrads(f)(args, dy) must be a tensor");let a=L.gradients(()=>e(...t),t,n);return n!=null&&Tn(a.value.shape,n.shape,"The shape of dy passed in valueAndGrads(f)([x1,...], dy) must match the shape returned by f([x1,...])"),df(a.grads),a}}function XS(e,t){R(os(e),()=>"The f passed in variableGrads(f) must be a function"),R(t==null||Array.isArray(t)&&t.every(u=>u instanceof ls),()=>"The varList passed in variableGrads(f, varList) must be an array of variables");let n=t!=null;if(!n){t=[];for(let u in L.registeredVariables)t.push(L.registeredVariables[u])}let a=n?t.filter(u=>!u.trainable):null,r=t.length;t=t.filter(u=>u.trainable),R(t.length>0,()=>`variableGrads() expects at least one of the input variables to be trainable, but none of the ${r} variables is trainable.`);let s=!0,{value:i,grads:o}=L.gradients(e,t,null,s);R(o.some(u=>u!=null),()=>"Cannot find a connection between any variable and the result of the loss function y=f(x). Please make sure the operations that use variables are inside the function f passed to minimize()."),R(i.rank===0,()=>`The f passed in variableGrads(f) must return a scalar, but it returned a rank-${i.rank} tensor`);let l={};return t.forEach((u,p)=>{o[p]!=null&&(l[u.name]=o[p])}),a!=null&&a.forEach(u=>l[u.name]=null),{value:i,grads:l}}function cr(e){return L.customGrad(e)}function df(e){if(e.filter(t=>t==null).length>0)throw new Error(`Cannot compute gradient of y=f(x) with respect to x. Make sure that
the f you passed encloses all operations that lead from x to y.`)}function iO(e){let t={x:F(e,"x","softplus")};return L.runKernel(Fu,t)}var xo=z({softplus_:iO});function oO(e){let t=F(e,"x","logSigmoid");return cr(n=>({value:It(xo(It(n))),gradFunc:a=>B(a,ma(It(n)))}))(t)}var YS=z({logSigmoid_:oO});function lO(e,t){let n=F(e,"a","sub"),a=F(t,"b","sub");[n,a]=At(n,a);let r={a:n,b:a};return L.runKernel(fo,r)}var ce=z({sub_:lO});function uO(e,t=-1){let n=F(e,"logits","logSoftmax");if(t===-1&&(t=n.rank-1),t!==n.rank-1)throw Error(`Log Softmax along a non-last dimension is not yet supported. Logits was rank ${n.rank} and axis was ${t}`);return cr((a,r)=>{let s=Ta(a,t,!0),i=ce(a,s),o=ce(oe(i,"float32"),ea(be(gn(i),t,!0)));return r([o]),{value:o,gradFunc:(l,u)=>{let[p]=u,d=!0,c=gn(p);return ce(l,B(be(l,t,d),c))}}})(n)}var hf=z({logSoftmax_:uO});function pO(e,t=null,n=!1){let a=F(e,"x","logSumExp"),r=Ea(t,a.shape),s=Ta(a,r,!0),i=ce(a,s),o=gn(i),l=be(o,r),u=ea(l),p=J(W(s,u.shape),u);if(n){let d=pi(p.shape,r);return W(p,d)}return p}var wv=z({logSumExp_:pO});function cO(e,t){let n=F(e,"a","logicalAnd","bool"),a=F(t,"b","logicalAnd","bool");pt(n.shape,a.shape);let r={a:n,b:a};return L.runKernel(hu,r)}var Ca=z({logicalAnd_:cO});function dO(e){let t={x:F(e,"x","logicalNot","bool")};return L.runKernel(Ic,t)}var jc=z({logicalNot_:dO});function hO(e,t){let n=F(e,"a","logicalOr","bool"),a=F(t,"b","logicalOr","bool");pt(n.shape,a.shape);let r={a:n,b:a};return L.runKernel(Sc,r)}var mf=z({logicalOr_:hO});function mO(e,t){let n=F(e,"a","logicalXor","bool"),a=F(t,"b","logicalXor","bool");return pt(n.shape,a.shape),Ca(mf(e,t),jc(Ca(e,t)))}var QS=z({logicalXor_:mO}),xh=2147483648;function fO(e,t,n="left"){let a=F(e,"sortedSequence","searchSorted"),r=F(t,"values","searchSorted"),s=a.shape[a.shape.length-1],i=r.shape[r.shape.length-1],o=W(a,[-1,s]),l=W(r,[-1,i]);if(o.rank<2)throw new Error("Sorted input argument must be at least 2-dimensional");if(o.shape[0]!==l.shape[0])throw new Error("Leading dimension of 'sortedSequence' and 'values' must match.");if(bt(l.shape)>=xh)throw new Error(`values tensor size must less than ${xh}`);if(o.shape[1]>=xh)throw new Error(`trailing dim_size must less than ${xh} for int32 output type, was ${o.shape[1]}`);let u={sortedSequence:o,values:l},p={side:n};return L.runKernel(Km,u,p)}var kv=z({searchSorted_:fO});function JS(e,t){return kv(e,t,"left")}function gO(e,t,n,a,r){let s=F(e,"x","maxPool"),i=1,o=s,l=!1;s.rank===3&&(l=!0,o=W(s,[1,s.shape[0],s.shape[1],s.shape[2]])),R(o.rank===4,()=>`Error in maxPool: input must be rank 4 but got rank ${o.rank}.`),R(mr(n,i),()=>`Error in maxPool: Either strides or dilations must be 1. Got strides ${n} and dilations '${i}'`),_n("maxPool",a,r);let u={x:o},p={filterSize:t,strides:n,pad:a,dimRoundingMode:r},d=L.runKernel(Hi,u,p);return l?W(d,[d.shape[1],d.shape[2],d.shape[3]]):d}var Pt=z({maxPool_:gO});function yO(e,t=[1,1,1],n,a,r,s="NDHWC"){let i=F(e,"x","maxPool3d"),o=i,l=!1;i.rank===4&&(l=!0,o=W(i,[1,i.shape[0],i.shape[1],i.shape[2],i.shape[3]])),R(o.rank===5,()=>`Error in maxPool3d: x must be rank 5 but got rank ${o.rank}.`),R(s==="NDHWC",()=>`Error in maxPool3d: Only NDHWC is currently supported, but got dataFormat of ${s}`),_n("maxPool3d",a,r);let u={x:o},p={filterSize:t,strides:n,pad:a,dimRoundingMode:r,dataFormat:s},d=L.runKernel(Tc,u,p);return l?W(d,[d.shape[1],d.shape[2],d.shape[3],d.shape[4]]):d}var Iv=z({maxPool3d_:yO});function bO(e,t,n,a,r=!1){let s={x:F(e,"x","maxPoolWithArgmax")},i={filterSize:t,strides:n,pad:a,includeBatchInIndex:r},o=L.runKernel(Um,s,i);return{result:o[0],indexes:o[1]}}var ZS=z({maxPoolWithArgmax_:bO});function xO(e,t){let n=F(e,"a","maximum"),a=F(t,"b","maximum");[n,a]=At(n,a),n.dtype==="bool"&&(n=oe(n,"int32"),a=oe(a,"int32")),pt(n.shape,a.shape);let r={a:n,b:a};return L.runKernel(Gi,r)}var fr=z({maximum_:xO});function vO(e,t=null,n=!1){let a={x:F(e,"x","mean")},r={axis:t,keepDims:n};return L.runKernel(ji,a,r)}var Et=z({mean_:vO});function St(e,t="float32"){if(t==="complex64"){let a=St(e,"float32"),r=St(e,"float32");return Ar(a,r)}let n=wm(bt(e),t);return L.makeTensor(n,e,t)}function Qn(e,t="float32"){if(t==="complex64"){let a=Qn(e,"float32"),r=St(e,"float32");return Ar(a,r)}let n=Ax(bt(e),t);return L.makeTensor(n,e,t)}function wO(e,t,{indexing:n="xy"}={}){if(n!=="xy"&&n!=="ij")throw new TypeError(`${n} is not a valid third argument to meshgrid`);if(e===void 0)return[];let a=F(e,"x","meshgrid",e instanceof Fe?e.dtype:"float32");if(t===void 0)return[a];let r=F(t,"y","meshgrid",t instanceof Fe?t.dtype:"float32"),s=bt(a.shape),i=bt(r.shape);return n==="xy"?(a=W(a,[1,-1]),r=W(r,[-1,1]),[De(Qn([i,1],a.dtype),a),De(r,Qn([1,s],r.dtype))]):(a=W(a,[-1,1]),r=W(r,[1,-1]),[De(a,Qn([1,i],a.dtype)),De(Qn([s,1],r.dtype),r)])}function kO(e,t){let n=F(e,"a","minimum"),a=F(t,"b","minimum");[n,a]=At(n,a),n.dtype==="bool"&&(n=oe(n,"int32"),a=oe(a,"int32")),pt(n.shape,a.shape);let r={a:n,b:a};return L.runKernel(Ki,r)}var Uu=z({minimum_:kO});function IO(e,t,n){R(n==="reflect"||n==="symmetric",()=>`Invalid mode. Mode must be either reflect or symmetric. Got ${n}.`);let a=F(e,"x","mirrorPad");if(a.rank===0)throw new Error("mirrorPad(scalar) is not defined. Pass non-scalar to mirrorPad");R(t.length===a.rank,()=>`Padding doesn't match input. Must be ${a.rank}. Got ${t.length}.`);let r=n==="reflect"?1:0;for(let o=0;o<a.rank;o++)R(t[o].length===2,()=>"Invalid number of paddings. Must be length of 2 each."),R(t[o][0]>=0&&t[o][0]<=a.shape[o]-r&&t[o][1]>=0&&t[o][1]<=a.shape[o]-r,()=>`Padding in dimension ${o} cannot be greater than or equal to ${a.shape[o]-r} or less than 0 for input of shape ${a.shape}`);let s={paddings:t,mode:n},i={x:a};return L.runKernel(Xi,i,s)}var Sv=z({mirrorPad_:IO});function SO(e,t){let n=F(e,"a","mod"),a=F(t,"b","mod");[n,a]=At(n,a);let r={a:n,b:a};return L.runKernel(mu,r)}var Nv=z({mod_:SO});function NO(e,t=null,n=!1){e=F(e,"x","moments");let a=Ea(t,e.shape),r=Et(e,a,n),s=r.shape;n||(s=pi(r.shape,a));let i=ut(ce(oe(e,"float32"),W(r,s))),o=Et(i,a,n);return{mean:r,variance:o}}var ff=z({moments_:NO});function TO(e,t,n,a){let r=F(t,"data","multiRNNCell"),s=ac(n,"c","multiRNNCell"),i=ac(a,"h","multiRNNCell"),o=r,l=[];for(let d=0;d<e.length;d++){let c=e[d](o,s[d],i[d]);l.push(c[0]),l.push(c[1]),o=c[1]}let u=[],p=[];for(let d=0;d<l.length;d+=2)u.push(l[d]),p.push(l[d+1]);return[u,p]}var _O=z({multiRNNCell_:TO});function CO(e,t,n,a=!1){let r=F(e,"logits","multinomial"),s=r.size,i=r.rank;if(s<2)throw new Error(`Error in multinomial: you need at least 2 outcomes, but got ${s}.`);if(i>2)throw new Error(`Rank of probabilities must be 1 or 2, but is ${i}`);n=n||Math.random();let o={logits:i===1?W(r,[1,-1]):r},l={numSamples:t,seed:n,normalized:a},u=L.runKernel(Gm,o,l);return i===1?W(u,[u.size]):u}var e2=z({multinomial_:CO});function EO(e,t){let n=F(e,"a","notEqual","string_or_numeric"),a=F(t,"b","notEqual","string_or_numeric");[n,a]=At(n,a),pt(n.shape,a.shape);let r={a:n,b:a};return L.runKernel(gu,r)}var di=z({notEqual_:EO});function FO(e){let t={x:F(e,"x","onesLike")};return L.runKernel(vu,t)}var ta=z({onesLike_:FO});function AO(e,t){let n=F(e,"v1","outerProduct"),a=F(t,"v2","outerProduct");R(n.rank===1&&a.rank===1,()=>`Error in outerProduct: inputs must be rank 1, but got ranks ${n.rank} and ${a.rank}.`);let r=W(n,[-1,1]),s=W(a,[1,-1]);return De(r,s)}var $O=z({outerProduct_:AO});function DO(e,t,n=0){let a=F(e,"x","pad");if(a.rank===0)throw new Error("pad(scalar) is not defined. Pass non-scalar to pad");let r={paddings:t,constantValue:n},s={x:a};return L.runKernel(Ji,s,r)}var ya=z({pad_:DO});function RO(e,t,n=0){return R(t.length===2,()=>"Invalid number of paddings. Must be length of 2."),ya(e,[t],n)}var MO=z({pad1d_:RO});function PO(e,t,n=0){return R(t.length===2&&t[0].length===2&&t[1].length===2,()=>"Invalid number of paddings. Must be length of 2 each."),ya(e,t,n)}var OO=z({pad2d_:PO});function LO(e,t,n=0){return R(t.length===3&&t[0].length===2&&t[1].length===2&&t[2].length===2,()=>"Invalid number of paddings. Must be length of 2 each."),ya(e,t,n)}var zO=z({pad3d_:LO});function BO(e,t,n=0){return R(t.length===4&&t[0].length===2&&t[1].length===2&&t[2].length===2&&t[3].length===2,()=>"Invalid number of paddings. Must be length of 2 each."),ya(e,t,n)}var WO=z({pad4d_:BO});function VO(e,t,n){let a=F(e,"x","spaceToBatchND");R(a.rank>=1+t.length,()=>`input rank ${a.rank} should be > than [blockShape] ${t.length}`),R(n.length===t.length,()=>`paddings.shape[0] ${n.length} must be equal to [blockShape] ${t.length}`),R(a.shape.reduce((i,o,l)=>l>0&&l<=t.length?i&&(o+n[l-1][0]+n[l-1][1])%t[l-1]===0:i,!0),()=>`input spatial dimensions ${a.shape.slice(1)} with paddings ${n.toString()} must be divisible by blockShapes ${t.toString()}`);let r={x:a},s={blockShape:t,paddings:n};return L.runKernel(Au,r,s)}var qc=z({spaceToBatchND_:VO});function UO(e,t,n,a,r,s,i){r==null&&(r=[1,1]),s==null&&(s=1),a===0&&(a="valid");let o=F(e,"x","maxPool"),l=o,u=!1;o.rank===3&&(u=!0,l=W(o,[1,o.shape[0],o.shape[1],o.shape[2]])),R(mr(s,r),()=>`Error in pool: Either strides or dilations must be 1. Got strides ${s} and dilations '${r}'`);let p=TS(l.shape,t,s,r,a),d=[p.dilationHeight,p.dilationWidth],c;a==="same"?c=HO([p.filterHeight,p.filterWidth],d):c=[[0,0],[0,0]];let h=d[0]===1&&d[1]===1,[m,f]=GO([p.inHeight,p.inWidth],d,c),g=h?a:"valid",y=h?l:qc(l,d,m),b=(n==="avg"?()=>ga(y,t,s,g,i):()=>Pt(y,t,s,g,i))(),x=h?b:Wc(b,d,f);return u?W(x,[x.shape[1],x.shape[2],x.shape[3]]):x}function GO(e,t,n){let a=n.map(p=>p[0]),r=n.map(p=>p[1]),s=e.concat(a,r),i=t.map((p,d)=>(p-s[d]%p)%p),o=r.map((p,d)=>p+i[d]),l=t.map((p,d)=>[a[d],o[d]]),u=t.map((p,d)=>[0,i[d]]);return[l,u]}function HO(e,t){let n=e.map((s,i)=>s+(s-1)*(t[i]-1)).map(s=>s-1),a=n.map(s=>Math.floor(s/2)),r=n.map((s,i)=>s-a[i]);return n.map((s,i)=>[a[i],r[i]])}var t2=z({pool_:UO});function jO(e,t){let n=F(e,"x","prelu"),a=F(t,"alpha","prelu"),r={x:n,alpha:a};return L.runKernel(eo,r)}var Kc=z({prelu_:jO});function qO(e,t=null,n=!1){let a=F(e,"x","prod");a.dtype==="bool"&&(a=oe(a,"int32"));let r={x:a},s={axis:t,keepDims:n};return L.runKernel(to,r,s)}var gf=z({prod_:qO});function KO(e,t,n){let a=bt(e),r=null;if(n==null||n==="float32")r=new Float32Array(a);else if(n==="int32")r=new Int32Array(a);else if(n==="bool")r=new Uint8Array(a);else throw new Error(`Unknown data type ${n}`);for(let s=0;s<a;s++)r[s]=t();return L.makeTensor(r,e,n)}var XO=z({rand_:KO}),Tv=xi(bI()),_v=class{constructor(e,t,n,a,r){this.mean=e,this.stdDev=t,this.dtype=n,this.nextVal=NaN,this.truncated=a,this.truncated&&(this.upper=this.mean+this.stdDev*2,this.lower=this.mean-this.stdDev*2);let s=r||Math.random();this.random=Tv.alea(s.toString())}nextValue(){if(!isNaN(this.nextVal)){let a=this.nextVal;return this.nextVal=NaN,a}let e,t,n=!1;for(;!n;){let a,r,s;do a=2*this.random()-1,r=2*this.random()-1,s=a*a+r*r;while(s>=1||s===0);let i=Math.sqrt(-2*Math.log(s)/s);e=this.mean+this.stdDev*a*i,t=this.mean+this.stdDev*r*i,(!this.truncated||this.isValidTruncated(e))&&(n=!0)}return(!this.truncated||this.isValidTruncated(t))&&(this.nextVal=this.convertValue(t)),this.convertValue(e)}convertValue(e){return this.dtype==null||this.dtype==="float32"?e:Math.round(e)}isValidTruncated(e){return e<=this.upper&&e>=this.lower}},YO=class{constructor(e,t,n,a){this.alpha=e,this.beta=1/t,this.dtype=n;let r=a||Math.random();this.randu=Tv.alea(r.toString()),this.randn=new _v(0,1,n,!1,this.randu()),e<1?this.d=e+2/3:this.d=e-1/3,this.c=1/Math.sqrt(9*this.d)}nextValue(){let e,t,n,a,r,s;for(;;){do a=this.randn.nextValue(),s=1+this.c*a;while(s<=0);if(s*=s*s,e=a*a,t=1-.331*e*e,n=.5*e+this.d*(1-s+Math.log(s)),r=this.randu(),r<t||Math.log(r)<n)break}return s=1/this.beta*this.d*s,this.alpha<1&&(s*=Math.pow(this.randu(),1/this.alpha)),this.convertValue(s)}convertValue(e){return this.dtype==="float32"?e:Math.round(e)}},QO=class{constructor(e=0,t=1,n,a){if(this.canReturnFloat=()=>this.dtype==null||this.dtype==="float32",this.min=e,this.range=t-e,this.dtype=n,a==null&&(a=Math.random()),typeof a=="number"&&(a=a.toString()),!this.canReturnFloat()&&this.range<=1)throw new Error(`The difference between ${e} - ${t} <= 1 and dtype is not float`);this.random=Tv.alea(a)}convertValue(e){return this.canReturnFloat()?e:Math.round(e)}nextValue(){return this.convertValue(this.min+this.range*this.random())}};function JO(e,t,n=1,a="float32",r){if(n==null&&(n=1),a==null&&(a="float32"),a!=="float32"&&a!=="int32")throw new Error(`Unsupported data type ${a}`);let s=new YO(t,n,a,r),i=Ve(e,a);for(let o=0;o<i.values.length;o++)i.values[o]=s.nextValue();return i.toTensor()}var ZO=z({randomGamma_:JO});function e3(e,t=0,n=1,a,r){if(a!=null&&a==="bool")throw new Error(`Unsupported data type ${a}`);let s=new _v(t,n,a,!1,r),i=Ve(e,a);for(let o=0;o<i.values.length;o++)i.values[o]=s.nextValue();return i.toTensor()}var n2=z({randomNormal_:e3});function t3(e,t=0,n=1,a="float32",r){let s=Ve(e,a),i=new QO(t,n,null,r);for(let o=0;o<s.values.length;o++)s.values[o]=i.nextValue();return s.toTensor()}var Gu=z({randomUniform_:t3});function Tl(e,t,n=1,a="float32"){if(n===0)throw new Error("Cannot have a step of zero");let r={start:e,stop:t,step:n,dtype:a};return L.runKernel(_c,{},r)}function n3(e){let t={x:F(e,"x","reciprocal")};return L.runKernel(ku,t)}var Cv=z({reciprocal_:n3});function a3(e){let t={x:F(e,"x","relu")};return L.runKernel(no,t)}var Xe=z({relu_:a3});function r3(e){let t={x:F(e,"x","relu6")};return L.runKernel(ro,t)}var yf=z({relu6_:r3});function s3(e,t){let n={x:F(e,"x","reverse")},a={dims:t};return L.runKernel(so,n,a)}var na=z({reverse_:s3});function i3(e){let t=F(e,"x","reverse");return R(t.rank===1,()=>`Error in reverse1D: x must be rank 1 but got rank ${t.rank}.`),na(t,0)}var o3=z({reverse1d_:i3});function l3(e,t){let n=F(e,"x","reverse");return R(n.rank===2,()=>`Error in reverse2D: x must be rank 2 but got rank ${n.rank}.`),na(n,t)}var u3=z({reverse2d_:l3});function p3(e,t){let n=F(e,"x","reverse");return R(n.rank===3,()=>`Error in reverse3D: x must be rank 3 but got rank ${n.rank}.`),na(n,t)}var c3=z({reverse3d_:p3});function d3(e,t){let n=F(e,"x","reverse");return R(n.rank===4,()=>`Error in reverse4D: x must be rank 4 but got rank ${n.rank}.`),na(n,t)}var h3=z({reverse4d_:d3});function m3(e){let t={x:F(e,"x","round")};return L.runKernel(io,t)}var bf=z({round_:m3});function f3(e){let t={x:F(e,"x","rsqrt","float32")};return L.runKernel(oo,t)}var xf=z({rsqrt_:f3});function g3(e){let t={x:F(e,"x","selu")};return L.runKernel(Tu,t)}var vf=z({selu_:g3});function y3(e,t,n,a,r,s=[1,1],i="NHWC"){let o=F(e,"x","separableConv2d"),l=F(t,"depthwiseFilter","separableConv2d"),u=F(n,"pointwiseFilter","separableConv2d"),p=o,d=!1;if(o.rank===3&&(d=!0,p=W(o,[1,o.shape[0],o.shape[1],o.shape[2]])),i==="NCHW")throw new Error("separableConv2d currently does not support dataFormat NCHW; only NHWC is supported");R(p.rank===4,()=>`Error in separableConv2d: input must be rank 4, but got rank ${p.rank}.`),R(l.rank===4,()=>`Error in separableConv2d: depthwise filter must be rank 4, but got rank ${l.rank}.`),R(u.rank===4,()=>`Error in separableConv2d: pointwise filter must be rank 4, but got rank ${l.rank}.`),R(u.shape[0]===1,()=>`Error in separableConv2d: the first dimension of pointwise filter must be 1, but got ${u.shape[0]}.`),R(u.shape[1]===1,()=>`Error in separableConv2d: the second dimension of pointwise filter must be 1, but got ${u.shape[1]}.`);let c=l.shape[2],h=l.shape[3];R(u.shape[2]===c*h,()=>`Error in separableConv2d: the third dimension of pointwise filter must be ${c*h}, but got ${u.shape[2]}.`);let m=Ss(p,l,a,r,i,s),f=Rt(m,u,1,"valid",i);return d?W(f,[f.shape[1],f.shape[2],f.shape[3]]):f}var vo=z({separableConv2d_:y3});async function b3(e,t){let n=F(e,"x","setdiff1d"),a=F(t,"y","setdiff1d");R(n.dtype===a.dtype,()=>`x and y should have the same dtype, but got x (${n.dtype}) and y (${a.dtype}).`),R(n.rank===1,()=>`x should be 1D tensor, but got x (${n.shape}).`),R(a.rank===1,()=>`y should be 1D tensor, but got y (${a.shape}).`);let r=await n.data(),s=await a.data(),i=new Set(s),o=0;for(let p=0;p<r.length;p++)i.has(r[p])||o++;let l=new qt([o],n.dtype),u=new qt([o],"int32");for(let p=0,d=0;p<r.length;p++)i.has(r[p])||(l.values[d]=r[p],u.values[d]=p,d++);return[l.toTensor(),u.toTensor()]}var a2=b3;function x3(e){let t={x:F(e,"x","sign")};return L.runKernel(Eu,t)}var Ev=z({sign_:x3});function v3(e){let t={x:F(e,"x","sin","float32")};return L.runKernel(lo,t)}var wf=z({sin_:v3});function w3(e){let t={x:F(e,"x","sinh")};return L.runKernel(Cu,t)}var kf=z({sinh_:w3});function k3(e,t,n){let a=F(e,"x","slice1d");return R(a.rank===1,()=>`slice1d expects a rank-1 tensor, but got a rank-${a.rank} tensor`),He(a,[t],[n])}var If=z({slice1d_:k3});function I3(e,t,n){let a=F(e,"x","slice2d");return R(a.rank===2,()=>`slice2d expects a rank-2 tensor, but got a rank-${a.rank} tensor`),He(a,t,n)}var Fv=z({slice2d_:I3});function S3(e,t,n){let a=F(e,"x","slice3d");return R(a.rank===3,()=>`slice3d expects a rank-3 tensor, but got a rank-${a.rank} tensor`),He(a,t,n)}var Hu=z({slice3d_:S3});function N3(e,t,n){let a=F(e,"x","slice4d");return R(a.rank===4,()=>`slice4d expects a rank-4 tensor, but got a rank-${a.rank} tensor`),He(a,t,n)}var oc=z({slice4d_:N3});function T3(e,t=-1){let n=F(e,"logits","softmax","float32");if(t===-1&&(t=n.rank-1),t!==n.rank-1)throw Error(`Softmax along a non-last dimension is not yet supported. Logits was rank ${n.rank} and dim was ${t}`);let a={logits:n},r={dim:t};return L.runKernel(ho,a,r)}var Qa=z({softmax_:T3});function _3(e){R(e.dtype==="complex64",()=>`The dtype for tf.spectral.fft() must be complex64 but got ${e.dtype}.`);let t={input:e};return L.runKernel(Pm,t)}var Xc=z({fft_:_3});function C3(e){R(e.dtype==="complex64",()=>`The dtype for tf.spectral.ifft() must be complex64 but got ${e.dtype}.`);let t={input:e};return L.runKernel(Om,t)}var _l=z({ifft_:C3});function E3(e){let t=e.shape[e.shape.length-1],n=e.size/t,a;if(t<=2){let r=W(e,[n,t]);a=_l(r)}else{let r=[n,2*(t-1)],s=W(Nl(e),[n,t]),i=W(zc(e),[n,t]),o=na(He(s,[0,1],[n,t-2]),1),l=B(na(He(i,[0,1],[n,t-2]),1),we(-1)),u=Ze([s,o],1),p=Ze([i,l],1),d=W(Ar(u,p),[r[0],r[1]]);a=_l(d)}if(a=Nl(a),e.rank===3&&e.shape[0]!==0){let r=a,s=e.shape[0];a=W(a,[s,a.shape[0]/s,a.shape[1]]),r.dispose()}return a}var Sf=z({irfft_:E3});function F3(e,t,n=0){let a={x:F(e,"x","split")},r={numOrSizeSplits:t,axis:n};return L.runKernel($u,a,r)}var zn=z({split_:F3});function A3(e,t){R(e.dtype==="float32",()=>`The dtype for rfft() must be real value but got ${e.dtype}`);let n=e.shape[e.shape.length-1],a=e.size/n,r;if(t!=null&&t<n){let m=e.shape.map(g=>0),f=e.shape.map(g=>g);f[e.shape.length-1]=t,r=He(e,m,f),n=t}else if(t!=null&&t>n){let m=e.shape.map(f=>f);m[e.shape.length-1]=t-n,r=Ze([e,St(m)],e.shape.length-1),n=t}else r=e;let s=Ke(r),i=W(Ar(r,s),[a,n]),o=Xc(i),l=Math.floor(n/2)+1,u=Nl(o),p=zc(o),d=zn(u,[l,n-l],u.shape.length-1),c=zn(p,[l,n-l],p.shape.length-1),h=r.shape.slice();return h[r.shape.length-1]=l,W(Ar(d[0],c[0]),h)}var Yc=z({rfft_:A3});function $3(e,t){let n=F(e,"a","squaredDifference"),a=F(t,"b","squaredDifference");[n,a]=At(n,a),pt(n.shape,a.shape);let r={a:n,b:a},s={};return L.runKernel(mo,r,s)}var Nf=z({squaredDifference_:$3});function D3(e,t){let n=F(e,"x","squeeze");return W(n,kI(n.shape,t).newShape)}var dr=z({squeeze_:D3});function R3(e,t=0){let n=ac(e,"tensors","stack","string_or_numeric");R(n.length>=1,()=>"Pass at least one tensor to tf.stack"),n.length>0&&R(t<=n[0].rank,()=>"Axis must be <= rank of the tensor");let a=n,r={axis:t};return L.runKernel(wu,a,r)}var Mt=z({stack_:R3});function M3(e,t=0){let n={x:F(e,"x","step")},a={alpha:t};return L.runKernel(ws,n,a)}var ju=z({step_:M3});function P3(e,t,n,a,r=0,s=0,i=0,o=0,l=0){let u={x:F(e,"x","stridedSlice","string_or_numeric")},p={begin:t,end:n,strides:a,beginMask:r,endMask:s,ellipsisMask:i,newAxisMask:o,shrinkAxisMask:l};return L.runKernel(Ru,u,p)}var Av=z({stridedSlice_:P3});function O3(e){let t={x:F(e,"x","tan","float32")};return L.runKernel(go,t)}var $v=z({tan_:O3});function qe(e,t){vi(e);let n=pr(e,t);if(n.length!==1)throw new Error("tensor1d() requires values to be a flat/TypedArray");return ks(e,null,n,t)}function Ha(e,t,n){if(vi(e),t!=null&&t.length!==2)throw new Error("tensor2d() requires shape to have two numbers");let a=pr(e,n);if(a.length!==2&&a.length!==1)throw new Error("tensor2d() requires values to be number[][] or flat/TypedArray");if(a.length===1&&t==null)throw new Error("tensor2d() requires shape to be provided when `values` are a flat/TypedArray");return ks(e,t,a,n)}function Ja(e,t,n){if(vi(e),t!=null&&t.length!==4)throw new Error("tensor4d() requires shape to have four numbers");let a=pr(e,n);if(a.length!==4&&a.length!==1)throw new Error("tensor4d() requires values to be number[][][][] or flat/TypedArray");if(a.length===1&&t==null)throw new Error("tensor4d() requires shape to be provided when `values` are a flat array");return ks(e,t,a,n)}function L3(e,t,n){if(vi(e),t!=null&&t.length!==5)throw new Error("tensor5d() requires shape to have five numbers");let a=pr(e,n);if(a.length!==5&&a.length!==1)throw new Error("tensor5d() requires values to be number[][][][][] or flat/TypedArray");if(a.length===1&&t==null)throw new Error("tensor5d() requires shape to be provided when `values` are a flat array");return ks(e,t,a,n)}function z3(e,t,n){if(vi(e),t!=null&&t.length!==6)throw new Error("tensor6d() requires shape to have six numbers");let a=pr(e,n);if(a.length!==6&&a.length!==1)throw new Error("tensor6d() requires values to be number[][][][][][] or flat/TypedArray");if(a.length===1&&t==null)throw new Error("tensor6d() requires shape to be provided when `values` are a flat array");return t=t||a,ks(e,t,a,n)}function B3(e,t=1,n=!0){let a=F(e,"x","topk");if(a.rank===0)throw new Error("topk() expects the input to be of rank 1 or higher");let r=a.shape[a.shape.length-1];if(t<0)throw new Error(`'k' passed to topk() must be >= 0 but got ${t}`);if(t>r)throw new Error(`'k' passed to topk() must be <= the last dimension (${r}) but got ${t}`);let s={x:a},i={k:t,sorted:n},[o,l]=L.runKernel(Mu,s,i);return{values:o,indices:l}}var Dv=z({topk_:B3});function W3(e,t=0,n=1,a,r){if(a!=null&&a==="bool")throw new Error("Unsupported data type $ { dtype }");let s=new _v(t,n,a,!0,r),i=Ve(e,a);for(let o=0;o<i.values.length;o++)i.values[o]=s.nextValue();return i.toTensor()}var Tf=z({truncatedNormal_:W3});function V3(e,t=0){let n=F(e,"x","unique","string_or_numeric");R(n.rank>0,()=>"The input tensor must be at least 1D");let a={x:n},r={axis:t},[s,i]=L.runKernel(Zm,a,r);return{values:s,indices:i}}var Zh=z({unique_:V3});function U3(e,t,n){let a=F(e,"x","unsortedSegmentSum"),r=F(t,"segmentIds","unsortedSegmentSum","int32");R(wl(n),()=>"numSegments must be of dtype int");let s={x:a,segmentIds:r},i={numSegments:n};return L.runKernel(Dc,s,i)}var Rv=z({unsortedSegmentSum_:U3});function G3(e,t=0){let n=F(e,"x","unstack","string_or_numeric");R(t>=-n.shape.length&&t<n.shape.length,()=>`Axis = ${t} is not in [-${n.shape.length}, ${n.shape.length})`);let a={value:n},r={axis:t};return L.runKernel(Ou,a,r)}var ht=z({unstack_:G3});function r2(e,t){return kv(e,t,"right")}function s2(e,t=!0,n,a){return L.makeVariable(e,t,n,a)}function i2(e,t){let n=[];for(let s=0;s<t.length;s++)t[s]&&n.push(s);let a=Ve(e,"int32"),r=Ve([n.length,e.length],"int32");for(let s=0;s<n.length;s++){let i=a.indexToLoc(n[s]),o=s*e.length;r.values.set(i,o)}return r.toTensor()}async function H3(e){let t=F(e,"condition","whereAsync","bool"),n=await t.data(),a=i2(t.shape,n);return e!==t&&t.dispose(),a}var Mv=H3;async function j3(e,t,n){let a=F(e,"tensor","boolMask"),r=F(t,"mask","boolMask","bool"),s=n==null?0:n,i=r.rank,o=a.shape;R(i>0,()=>"mask cannot be scalar"),Tn(o.slice(s,s+i),r.shape,"mask's shape must match the first K dimensions of tensor's shape,");let l=1;for(let f=s;f<s+i;f++)l*=o[f];let u=o.slice(0,s).concat([l],o.slice(s+i)),p=W(a,u),d=W(r,[-1]),c=await Mv(d),h=dr(c,[1]),m=ci(p,h,s);return e!==a&&a.dispose(),t!==r&&r.dispose(),h.dispose(),p.dispose(),d.dispose(),c.dispose(),m}var q3=j3;function K3(e,t,n,a,r=!0){let s=F(e,"v","movingAverage"),i=F(t,"x","movingAverage"),o=F(n,"decay","movingAverage");zI(s,i),R(ys(s.shape,i.shape),()=>"Shape mismatch in v and x");let l=we(1),u=ce(l,o),p=B(ce(i,s),u);if(r){R(a!=null,()=>"When using zeroDebias: true, step is required.");let d=F(a,"step","movingAverage");p=fe(p,ce(l,Dr(o,d)))}return J(s,p)}var X3=z({movingAverage_:K3});function Y3(e,t,n){let a=F(e,"indices","scatterND","int32"),r=F(t,"updates","scatterND");qx(r,a,n);let s={indices:a,updates:r},i={shape:n};return L.runKernel(Su,s,i)}var o2=z({scatterND_:Y3});function Q3(e,t,n,a){if(e.dtype!=="int32")throw new Error(`tf.sparseToDense() expects the indices to be int32 type, but the dtype was ${e.dtype}.`);if(e.rank>2)throw new Error(`sparseIndices should be a scalar, vector, or matrix, but got shape ${e.shape}.`);let r=e.rank>0?e.shape[0]:1,s=e.rank>1?e.shape[1]:1;if(n.length!==s)throw new Error(`outputShape has incorrect number of elements:, ${n.length}, should be: ${s}.`);let i=t.size;if(!(t.rank===0||t.rank===1&&i===r))throw new Error(`sparseValues has incorrect shape ${t.shape}, should be [] or [${r}]`);if(t.dtype!==a.dtype)throw new Error("sparseValues.dtype must match defaultValues.dtype")}function J3(e,t,n,a=0){let r=F(e,"sparseIndices","sparseToDense","int32"),s=F(t,"sparseValues","sparseToDense","string_or_numeric"),i=F(a,"defaultValue","sparseToDense",s.dtype);Q3(r,s,n,i);let o={sparseIndices:r,sparseValues:s,defaultValue:i},l={outputShape:n};return L.runKernel(Xm,o,l)}var Pv=z({sparseToDense_:J3});function Z3(e,t){let n=F(t,"indices","gatherND","int32"),a={params:F(e,"x","gatherND","string_or_numeric"),indices:n};return L.runKernel(su,a)}var l2=z({gatherND_:Z3});function eL(e,t){if(t==null)return e.shape.slice();if(ys(e.shape,t))return t;if(e.shape.length===t.length){let n=[];for(let a=0;a<e.shape.length;a++)t[a]==null&&e.shape[a]!=null?n.push(e.shape[a]):n.push(t[a]);return n}return t}function tL(e,t,n,a){let r=F(e,"x","dropout");if(R(r.dtype==="float32",()=>`x has to be a floating point tensor since it's going to be scaled, but got a ${r.dtype} tensor instead.`),R(t>=0&&t<1,()=>`rate must be a float in the range [0, 1), but got ${t}.`),t===0)return e instanceof Fe?r.clone():r;let s=eL(r,n),i=1-t,o=fe(Vu(J(Gu(s,0,1,"float32",a),i)),i);return B(r,o)}var u2=z({dropout_:tL});function p2(e){return Math.floor(Math.pow(2,Math.ceil(Math.log(e)/Math.log(2))))}function Ov(e,t,n){let a=1-e%2,r=new Float32Array(e);for(let s=0;s<e;++s){let i=2*Math.PI*s/(e+a-1);r[s]=t-n*Math.cos(i)}return qe(r,"float32")}async function nL(e,t,n=1){let a=F(e,"predictions","inTopK"),r=F(t,"targets","inTopK");R(a.rank>1,()=>`inTopK() expects the predictions to be of rank 2 or higher, but got ${a.rank}`),R(a.rank-1===r.rank,()=>`predictions rank should be 1 larger than targets rank, but got predictions rank ${a.rank} and targets rank ${r.rank}`),Tn(a.shape.slice(0,a.shape.length-1),r.shape,"predictions's shape should be align with the targets' shape, except the last dimension.");let s=a.shape[a.shape.length-1];R(n>0&&n<=s,()=>`'k' passed to inTopK() must be > 0 && <= the predictions last dimension (${s}), but got ${n}`);let i=await a.data(),o=await r.data(),[l,u]=[i.length/s,s],p=II("bool",l);for(let d=0;d<l;d++){let c=d*u,h=i.subarray(c,c+u),m=[];for(let f=0;f<h.length;f++)m.push({value:h[f],index:f});m.sort((f,g)=>g.value-f.value),p[d]=0;for(let f=0;f<n;f++)if(m[f].index===o[d]){p[d]=1;break}}return e!==a&&a.dispose(),t!==r&&r.dispose(),Jn(p,r.shape,"bool")}var aL=nL,ps={};Me(ps,{conv2d:()=>iL,depthwiseConv2d:()=>pL,matMul:()=>dL});function rL(e,t,n,a,r,s="NHWC",i){let o=e;e.rank===3&&(o=W(e,[1,e.shape[0],e.shape[1],e.shape[2]]));let l=t;l.rank===3&&(l=W(t,[1,t.shape[0],t.shape[1],t.shape[2]])),R(o.rank===4,()=>`Error in conv2dDerFilter: input must be rank 4, but got shape ${o.shape}.`),R(l.rank===4,()=>`Error in conv2dDerFilter: dy must be rank 4, but got shape ${l.shape}.`),R(n.length===4,()=>`Error in conv2dDerFilter: filterShape must be length 4, but got ${n}.`);let u=s==="NHWC"?o.shape[3]:o.shape[1],p=s==="NHWC"?l.shape[3]:l.shape[1];R(u===n[2],()=>`Error in conv2dDerFilter: depth of input ${u}) must match input depth in filter (${n[2]}.`),R(p===n[3],()=>`Error in conv2dDerFilter: depth of dy (${p}) must match output depth for filter (${n[3]}).`),_n("conv2dDerFilter",r,i);let d={x:o,dy:l},c={strides:a,pad:r,dataFormat:s,dimRoundingMode:i,filterShape:n};return L.runKernel(_m,d,c)}var Lv=z({conv2DBackpropFilter_:rL});function _f(e,t,n){if(n==null||n==="linear")return e;if(n==="relu")return B(e,ju(t));throw new Error(`Cannot compute gradient for fused activation ${n}.`)}function Cf(e,t){let n=t,a=Vt(e.shape,t.shape);return a.length>0&&(n=be(n,a)),W(n,e.shape)}function Ef(e,t,n,a){if(t==="linear")return e;if(t==="relu")return Xe(e);if(t==="elu")return Wu(e);if(t==="relu6")return yf(e);if(t==="prelu")return Kc(e,n);if(t==="leakyrelu")return Gc(e,a);if(t==="sigmoid")return ma(e);throw new Error(`Unknown fused activation ${t}.`)}var Ff=(e,t)=>!(e>0)||t==="linear";function sL({x:e,filter:t,strides:n,pad:a,dataFormat:r="NHWC",dilations:s=[1,1],dimRoundingMode:i,bias:o,activation:l="linear",preluActivationWeights:u,leakyreluAlpha:p}){if(l=l||"linear",Ff(L.state.gradientDepth,l)===!1){R(r==="NHWC",()=>`Error in fused conv2d: got dataFormat of ${r} but only NHWC is currently supported for the case of gradient depth is 0 and the activation is not linear.`);let T=Rt(e,t,n,a,r,s,i);return o!=null&&(T=J(T,o)),Ef(T,l,u,p)}let d=F(e,"x","conv2d","float32"),c=F(t,"filter","conv2d","float32"),h=d,m=!1;d.rank===3&&(m=!0,h=W(d,[1,d.shape[0],d.shape[1],d.shape[2]])),R(h.rank===4,()=>`Error in fused conv2d: input must be rank 4, but got rank ${h.rank}.`),R(c.rank===4,()=>`Error in fused conv2d: filter must be rank 4, but got rank ${c.rank}.`),_n("fused conv2d",a,i);let f=r==="NHWC"?h.shape[3]:h.shape[1];R(c.shape[2]===f,()=>`Error in conv2d: depth of input (${f}) must match input depth for filter ${c.shape[2]}.`),R(mr(n,s),()=>`Error in conv2D: Either strides or dilations must be 1. Got strides ${n} and dilations '${s}'`);let g=Bc(h.shape,c.shape,n,s,a,i),y;o!=null&&(y=F(o,"bias","fused conv2d"),[y]=At(y,d),r==="NHWC"?pt(g.outShape,y.shape):(R(y.shape.length<=1,()=>`Error in fused conv2d: only supports scalar or 1-D Tensor bias for NCHW format but got the bias of rank-${y.shape.length}.`),R(y.shape.length===0||y.shape[0]===g.outChannels||y.shape[0]===1,()=>`Error in fused conv2d: bias shape (${y.shape}) is not compatible with the number of output channels (${g.outChannels})`)));let b;if(u!=null){let T=u.shape;if(R(T.length<=1||T.length===3,()=>`Error in fused conv2d: only supports scalar, 1-D Tensor or 3-D Tensor PReLU activation weights but got a tensor of rank-${T.length}.`),T.length===1)R(T[0]===1||T[0]===g.outChannels,()=>`Error in fused conv2d: PReLU activation weights (${T}) is not compatible with the number of output channels (${g.outChannels}).`);else if(T.length===3)try{pt(T,g.outShape)}catch(_){let E=`Error in fused conv2d: PReLU activation weights (${T}) is not compatible with the output shape of the conv2d (${g.outShape}).`;throw Error(E)}b=F(u,"prelu weights","fused conv2d")}let x=(T,_)=>{R(r==="NHWC",()=>`Error in gradient of fused conv2D: got dataFormat of ${r} but only NHWC is currently supported.`);let[E,A,M,$]=_,S=_f(T,M,l);R(us(s),()=>`Error in gradient of fused conv2D: dilation rates greater than 1 are not yet supported in gradients. Got dilations '${s}'`);let P=lv(A.shape,S,E,n,a),V=Lv(A,S,E.shape,n,a),j=[P,V];if($!=null){let q=Cf($,S);j.push(q)}return j},v={x:h,filter:c,bias:y,preluActivationWeights:b},k={strides:n,pad:a,dataFormat:r,dilations:s,dimRoundingMode:i,activation:l,leakyreluAlpha:p};return o==null?cr((T,_,E)=>{let A=L.runKernel(ri,v,k);return E([_,T,A]),m&&(A=W(A,[A.shape[1],A.shape[2],A.shape[3]])),{value:A,gradFunc:x}})(h,c):cr((T,_,E,A)=>{let M=L.runKernel(ri,v,k);return A([_,T,M,E]),m&&(M=W(M,[M.shape[1],M.shape[2],M.shape[3]])),{value:M,gradFunc:x}})(h,c,y)}var iL=z({fusedConv2d_:sL});function oL(e,t,n,a,r,s=[1,1],i){let o=e;e.rank===3&&(o=W(e,[1,e.shape[0],e.shape[1],e.shape[2]]));let l=t;l.rank===3&&(l=W(t,[1,t.shape[0],t.shape[1],t.shape[2]]));let u={x:o,dy:l},p={strides:a,pad:r,dimRoundingMode:i,dilations:s,filterShape:n};return L.runKernel(Am,u,p)}var c2=z({depthwiseConv2dNativeBackpropFilter_:oL});function lL(e,t,n,a,r,s=[1,1],i){let o=t,l=!1;t.rank===3&&(l=!0,o=W(t,[1,t.shape[0],t.shape[1],t.shape[2]]));let u={dy:o,filter:n},p={strides:a,pad:r,dimRoundingMode:i,dilations:s,inputShape:e},d=L.runKernel($m,u,p);return l?W(d,[d.shape[1],d.shape[2],d.shape[3]]):d}var d2=z({depthwiseConv2dNativeBackpropInput_:lL});function uL({x:e,filter:t,strides:n,pad:a,dataFormat:r="NHWC",dilations:s=[1,1],dimRoundingMode:i,bias:o,activation:l="linear",preluActivationWeights:u,leakyreluAlpha:p}){if(Ff(L.state.gradientDepth,l)===!1){let k=Ss(e,t,n,a,r,s,i);return o!=null&&(k=J(k,o)),Ef(k,l,u,p)}let d=F(e,"x","depthwiseConv2d","float32"),c=F(t,"filter","depthwiseConv2d","float32"),h=d,m=!1;d.rank===3&&(m=!0,h=W(d,[1,d.shape[0],d.shape[1],d.shape[2]])),R(h.rank===4,()=>`Error in fused depthwiseConv2d: input must be rank 4, but got rank ${h.rank}.`),R(c.rank===4,()=>`Error in fused depthwiseConv2d: filter must be rank 4, but got rank ${c.rank}.`),R(h.shape[3]===c.shape[2],()=>`Error in fused depthwiseConv2d: number of input channels (${h.shape[3]}) must match the inChannels dimension in filter ${c.shape[2]}.`),s==null&&(s=[1,1]),R(mr(n,s),()=>`Error in fused depthwiseConv2d: Either strides or dilations must be 1. Got strides ${n} and dilations '${s}'`),_n("fused depthwiseConv2d",a,i);let f=Bc(h.shape,c.shape,n,s,a,i,!0),g;o!=null&&(g=F(o,"bias","fused conv2d"),[g]=At(g,d),pt(f.outShape,g.shape));let y;u!=null&&(y=F(u,"prelu weights","fused depthwiseConv2d"));let b=(k,T)=>{R(us(s),()=>`Error in gradient of fused depthwiseConv2d: dilation rates greater than 1 are not yet supported. Got dilations '${s}'`);let[_,E,A,M]=T,$=_f(k,A,l),S=d2(E.shape,$,_,n,a,s,i),P=c2(E,$,_.shape,n,a,s,i);if(M!=null){let V=Cf(g,$);return[S,P,V]}return[S,P]},x={x:h,filter:c,bias:g,preluActivationWeights:y},v={strides:n,pad:a,dataFormat:r,dilations:s,dimRoundingMode:i,activation:l,leakyreluAlpha:p};return o==null?cr((k,T,_)=>{let E=L.runKernel(si,x,v);return _([T,k,E]),m&&(E=W(E,[E.shape[1],E.shape[2],E.shape[3]])),{value:E,gradFunc:b}})(h,c):cr((k,T,_,E)=>{let A=L.runKernel(si,x,v);return E([T,k,A,_]),m&&(A=W(A,[A.shape[1],A.shape[2],A.shape[3]])),{value:A,gradFunc:b}})(h,c,g)}var pL=z({fusedDepthwiseConv2d_:uL});function cL({a:e,b:t,transposeA:n=!1,transposeB:a=!1,bias:r,activation:s="linear",preluActivationWeights:i,leakyreluAlpha:o}){if(Ff(L.state.gradientDepth,s)===!1){let M=De(e,t,n,a);return r!=null&&(M=J(M,r)),Ef(M,s,i,o)}let l=F(e,"a","fused matMul"),u=F(t,"b","fused matMul");[l,u]=At(l,u);let p=n?l.shape[l.rank-2]:l.shape[l.rank-1],d=a?u.shape[u.rank-1]:u.shape[u.rank-2],c=n?l.shape[l.rank-1]:l.shape[l.rank-2],h=a?u.shape[u.rank-2]:u.shape[u.rank-1],m=l.shape.slice(0,-2),f=u.shape.slice(0,-2),g=bt(m),y=bt(f);R(p===d,()=>`Error in fused matMul: inner shapes (${p}) and (${d}) of Tensors with shapes ${l.shape} and ${u.shape} and transposeA=${n} and transposeB=${a} must match.`);let b=pt(l.shape.slice(0,-2),u.shape.slice(0,-2)).concat([c,h]),x=n?W(l,[g,p,c]):W(l,[g,c,p]),v=a?W(u,[y,h,d]):W(u,[y,d,h]),k;r!=null&&(k=F(r,"bias","fused matMul"),[k]=At(k,l),pt(b,k.shape));let T;i!=null&&(T=F(i,"prelu weights","fused matMul"));let _=(M,$)=>{let[S,P,V,j]=$,q=_f(W(M,V.shape),V,s),K,Z;if(!n&&!a?(K=De(q,P,!1,!0),Z=De(S,q,!0,!1)):!n&&a?(K=De(q,P,!1,!1),Z=De(q,S,!0,!1)):n&&!a?(K=De(P,q,!1,!0),Z=De(S,q,!1,!1)):(K=De(P,q,!0,!0),Z=De(q,S,!0,!0)),r!=null){let ee=Cf(j,q);return[K,Z,ee]}else return[K,Z]},E={a:x,b:v,bias:k,preluActivationWeights:T},A={transposeA:n,transposeB:a,activation:s,leakyreluAlpha:o};return r==null?cr((M,$,S)=>{let P=L.runKernel(ai,E,A);return S([M,$,P]),{value:W(P,b),gradFunc:_}})(x,v):cr((M,$,S,P)=>{let V=L.runKernel(ai,E,A);return P([M,$,V,S]),{value:W(V,b),gradFunc:_}})(x,v,k)}var dL=z({fusedMatMul_:cL});function hL(e){return Ov(e,.54,.46)}var mL=z({hammingWindow_:hL});function fL(e){return Ov(e,.5,.5)}var h2=z({hannWindow_:fL});function gL(e,t,n,a=!1,r=0){let s=0,i=[];for(;s+t<=e.size;)i.push(He(e,s,t)),s+=n;if(a)for(;s<e.size;){let o=s+t-e.size,l=Ze([He(e,s,t-o),Cn([o],r)]);i.push(l),s+=n}return i.length===0?Ha([],[0,t]):W(Ze(i),[i.length,t])}var m2=z({frame_:gL});function yL(e,t,n,a,r=h2){a==null&&(a=p2(t));let s=m2(e,t,n),i=B(s,r(t));return Yc(i,a)}var bL=z({stft_:yL});function xL(e,t,n,a,r="bilinear",s=0){let i=F(e,"image","cropAndResize"),o=F(t,"boxes","cropAndResize","float32"),l=F(n,"boxInd","cropAndResize","int32"),u=o.shape[0];R(i.rank===4,()=>`Error in cropAndResize: image must be rank 4,but got rank ${i.rank}.`),R(o.rank===2&&o.shape[1]===4,()=>`Error in cropAndResize: boxes must be have size [${u},4] but had shape ${o.shape}.`),R(l.rank===1&&l.shape[0]===u,()=>`Error in cropAndResize: boxInd must be have size [${u}] but had shape ${o.shape}.`),R(a.length===2,()=>`Error in cropAndResize: cropSize must be of length 2, but got length ${a.length}.`),R(a[0]>=1&&a[1]>=1,()=>`cropSize must be atleast [1,1], but was ${a}`),R(r==="bilinear"||r==="nearest",()=>`method must be bilinear or nearest, but was ${r}`);let p={image:i,boxes:o,boxInd:l},d={method:r,extrapolationValue:s,cropSize:a};return L.runKernel(Ql,p,d)}var vL=z({cropAndResize_:xL});function wL(e){let t=F(e,"image","flipLeftRight","float32");R(t.rank===4,()=>`Error in flipLeftRight: image must be rank 4,but got rank ${t.rank}.`);let n={image:t};return L.runKernel(au,n,{})}var kL=z({flipLeftRight_:wL});function IL(e){let t=F(e,"image","grayscaleToRGB"),n=t.rank-1,a=t.shape[n];R(t.rank>=2,()=>`Error in grayscaleToRGB: images must be at least rank 2, but got rank ${t.rank}.`),R(a===1,()=>`Error in grayscaleToRGB: last dimension of a grayscale image should be size 1, but got size ${a}.`);let r=new Array(t.rank);return r.fill(1,0,n),r[n]=3,On(t,r)}var SL=z({grayscaleToRGB_:IL});function NL(e,t,n=0,a=.5){let r=F(e,"image","rotateWithOffset","float32");R(r.rank===4,()=>`Error in rotateWithOffset: image must be rank 4,but got rank ${r.rank}.`);let s={image:r},i={radians:t,fillValue:n,center:a};return L.runKernel(zu,s,i)}var TL=z({rotateWithOffset_:NL});function qu(e,t,n,a,r,s){a==null&&(a=.5),r==null&&(r=Number.NEGATIVE_INFINITY),s==null&&(s=0);let i=e.shape[0];return n=Math.min(n,i),R(0<=a&&a<=1,()=>`iouThreshold must be in [0, 1], but was '${a}'`),R(e.rank===2,()=>`boxes must be a 2D tensor, but was of rank '${e.rank}'`),R(e.shape[1]===4,()=>`boxes must have 4 columns, but 2nd dimension was ${e.shape[1]}`),R(t.rank===1,()=>"scores must be a 1D tensor"),R(t.shape[0]===i,()=>`scores has incompatible shape with boxes. Expected ${i}, but was ${t.shape[0]}`),R(0<=s&&s<=1,()=>`softNmsSigma must be in [0, 1], but was '${s}'`),{maxOutputSize:n,iouThreshold:a,scoreThreshold:r,softNmsSigma:s}}function _L(e,t,n,a=.5,r=Number.NEGATIVE_INFINITY){let s=F(e,"boxes","nonMaxSuppression","float32"),i=F(t,"scores","nonMaxSuppression","float32"),o=qu(s,i,n,a,r);n=o.maxOutputSize,a=o.iouThreshold,r=o.scoreThreshold;let l={maxOutputSize:n,iouThreshold:a,scoreThreshold:r};return L.runKernel(yu,{boxes:s,scores:i},l)}var CL=z({nonMaxSuppression_:_L});function EL(e,t,n){let a=FL(e,t,n),r=a<0?-(a+1):a;e.splice(r,0,t)}function FL(e,t,n){return $L(e,t,n||AL)}function AL(e,t){return e>t?1:e<t?-1:0}function $L(e,t,n){let a=0,r=e.length,s=0,i=!1;for(;a<r;){s=a+(r-a>>>1);let o=n(t,e[s]);o>0?a=s+1:(r=s,i=!o)}return i?a:-a-1}function f2(e,t,n,a,r){return zv(e,t,n,a,r,0)}function g2(e,t,n,a,r,s){return zv(e,t,n,a,r,0,!1,s,!0)}function y2(e,t,n,a,r,s){return zv(e,t,n,a,r,s,!0)}function zv(e,t,n,a,r,s,i=!1,o=!1,l=!1){let u=[];for(let g=0;g<t.length;g++)t[g]>r&&u.push({score:t[g],boxIndex:g,suppressBeginIndex:0});u.sort(ak);let p=s>0?-.5/s:0,d=[],c=[];for(;d.length<n&&u.length>0;){let g=u.pop(),{score:y,boxIndex:b,suppressBeginIndex:x}=g;if(y<r)break;let v=!1;for(let k=d.length-1;k>=x;--k){let T=DL(e,b,d[k]);if(T>=a){v=!0;break}if(g.score=g.score*RL(a,p,T),g.score<=r)break}g.suppressBeginIndex=d.length,v||(g.score===y?(d.push(b),c.push(g.score)):g.score>r&&EL(u,g,ak))}let h=d.length,m=n-h;o&&m>0&&(d.push(...new Array(m).fill(0)),c.push(...new Array(m).fill(0)));let f={selectedIndices:d};return i&&(f.selectedScores=c),l&&(f.validOutputs=h),f}function DL(e,t,n){let a=e.subarray(t*4,t*4+4),r=e.subarray(n*4,n*4+4),s=Math.min(a[0],a[2]),i=Math.min(a[1],a[3]),o=Math.max(a[0],a[2]),l=Math.max(a[1],a[3]),u=Math.min(r[0],r[2]),p=Math.min(r[1],r[3]),d=Math.max(r[0],r[2]),c=Math.max(r[1],r[3]),h=(o-s)*(l-i),m=(d-u)*(c-p);if(h<=0||m<=0)return 0;let f=Math.max(s,u),g=Math.max(i,p),y=Math.min(o,d),b=Math.min(l,c),x=Math.max(y-f,0)*Math.max(b-g,0);return x/(h+m-x)}function RL(e,t,n){let a=Math.exp(t*n*n);return n<=e?a:0}function ak(e,t){return e.score-t.score||e.score===t.score&&t.boxIndex-e.boxIndex}async function ML(e,t,n,a=.5,r=Number.NEGATIVE_INFINITY){let s=F(e,"boxes","nonMaxSuppressionAsync"),i=F(t,"scores","nonMaxSuppressionAsync"),o=qu(s,i,n,a,r);n=o.maxOutputSize,a=o.iouThreshold,r=o.scoreThreshold;let l=await Promise.all([s.data(),i.data()]),u=l[0],p=l[1],{selectedIndices:d}=f2(u,p,n,a,r);return s!==e&&s.dispose(),i!==t&&i.dispose(),qe(d,"int32")}var PL=ML;function OL(e,t,n,a=.5,r=Number.NEGATIVE_INFINITY,s=0){let i=F(e,"boxes","nonMaxSuppression"),o=F(t,"scores","nonMaxSuppression"),l=qu(i,o,n,a,r,s);n=l.maxOutputSize,a=l.iouThreshold,r=l.scoreThreshold,s=l.softNmsSigma;let u={boxes:i,scores:o},p={maxOutputSize:n,iouThreshold:a,scoreThreshold:r,softNmsSigma:s},d=L.runKernel(xu,u,p);return{selectedIndices:d[0],selectedScores:d[1]}}var LL=z({nonMaxSuppressionWithScore_:OL});async function zL(e,t,n,a=.5,r=Number.NEGATIVE_INFINITY,s=0){let i=F(e,"boxes","nonMaxSuppressionAsync"),o=F(t,"scores","nonMaxSuppressionAsync"),l=qu(i,o,n,a,r,s);n=l.maxOutputSize,a=l.iouThreshold,r=l.scoreThreshold,s=l.softNmsSigma;let u=await Promise.all([i.data(),o.data()]),p=u[0],d=u[1],{selectedIndices:c,selectedScores:h}=y2(p,d,n,a,r,s);return i!==e&&i.dispose(),o!==t&&o.dispose(),{selectedIndices:qe(c,"int32"),selectedScores:qe(h)}}var BL=zL;function WL(e,t,n,a=.5,r=Number.NEGATIVE_INFINITY,s=!1){let i=F(e,"boxes","nonMaxSuppression"),o=F(t,"scores","nonMaxSuppression"),l=qu(i,o,n,a,r,null),u=l.maxOutputSize,p=l.iouThreshold,d=l.scoreThreshold,c={boxes:i,scores:o},h={maxOutputSize:u,iouThreshold:p,scoreThreshold:d,padToMaxOutputSize:s},m=L.runKernel(bu,c,h);return{selectedIndices:m[0],validOutputs:m[1]}}var VL=z({nonMaxSuppressionPadded_:WL});async function UL(e,t,n,a=.5,r=Number.NEGATIVE_INFINITY,s=!1){let i=F(e,"boxes","nonMaxSuppressionAsync"),o=F(t,"scores","nonMaxSuppressionAsync"),l=qu(i,o,n,a,r,null),u=l.maxOutputSize,p=l.iouThreshold,d=l.scoreThreshold,[c,h]=await Promise.all([i.data(),o.data()]),{selectedIndices:m,validOutputs:f}=g2(c,h,u,p,d,s);return i!==e&&i.dispose(),o!==t&&o.dispose(),{selectedIndices:qe(m,"int32"),validOutputs:we(f,"int32")}}var GL=UL;function HL(e,t,n=!1,a=!1){let r=F(e,"images","resizeBilinear");R(r.rank===3||r.rank===4,()=>`Error in resizeBilinear: x must be rank 3 or 4, but got rank ${r.rank}.`),R(t.length===2,()=>`Error in resizeBilinear: new shape must 2D, but got shape ${t}.`),R(a===!1||n===!1,()=>"Error in resizeBilinear: If halfPixelCenters is true, alignCorners must be false.");let s=r,i=!1;r.rank===3&&(i=!0,s=W(r,[1,r.shape[0],r.shape[1],r.shape[2]]));let[]=t,o={images:s},l={alignCorners:n,halfPixelCenters:a,size:t},u=L.runKernel(ao,o,l);return i?W(u,[u.shape[1],u.shape[2],u.shape[3]]):u}var b2=z({resizeBilinear_:HL});function jL(e,t,n=!1,a=!1){let r=F(e,"images","resizeNearestNeighbor");R(r.rank===3||r.rank===4,()=>`Error in resizeNearestNeighbor: x must be rank 3 or 4, but got rank ${r.rank}.`),R(t.length===2,()=>`Error in resizeNearestNeighbor: new shape must 2D, but got shape ${t}.`),R(r.dtype==="float32"||r.dtype==="int32",()=>"`images` must have `int32` or `float32` as dtype"),R(a===!1||n===!1,()=>"Error in resizeNearestNeighbor: If halfPixelCenters is true, alignCorners must be false.");let s=r,i=!1;r.rank===3&&(i=!0,s=W(r,[1,r.shape[0],r.shape[1],r.shape[2]]));let[]=t,o={images:s},l={alignCorners:n,halfPixelCenters:a,size:t},u=L.runKernel(Cc,o,l);return i?W(u,[u.shape[1],u.shape[2],u.shape[3]]):u}var x2=z({resizeNearestNeighbor_:jL});function qL(e,t="binary",n=!1,a=.5){let r=F(e,"image","threshold"),s=.2989,i=.587,o=.114,l=r.shape[0]*r.shape[1],u=B(qe([a]),255),p,d,c,h;if(R(r.rank===3,()=>`Error in threshold: image must be rank 3,but got rank ${r.rank}.`),R(r.shape[2]===3||r.shape[2]===1,()=>`Error in threshold: image color channel must be equal to 3 or 1but got ${r.shape[2]}.`),R(r.dtype==="int32"||r.dtype==="float32",()=>`Error in dtype: image dtype must be int32 or float32,but got dtype ${r.dtype}.`),R(t==="otsu"||t==="binary",()=>`Method must be binary or otsu, but was ${t}`),r.shape[2]===3){[p,d,c]=zn(r,[1,1,1],-1);let f=B(p,s),g=B(d,i),y=B(c,o);h=J(J(f,g),y)}else h=e;if(t==="otsu"){let f=iv(oe(bf(h),"int32"),Jn([]),256);u=KL(f,l)}let m=n?Ts(h,u):Un(h,u);return oe(B(m,255),"int32")}function KL(e,t){let n=qe([-1]),a=qe([0]),r=qe([0]),s,i,o,l,u,p;for(let d=0;d<e.size-1;d++){s=He(e,0,d+1),i=He(e,d+1),u=fe(be(s),t),p=fe(be(i),t);let c=be(B(s,Tl(0,s.size)));o=fe(c,be(s));let h=Cn(i.shape,s.size),m=J(Tl(0,i.size),h),f=B(i,m);l=fe(be(f),be(i));let g=ce(o,l),y=ce(o,l),b=B(u,p);r=B(B(b,g),y);let x=Un(r,a);a=fn(x,r,a),n=fn(x,qe([d]),n)}return n}var XL=z({threshold_:qL});function YL(e,t,n="nearest",a="constant",r=0,s){let i=F(e,"image","transform","float32"),o=F(t,"transforms","transform","float32");R(i.rank===4,()=>`Error in transform: image must be rank 4,but got rank ${i.rank}.`),R(o.rank===2&&(o.shape[0]===i.shape[0]||o.shape[0]===1)&&o.shape[1]===8,()=>"Error in transform: Input transform should be batch x 8 or 1 x 8"),R(s==null||s.length===2,()=>`Error in transform: outputShape must be [height, width] or null, but got ${s}.`);let l={image:i,transforms:o},u={interpolation:n,fillMode:a,fillValue:r,outputShape:s};return L.runKernel(Pu,l,u)}var QL=z({transform_:YL});function JL(e,t,n){R(t%1===0,()=>`bandPart(): numLower must be an integer, got ${t}.`),R(n%1===0,()=>`bandPart(): numUpper must be an integer, got ${n}.`);let a=F(e,"a","bandPart");R(a.rank>=2,()=>`bandPart(): Rank must be at least 2, got ${a.rank}.`);let r=a.shape,[s,i]=a.shape.slice(-2);if(!(t<=s))throw new Error(`bandPart(): numLower (${t}) must not be greater than the number of rows (${s}).`);if(!(n<=i))throw new Error(`bandPart(): numUpper (${n}) must not be greater than the number of columns (${i}).`);t<0&&(t=s),n<0&&(n=i);let o=W(Tl(0,s,1,"int32"),[-1,1]),l=Tl(0,i,1,"int32"),u=ce(o,l),p=Ca(Ts(u,we(+t,"int32")),Ns(u,we(-n,"int32"))),d=St([s,i],a.dtype);return W(Mt(ht(W(a,[-1,s,i])).map(c=>fn(p,c,d))),r)}var ZL=z({bandPart_:JL});function ez(e){let t;if(Array.isArray(e)){t=!1,R(e!=null&&e.length>0,()=>"Gram-Schmidt process: input must not be null, undefined, or empty");let r=e[0].shape[0];for(let s=1;s<e.length;++s)R(e[s].shape[0]===r,()=>`Gram-Schmidt: Non-unique lengths found in the input vectors: (${e[s].shape[0]} vs. ${r})`)}else t=!0,e=zn(e,e.shape[0],0).map(r=>dr(r,[0]));R(e.length<=e[0].shape[0],()=>`Gram-Schmidt: Number of vectors (${e.length}) exceeds number of dimensions (${e[0].shape[0]}).`);let n=[],a=e;for(let r=0;r<e.length;++r)n.push(L.tidy(()=>{let s=a[r];if(r>0)for(let i=0;i<r;++i){let o=B(be(B(n[i],s)),n[i]);s=ce(s,o)}return fe(s,Uc(s,"euclidean"))}));return t?Mt(n,0):n}var tz=z({gramSchmidt_:ez});function nz(e,t=!1){if(R(e.rank>=2,()=>`qr() requires input tensor to have a rank >= 2, but got rank ${e.rank}`),e.rank===2)return rk(e,t);{let n=e.shape.slice(0,e.shape.length-2).reduce((l,u)=>l*u),a=ht(W(e,[n,e.shape[e.shape.length-2],e.shape[e.shape.length-1]]),0),r=[],s=[];a.forEach(l=>{let[u,p]=rk(l,t);r.push(u),s.push(p)});let i=W(Mt(r,0),e.shape),o=W(Mt(s,0),e.shape);return[i,o]}}function rk(e,t=!1){return L.tidy(()=>{R(e.shape.length===2,()=>`qr2d() requires a 2D Tensor, but got a ${e.shape.length}D Tensor.`);let n=e.shape[0],a=e.shape[1],r=bv(n),s=Er(e),i=Ha([[1]],[1,1]),o=Er(i),l=n>=a?a:n;for(let u=0;u<l;++u){let p=s,d=o,c=r;[o,s,r]=L.tidy(()=>{let h=He(s,[u,u],[n-u,1]),m=Uc(h),f=He(s,[u,u],[1,1]),g=fn(Un(f,0),Ha([[-1]]),Ha([[1]])),y=ce(f,B(g,m)),b=fe(h,y);b.shape[0]===1?o=Er(i):o=Ze([i,He(b,[1,0],[b.shape[0]-1,b.shape[1]])],0);let x=It(fe(De(g,y),m)),v=He(s,[u,0],[n-u,a]),k=B(x,o),T=Ae(o);if(u===0)s=ce(v,De(k,De(T,v)));else{let A=ce(v,De(k,De(T,v)));s=Ze([He(s,[0,0],[u,a]),A],0)}let _=Ae(k),E=He(r,[0,u],[n,r.shape[1]-u]);if(u===0)r=ce(E,De(De(E,o),_));else{let A=ce(E,De(De(E,o),_));r=Ze([He(r,[0,0],[n,u]),A],1)}return[o,s,r]}),Re([p,d,c])}return!t&&n>a&&(r=He(r,[0,0],[n,a]),s=He(s,[0,0],[a,a])),[r,s]})}var az=z({qr_:nz}),In;(function(e){e[e.NONE=0]="NONE",e[e.MEAN=1]="MEAN",e[e.SUM=2]="SUM",e[e.SUM_BY_NONZERO_WEIGHTS=3]="SUM_BY_NONZERO_WEIGHTS"})(In||(In={}));function rz(e,t,n=In.SUM_BY_NONZERO_WEIGHTS){let a=F(e,"losses","computeWeightedLoss"),r=null;t!=null&&(r=F(t,"weights","computeWeightedLoss"));let s=r==null?a:B(a,r);if(n===In.NONE)return s;if(n===In.SUM)return be(s);if(n===In.MEAN){if(r==null)return Et(s);{let i=a.size/r.size,o=fe(be(s),be(r));return i>1?fe(o,we(i)):o}}if(n===In.SUM_BY_NONZERO_WEIGHTS){if(r==null)return fe(be(s),we(a.size));{let i=B(r,Qn(a.shape)),o=oe(be(di(i,we(0))),"float32");return fe(be(s),o)}}throw Error(`Unknown reduction: ${n}`)}var Rr=z({computeWeightedLoss_:rz});function sz(e,t,n,a=In.SUM_BY_NONZERO_WEIGHTS){let r=F(e,"labels","absoluteDifference"),s=F(t,"predictions","absoluteDifference"),i=null;n!=null&&(i=F(n,"weights","absoluteDifference")),Tn(r.shape,s.shape,"Error in absoluteDifference: ");let o=zt(ce(r,s));return Rr(o,i,a)}var iz=z({absoluteDifference_:sz});function oz(e,t,n,a,r=In.SUM_BY_NONZERO_WEIGHTS){let s=F(e,"labels","cosineDistance"),i=F(t,"predictions","cosineDistance"),o=null;a!=null&&(o=F(a,"weights","cosineDistance")),Tn(s.shape,i.shape,"Error in cosineDistance: ");let l=we(1),u=ce(l,be(B(s,i),n,!0));return Rr(u,o,r)}var lz=z({cosineDistance_:oz});function uz(e,t,n,a=In.SUM_BY_NONZERO_WEIGHTS){let r=F(e,"labels","hingeLoss"),s=F(t,"predictions","hingeLoss"),i=null;n!=null&&(i=F(n,"weights","hingeLoss")),Tn(r.shape,s.shape,"Error in hingeLoss: ");let o=we(1);r=ce(B(we(2),r),o);let l=Xe(ce(o,B(r,s)));return Rr(l,i,a)}var pz=z({hingeLoss_:uz});function cz(e,t,n,a=1,r=In.SUM_BY_NONZERO_WEIGHTS){let s=F(e,"labels","huberLoss"),i=F(t,"predictions","huberLoss"),o=null;n!=null&&(o=F(n,"weights","huberLoss")),Tn(s.shape,i.shape,"Error in huberLoss: ");let l=we(a),u=zt(ce(i,s)),p=Uu(u,l),d=ce(u,p),c=J(B(we(.5),ut(p)),B(l,d));return Rr(c,o,r)}var dz=z({huberLoss_:cz});function hz(e,t,n,a=1e-7,r=In.SUM_BY_NONZERO_WEIGHTS){let s=F(e,"labels","logLoss"),i=F(t,"predictions","logLoss"),o=null;n!=null&&(o=F(n,"weights","logLoss")),Tn(s.shape,i.shape,"Error in logLoss: ");let l=we(1),u=we(a),p=It(B(s,ea(J(i,u)))),d=B(ce(l,s),ea(J(ce(l,i),u))),c=ce(p,d);return Rr(c,o,r)}var mz=z({logLoss_:hz});function fz(e,t,n,a=In.SUM_BY_NONZERO_WEIGHTS){let r=F(e,"labels","meanSquaredError"),s=F(t,"predictions","meanSquaredError"),i=null;n!=null&&(i=F(n,"weights","meanSquaredError")),Tn(r.shape,s.shape,"Error in meanSquaredError: ");let o=Nf(r,s);return Rr(o,i,a)}var gz=z({meanSquaredError_:fz});function yz(e,t){let n=F(e,"labels","sigmoidCrossEntropyWithLogits"),a=F(t,"logits","sigmoidCrossEntropyWithLogits");Tn(n.shape,a.shape,"Error in sigmoidCrossEntropyWithLogits: ");let r=Xe(a),s=B(a,n),i=Hc(gn(It(zt(a))));return J(ce(r,s),i)}function bz(e,t,n,a=0,r=In.SUM_BY_NONZERO_WEIGHTS){let s=F(e,"multiClassLabels","sigmoidCrossEntropy"),i=F(t,"logits","sigmoidCrossEntropy"),o=null;if(n!=null&&(o=F(n,"weights","sigmoidCrossEntropy")),Tn(s.shape,i.shape,"Error in sigmoidCrossEntropy: "),a>0){let u=we(a),p=we(1),d=we(.5);s=J(B(s,ce(p,u)),B(d,u))}let l=yz(s,i);return Rr(l,o,r)}var xz=z({sigmoidCrossEntropy_:bz});function vz(e,t,n=-1){if(n===-1&&(n=t.rank-1),n!==t.rank-1)throw Error(`Softmax cross entropy along a non-last dimension is not yet supported. Labels / logits was rank ${t.rank} and dim was ${n}`);return cr((a,r,s)=>{let i=wv(r,[n],!0),o=ce(oe(r,"float32"),i);s([a,o]);let l=It(B(o,a));return{value:be(l,[n]),gradFunc:(u,p)=>{let[d,c]=p,h=pi(u.shape,[n]);return[B(W(u,h),ce(oe(d,"float32"),gn(c))),B(W(u,h),ce(gn(c),oe(d,"float32")))]}}})(e,t)}function wz(e,t,n,a=0,r=In.SUM_BY_NONZERO_WEIGHTS){let s=F(e,"onehotLabels","softmaxCrossEntropy"),i=F(t,"logits","softmaxCrossEntropy"),o=null;if(n!=null&&(o=F(n,"weights","softmaxCrossEntropy")),Tn(s.shape,i.shape,"Error in softmaxCrossEntropy: "),a>0){let u=we(a),p=we(1),d=we(s.shape[1]);s=J(B(s,ce(p,u)),fe(u,d))}let l=vz(s,i);return Rr(l,o,r)}var kz=z({softmaxCrossEntropy_:wz});function Iz(e,t,n,a){let r=F(e,"indices","sparseFillEmptyRows","int32"),s=F(t,"values","sparseFillEmptyRows"),i=F(n,"denseShape","sparseFillEmptyRows","int32"),o=F(a,"defaultValue","sparseFillEmptyRows",s.dtype);if(r.rank!==2)throw new Error(`Indices should be Tensor2D but received shape
${r.shape}`);if(s.rank!==1)throw new Error(`Values should be Tensor1D but received shape ${s.shape}`);if(i.rank!==1)throw new Error(`Dense shape should be Tensor1D but received shape ${i.shape}`);if(o.rank!==0)throw new Error(`Default value should be a scalar but received shape ${o.shape}`);let l={indices:r,values:s,denseShape:i,defaultValue:o},u=L.runKernel(Ec,l);return{outputIndices:u[0],outputValues:u[1],emptyRowIndicator:u[2],reverseIndexMap:u[3]}}var Sz=z({sparseFillEmptyRows_:Iz});function Nz(e,t,n){let a=F(e,"inputIndices","sparseReshape","int32"),r=F(t,"inputShape","sparseReshape","int32"),s=F(n,"newShape","sparseReshape","int32");if(a.rank!==2)throw new Error(`Input indices should be Tensor2D but received shape
${a.shape}`);if(r.rank!==1)throw new Error(`Input shape should be Tensor1D but received shape ${r.shape}`);if(s.rank!==1)throw new Error(`New shape should be Tensor1D but received shape ${s.shape}`);let i={inputIndices:a,inputShape:r,newShape:s},o=L.runKernel(Du,i);return{outputIndices:o[0],outputShape:o[1]}}var Tz=z({sparseReshape_:Nz});function _z(e,t,n){let a=F(e,"data","sparseSegmentMean"),r=F(t,"indices","sparseSegmentMean","int32"),s=F(n,"segmentIds","sparseSegmentMean","int32");if(a.rank<1)throw new Error("Data should be at least 1 dimensional but received scalar");if(r.rank!==1)throw new Error(`Indices should be Tensor1D but received shape
${r.shape}`);if(s.rank!==1)throw new Error(`Segment ids should be Tensor1D but received shape
${s.shape}`);let i={data:a,indices:r,segmentIds:s};return L.runKernel(Fc,i)}var Cz=z({sparseSegmentMean_:_z});function Ez(e,t,n){let a=F(e,"data","sparseSegmentSum"),r=F(t,"indices","sparseSegmentSum","int32"),s=F(n,"segmentIds","sparseSegmentSum","int32");if(a.rank<1)throw new Error("Data should be at least 1 dimensional but received scalar");if(r.rank!==1)throw new Error(`Indices should be Tensor1D but received shape
${r.shape}`);if(s.rank!==1)throw new Error(`Segment ids should be Tensor1D but received shape
${s.shape}`);let i={data:a,indices:r,segmentIds:s};return L.runKernel(Ac,i)}var Fz=z({sparseSegmentSum_:Ez});function Az(e,t,n,a,r,s,i,o){let l=F(e,"data","stringNGrams","string");if(l.dtype!=="string")throw new Error("Data must be of datatype string");if(l.shape.length!==1)throw new Error(`Data must be a vector, saw: ${l.shape}`);let u=F(t,"dataSplits","stringNGrams");if(u.dtype!=="int32")throw new Error("Data splits must be of datatype int32");let p={separator:n,nGramWidths:a,leftPad:r,rightPad:s,padWidth:i,preserveShortSequences:o},d={data:l,dataSplits:u},c=L.runKernel(Ym,d,p);return{nGrams:c[0],nGramsSplits:c[1]}}var $z=z({stringNGrams_:Az});function Dz(e,t,n=!0){let a=F(e,"input","stringSplit","string"),r=F(t,"delimiter","stringSplit","string");if(a.rank!==1)throw new Error(`Input should be Tensor1D but received shape ${a.shape}`);if(r.rank!==0)throw new Error(`Delimiter should be a scalar but received shape ${r.shape}`);let s={skipEmpty:n},i={input:a,delimiter:r},o=L.runKernel(Qm,i,s);return{indices:o[0],values:o[1],shape:o[2]}}var Rz=z({stringSplit_:Dz});function Mz(e,t){let n=F(e,"input","stringToHashBucketFast","string"),a={numBuckets:t};if(t<=0)throw new Error("Number of buckets must be at least 1");let r={input:n};return L.runKernel(Jm,r,a)}var Pz=z({stringToHashBucketFast_:Mz}),Oz={fft:Xc,ifft:_l,rfft:Yc,irfft:Sf},Lz={hammingWindow:mL,hannWindow:h2,frame:m2,stft:bL},Ln={flipLeftRight:kL,grayscaleToRGB:SL,resizeNearestNeighbor:x2,resizeBilinear:b2,rotateWithOffset:TL,cropAndResize:vL,nonMaxSuppression:CL,nonMaxSuppressionAsync:PL,nonMaxSuppressionWithScore:LL,nonMaxSuppressionWithScoreAsync:BL,nonMaxSuppressionPadded:VL,nonMaxSuppressionPaddedAsync:GL,threshold:XL,transform:QL},v2={bandPart:ZL,gramSchmidt:tz,qr:az},zz={absoluteDifference:iz,computeWeightedLoss:Rr,cosineDistance:lz,hingeLoss:pz,huberLoss:dz,logLoss:mz,meanSquaredError:gz,sigmoidCrossEntropy:xz,softmaxCrossEntropy:kz},Wp={sparseFillEmptyRows:Sz,sparseReshape:Tz,sparseSegmentMean:Cz,sparseSegmentSum:Fz},$h={stringNGrams:$z,stringSplit:Rz,stringToHashBucketFast:Pz},Mr=class extends wS{minimize(e,t=!1,n){let{value:a,grads:r}=this.computeGradients(e,n);if(n!=null){let s=n.map(i=>({name:i.name,tensor:r[i.name]}));this.applyGradients(s)}else this.applyGradients(r);return Re(r),t?a:(a.dispose(),null)}get iterations(){return this.iterations_==null&&(this.iterations_=0),this.iterations_}incrementIterations(){this.iterations_=this.iterations+1}computeGradients(e,t){return XS(e,t)}dispose(){this.iterations_!=null&&Re(this.iterations_)}async saveIterations(){return this.iterations_==null&&(this.iterations_=0),{name:"iter",tensor:we(this.iterations_,"int32")}}async getWeights(){throw new Error("getWeights() is not implemented for this optimizer yet.")}async setWeights(e){throw new Error(`setWeights() is not implemented for this optimizer class ${this.getClassName()}`)}async extractIterations(e){return this.iterations_=(await e[0].tensor.data())[0],e.slice(1)}};Object.defineProperty(Mr,Symbol.hasInstance,{value:e=>e.minimize!=null&&e.computeGradients!=null&&e.applyGradients!=null});var Af=class extends Mr{constructor(e,t,n=null){super(),this.learningRate=e,this.rho=t,this.epsilon=n,this.accumulatedGrads=[],this.accumulatedUpdates=[],n==null&&(this.epsilon=L.backend.epsilon())}applyGradients(e){(Array.isArray(e)?e.map(t=>t.name):Object.keys(e)).forEach((t,n)=>{let a=L.registeredVariables[t],r=!1;this.accumulatedGrads[n]==null&&(this.accumulatedGrads[n]={originalName:`${t}/accum_grad`,variable:O(()=>Ke(a).variable(r))}),this.accumulatedUpdates[n]==null&&(this.accumulatedUpdates[n]={originalName:`${t}/accum_var`,variable:O(()=>Ke(a).variable(r))});let s=Array.isArray(e)?e[n].tensor:e[t];if(s==null)return;let i=this.accumulatedGrads[n].variable,o=this.accumulatedUpdates[n].variable;O(()=>{let l=J(B(i,this.rho),B(ut(s),1-this.rho)),u=B(fe(pn(J(o,this.epsilon)),pn(J(i,this.epsilon))),s),p=J(B(o,this.rho),B(ut(u),1-this.rho));i.assign(l),o.assign(p);let d=J(B(u,-this.learningRate),a);a.assign(d)})}),this.incrementIterations()}dispose(){this.accumulatedUpdates!=null&&(Re(this.accumulatedGrads.map(e=>e.variable)),Re(this.accumulatedUpdates.map(e=>e.variable)))}async getWeights(){let e=[...this.accumulatedGrads,...this.accumulatedUpdates];return[await this.saveIterations()].concat(e.map(t=>({name:t.originalName,tensor:t.variable})))}async setWeights(e){e=await this.extractIterations(e);let t=e.length/2,n=!1;this.accumulatedGrads=e.slice(0,t).map(a=>({originalName:a.name,variable:a.tensor.variable(n)})),this.accumulatedUpdates=e.slice(t,t*2).map(a=>({originalName:a.name,variable:a.tensor.variable(n)}))}getConfig(){return{learningRate:this.learningRate,rho:this.rho,epsilon:this.epsilon}}static fromConfig(e,t){return new e(t.learningRate,t.rho,t.epsilon)}};Af.className="Adadelta";Is(Af);var $f=class extends Mr{constructor(e,t=.1){super(),this.learningRate=e,this.initialAccumulatorValue=t,this.accumulatedGrads=[]}applyGradients(e){(Array.isArray(e)?e.map(t=>t.name):Object.keys(e)).forEach((t,n)=>{let a=L.registeredVariables[t];this.accumulatedGrads[n]==null&&(this.accumulatedGrads[n]={originalName:`${t}/accumulator`,variable:O(()=>Cn(a.shape,this.initialAccumulatorValue).variable(!1))});let r=Array.isArray(e)?e[n].tensor:e[t];if(r==null)return;let s=this.accumulatedGrads[n].variable;O(()=>{let i=J(s,ut(r));s.assign(i);let o=J(B(fe(r,pn(J(i,L.backend.epsilon()))),-this.learningRate),a);a.assign(o)})}),this.incrementIterations()}dispose(){this.accumulatedGrads!=null&&Re(this.accumulatedGrads.map(e=>e.variable))}async getWeights(){return[await this.saveIterations()].concat(this.accumulatedGrads.map(e=>({name:e.originalName,tensor:e.variable})))}async setWeights(e){e=await this.extractIterations(e);let t=!1;this.accumulatedGrads=e.map(n=>({originalName:n.name,variable:n.tensor.variable(t)}))}getConfig(){return{learningRate:this.learningRate,initialAccumulatorValue:this.initialAccumulatorValue}}static fromConfig(e,t){return new e(t.learningRate,t.initialAccumulatorValue)}};$f.className="Adagrad";Is($f);var Df=class extends Mr{constructor(e,t,n,a=null){super(),this.learningRate=e,this.beta1=t,this.beta2=n,this.epsilon=a,this.accumulatedFirstMoment=[],this.accumulatedSecondMoment=[],O(()=>{this.accBeta1=we(t).variable(),this.accBeta2=we(n).variable()}),a==null&&(this.epsilon=L.backend.epsilon())}applyGradients(e){let t=Array.isArray(e)?e.map(n=>n.name):Object.keys(e);O(()=>{let n=ce(1,this.accBeta1),a=ce(1,this.accBeta2);t.forEach((r,s)=>{let i=L.registeredVariables[r],o=!1;this.accumulatedFirstMoment[s]==null&&(this.accumulatedFirstMoment[s]={originalName:`${r}/m`,variable:O(()=>Ke(i).variable(o))}),this.accumulatedSecondMoment[s]==null&&(this.accumulatedSecondMoment[s]={originalName:`${r}/v`,variable:O(()=>Ke(i).variable(o))});let l=Array.isArray(e)?e[s].tensor:e[r];if(l==null)return;let u=this.accumulatedFirstMoment[s].variable,p=this.accumulatedSecondMoment[s].variable,d=J(B(u,this.beta1),B(l,1-this.beta1)),c=J(B(p,this.beta2),B(ut(l),1-this.beta2)),h=fe(d,n),m=fe(c,a);u.assign(d),p.assign(c);let f=J(B(fe(h,J(pn(m),this.epsilon)),-this.learningRate),i);i.assign(f)}),this.accBeta1.assign(B(this.accBeta1,this.beta1)),this.accBeta2.assign(B(this.accBeta2,this.beta2))}),this.incrementIterations()}dispose(){this.accBeta1.dispose(),this.accBeta2.dispose(),this.accumulatedFirstMoment!=null&&Re(this.accumulatedFirstMoment.map(e=>e.variable)),this.accumulatedSecondMoment!=null&&Re(this.accumulatedSecondMoment.map(e=>e.variable))}async getWeights(){let e=[...this.accumulatedFirstMoment,...this.accumulatedSecondMoment];return[await this.saveIterations()].concat(e.map(t=>({name:t.originalName,tensor:t.variable})))}async setWeights(e){e=await this.extractIterations(e),O(()=>{this.accBeta1.assign(Dr(this.beta1,this.iterations_+1)),this.accBeta2.assign(Dr(this.beta2,this.iterations_+1))});let t=e.length/2,n=!1;this.accumulatedFirstMoment=e.slice(0,t).map(a=>({originalName:a.name,variable:a.tensor.variable(n)})),this.accumulatedSecondMoment=e.slice(t,t*2).map(a=>({originalName:a.name,variable:a.tensor.variable(n)}))}getConfig(){return{learningRate:this.learningRate,beta1:this.beta1,beta2:this.beta2,epsilon:this.epsilon}}static fromConfig(e,t){return new e(t.learningRate,t.beta1,t.beta2,t.epsilon)}};Df.className="Adam";Is(Df);var Rf=class extends Mr{constructor(e,t,n,a=null,r=0){super(),this.learningRate=e,this.beta1=t,this.beta2=n,this.epsilon=a,this.decay=r,this.accumulatedFirstMoment=[],this.accumulatedWeightedInfNorm=[],O(()=>{this.iteration=we(0).variable(),this.accBeta1=we(t).variable()}),a==null&&(this.epsilon=L.backend.epsilon())}applyGradients(e){let t=Array.isArray(e)?e.map(n=>n.name):Object.keys(e);O(()=>{let n=ce(1,this.accBeta1),a=fe(-this.learningRate,J(B(this.iteration,this.decay),1));t.forEach((r,s)=>{let i=L.registeredVariables[r],o=!1;this.accumulatedFirstMoment[s]==null&&(this.accumulatedFirstMoment[s]={originalName:`${r}/m`,variable:Ke(i).variable(o)}),this.accumulatedWeightedInfNorm[s]==null&&(this.accumulatedWeightedInfNorm[s]={originalName:`${r}/v`,variable:Ke(i).variable(o)});let l=Array.isArray(e)?e[s].tensor:e[r];if(l==null)return;let u=this.accumulatedFirstMoment[s].variable,p=this.accumulatedWeightedInfNorm[s].variable,d=J(B(u,this.beta1),B(l,1-this.beta1)),c=B(p,this.beta2),h=zt(l),m=fr(c,h);u.assign(d),p.assign(m);let f=J(B(fe(a,n),fe(d,J(m,this.epsilon))),i);i.assign(f)}),this.iteration.assign(J(this.iteration,1)),this.accBeta1.assign(B(this.accBeta1,this.beta1))}),this.incrementIterations()}dispose(){this.accBeta1.dispose(),this.iteration.dispose(),this.accumulatedFirstMoment!=null&&Re(this.accumulatedFirstMoment.map(e=>e.variable)),this.accumulatedWeightedInfNorm!=null&&Re(this.accumulatedWeightedInfNorm.map(e=>e.variable))}async getWeights(){throw new Error("getWeights() is not implemented for Adamax yet.")}async setWeights(e){throw new Error("setWeights() is not implemented for Adamax yet.")}getConfig(){return{learningRate:this.learningRate,beta1:this.beta1,beta2:this.beta2,epsilon:this.epsilon,decay:this.decay}}static fromConfig(e,t){return new e(t.learningRate,t.beta1,t.beta2,t.epsilon,t.decay)}};Rf.className="Adamax";Is(Rf);var Qc=class extends Mr{constructor(e){super(),this.learningRate=e,this.setLearningRate(e)}applyGradients(e){(Array.isArray(e)?e.map(t=>t.name):Object.keys(e)).forEach((t,n)=>{let a=Array.isArray(e)?e[n].tensor:e[t];if(a==null)return;let r=L.registeredVariables[t];O(()=>{let s=J(B(this.c,a),r);r.assign(s)})}),this.incrementIterations()}setLearningRate(e){this.learningRate=e,this.c!=null&&this.c.dispose(),this.c=tn(we(-e))}dispose(){this.c.dispose()}async getWeights(){return[await this.saveIterations()]}async setWeights(e){if(e=await this.extractIterations(e),e.length!==0)throw new Error("SGD optimizer does not have settable weights.")}getConfig(){return{learningRate:this.learningRate}}static fromConfig(e,t){return new e(t.learningRate)}};Qc.className="SGD";Is(Qc);var Mf=class extends Qc{constructor(e,t,n=!1){super(e),this.learningRate=e,this.momentum=t,this.useNesterov=n,this.accumulations=[],this.m=we(this.momentum)}applyGradients(e){(Array.isArray(e)?e.map(t=>t.name):Object.keys(e)).forEach((t,n)=>{let a=L.registeredVariables[t];this.accumulations[n]==null&&(this.accumulations[n]={originalName:`${t}/momentum`,variable:O(()=>Ke(a).variable(!1))});let r=this.accumulations[n].variable,s=Array.isArray(e)?e[n].tensor:e[t];s!=null&&O(()=>{let i,o=J(B(this.m,r),s);this.useNesterov?i=J(B(this.c,J(s,B(o,this.m))),a):i=J(B(this.c,o),a),r.assign(o),a.assign(i)})}),this.incrementIterations()}dispose(){this.m.dispose(),this.accumulations!=null&&Re(this.accumulations.map(e=>e.variable))}setMomentum(e){this.momentum=e}async getWeights(){return[await this.saveIterations()].concat(this.accumulations.map(e=>({name:e.originalName,tensor:e.variable})))}async setWeights(e){e=await this.extractIterations(e);let t=!1;this.accumulations=e.map(n=>({originalName:n.name,variable:n.tensor.variable(t)}))}getConfig(){return{learningRate:this.learningRate,momentum:this.momentum,useNesterov:this.useNesterov}}static fromConfig(e,t){return new e(t.learningRate,t.momentum,t.useNesterov)}};Mf.className="Momentum";Is(Mf);var Pf=class extends Mr{constructor(e,t=.9,n=0,a=null,r=!1){if(super(),this.learningRate=e,this.decay=t,this.momentum=n,this.epsilon=a,this.accumulatedMeanSquares=[],this.accumulatedMoments=[],this.accumulatedMeanGrads=[],this.centered=r,a==null&&(this.epsilon=L.backend.epsilon()),e==null)throw new Error("learningRate for RMSPropOptimizer must be defined.")}applyGradients(e){(Array.isArray(e)?e.map(t=>t.name):Object.keys(e)).forEach((t,n)=>{let a=L.registeredVariables[t],r=!1;this.accumulatedMeanSquares[n]==null&&(this.accumulatedMeanSquares[n]={originalName:`${t}/rms`,variable:O(()=>Ke(a).variable(r))}),this.accumulatedMoments[n]==null&&(this.accumulatedMoments[n]={originalName:`${t}/momentum`,variable:O(()=>Ke(a).variable(r))}),this.accumulatedMeanGrads[n]==null&&this.centered&&(this.accumulatedMeanGrads[n]={originalName:`${t}/mg`,variable:O(()=>Ke(a).variable(r))});let s=Array.isArray(e)?e[n].tensor:e[t];if(s==null)return;let i=this.accumulatedMeanSquares[n].variable,o=this.accumulatedMoments[n].variable;O(()=>{let l=J(B(i,this.decay),B(ut(s),1-this.decay));if(this.centered){let u=this.accumulatedMeanGrads[n].variable,p=J(B(u,this.decay),B(s,1-this.decay)),d=fe(B(s,this.learningRate),pn(ce(l,J(ut(p),this.epsilon)))),c=J(B(o,this.momentum),d);i.assign(l),u.assign(p),o.assign(c);let h=ce(a,c);a.assign(h)}else{let u=J(B(i,this.decay),B(ut(s),1-this.decay)),p=J(B(o,this.momentum),fe(B(s,this.learningRate),pn(J(u,this.epsilon))));i.assign(u),o.assign(p);let d=ce(a,p);a.assign(d)}})}),this.incrementIterations()}dispose(){this.accumulatedMeanSquares!=null&&Re(this.accumulatedMeanSquares.map(e=>e.variable)),this.accumulatedMeanGrads!=null&&this.centered&&Re(this.accumulatedMeanGrads.map(e=>e.variable)),this.accumulatedMoments!=null&&Re(this.accumulatedMoments.map(e=>e.variable))}async getWeights(){let e=[...this.accumulatedMeanSquares,...this.accumulatedMoments];return this.centered&&e.push(...this.accumulatedMeanGrads),[await this.saveIterations()].concat(e.map(t=>({name:t.originalName,tensor:t.variable})))}async setWeights(e){e=await this.extractIterations(e);let t=this.centered?e.length/3:e.length/2,n=!1;this.accumulatedMeanSquares=e.slice(0,t).map(a=>({originalName:a.name,variable:a.tensor.variable(n)})),this.accumulatedMoments=e.slice(t,t*2).map(a=>({originalName:a.name,variable:a.tensor.variable(n)})),this.centered&&(this.accumulatedMeanGrads=e.slice(t*2,t*3).map(a=>({originalName:a.name,variable:a.tensor.variable(n)})))}getConfig(){return{learningRate:this.learningRate,decay:this.decay,momentum:this.momentum,epsilon:this.epsilon,centered:this.centered}}static fromConfig(e,t){return new e(t.learningRate,t.decay,t.momentum,t.epsilon,t.centered)}};Pf.className="RMSProp";Is(Pf);var Jr=class{static sgd(e){return new Qc(e)}static momentum(e,t,n=!1){return new Mf(e,t,n)}static rmsprop(e,t=.9,n=0,a=null,r=!1){return new Pf(e,t,n,a,r)}static adam(e=.001,t=.9,n=.999,a=null){return new Df(e,t,n,a)}static adadelta(e=.001,t=.95,n=null){return new Af(e,t,n)}static adamax(e=.002,t=.9,n=.999,a=null,r=0){return new Rf(e,t,n,a,r)}static adagrad(e,t=.1){return new $f(e,t)}},Hs={sgd:Jr.sgd,momentum:Jr.momentum,adadelta:Jr.adadelta,adagrad:Jr.adagrad,rmsprop:Jr.rmsprop,adamax:Jr.adamax,adam:Jr.adam},Bz=(()=>typeof requestAnimationFrame!="undefined"?requestAnimationFrame:typeof setImmediate!="undefined"?setImmediate:e=>e())();function Bv(){return new Promise(e=>Bz(()=>e()))}var C={};Me(C,{ERF_A1:()=>Qz,ERF_A2:()=>Jz,ERF_A3:()=>Zz,ERF_A4:()=>eB,ERF_A5:()=>tB,ERF_P:()=>Yz,PARALLELIZE_THRESHOLD:()=>Wv,SELU_SCALE:()=>k2,SELU_SCALEALPHA:()=>w2,applyActivation:()=>Ef,assertAndGetBroadcastShape:()=>pt,assertAxesAreInnerMostDims:()=>CP,assertParamsConsistent:()=>Wz,assignToTypedArray:()=>oB,axesAreInnerMostDims:()=>mv,calculateShapes:()=>cS,checkEinsumDimSizes:()=>hB,checkPadOnDimRoundingMode:()=>_n,combineLocations:()=>VS,complexWithEvenIndex:()=>rB,complexWithOddIndex:()=>sB,computeConv2DInfo:()=>Bc,computeConv3DInfo:()=>_S,computeDefaultPad:()=>rv,computeDilation2DInfo:()=>_M,computeOptimalWindowSize:()=>Uz,computeOutAndReduceShapes:()=>US,computeOutShape:()=>Vz,computePool2DInfo:()=>TS,computePool3DInfo:()=>CM,convertConv2DDataFormat:()=>CS,decodeEinsumEquation:()=>cB,eitherStridesOrDilationsAreOne:()=>mr,expandShapeToKeepDim:()=>pi,exponent:()=>uB,exponents:()=>lB,fromStringArrayToUint8:()=>RB,fromUint8ToStringArray:()=>DB,getAxesPermutation:()=>GS,getBroadcastDims:()=>lS,getComplexWithIndex:()=>iB,getEinsumComputePath:()=>mB,getEinsumPermutation:()=>dB,getFusedBiasGradient:()=>Cf,getFusedDyActivation:()=>_f,getImageCenter:()=>Gz,getInnerMostAxes:()=>EP,getPermuted:()=>jz,getReductionAxes:()=>Vt,getReshaped:()=>Hz,getReshapedPermuted:()=>qz,getSliceBeginCoords:()=>Kz,getSliceSize:()=>Xz,getSparseFillEmptyRowsIndicesDenseShapeMismatch:()=>bB,getSparseFillEmptyRowsNegativeIndexErrorMessage:()=>xB,getSparseFillEmptyRowsOutOfRangeIndexErrorMessage:()=>vB,getSparseReshapeEmptyTensorZeroOutputDimErrorMessage:()=>IB,getSparseReshapeInputOutputMismatchErrorMessage:()=>NB,getSparseReshapeInputOutputMultipleErrorMessage:()=>SB,getSparseReshapeMultipleNegativeOneOutputDimErrorMessage:()=>wB,getSparseReshapeNegativeOutputDimErrorMessage:()=>kB,getSparseSegmentReductionIndicesOutOfRangeErrorMessage:()=>EB,getSparseSegmentReductionNegativeSegmentIdsErrorMessage:()=>TB,getSparseSegmentReductionNonIncreasingSegmentIdsErrorMessage:()=>_B,getSparseSegmentReductionSegmentIdOutOfRangeErrorMessage:()=>CB,getUndoAxesPermutation:()=>fv,isIdentityPermutation:()=>fB,log:()=>j$,mergeRealAndImagArrays:()=>nB,prepareAndValidate:()=>pS,prepareSplitSize:()=>yB,segment_util:()=>I2,shouldFuse:()=>Ff,slice_util:()=>Kt,splitRealAndImagArrays:()=>aB,tupleValuesAreOne:()=>us,upcastType:()=>fa,validateInput:()=>qx,validateUpdateShape:()=>jx,warn:()=>es});function Wz(e,t){let n=e[0].length;e.forEach((r,s)=>{R(r.length===n,()=>`Error in concat${n}D: rank of tensors[${s}] must be the same as the rank of the rest (${n})`)}),R(t>=0&&t<n,()=>`Error in concat${n}D: axis must be between 0 and ${n-1}.`);let a=e[0];e.forEach((r,s)=>{for(let i=0;i<n;i++)R(i===t||r[i]===a[i],()=>`Error in concat${n}D: Shape of tensors[${s}] (${r}) does not match the shape of the rest (${a}) along the non-concatenated axis ${s}.`)})}function Vz(e,t){let n=e[0].slice();for(let a=1;a<e.length;a++)n[t]+=e[a][t];return n}var Wv=30;function Uz(e){return e<=Wv?e:Uh(e,Math.floor(Math.sqrt(e)))}function Gz(e,t,n){let a=n*(typeof e=="number"?e:e[0]),r=t*(typeof e=="number"?e:e[1]);return[a,r]}function Hz(e,t,n,a=!0){let r=[];if(a)r=r.concat(t.slice(0)),r.push(e[0]/n),r=r.concat(e.slice(1));else{r=r.concat(e[0]);let s=t.length;for(let i=0;i<s;++i)r=r.concat([e[i+1]/t[i],t[i]]);r=r.concat(e.slice(s+1))}return r}function jz(e,t,n=!0){let a=[];if(n){a.push(t);for(let r=t+1;r<e;++r)r<=2*t?(a.push(r),a.push(r-(t+1))):a.push(r)}else{let r=[],s=[];for(let i=1;i<e;++i)i>=t*2+1||i%2===1?s.push(i):r.push(i);a.push(...r),a.push(0),a.push(...s)}return a}function qz(e,t,n,a=!0){let r=[];a?r.push(e[0]/n):r.push(e[0]*n);for(let s=1;s<e.length;++s)s<=t.length?a?r.push(t[s-1]*e[s]):r.push(e[s]/t[s-1]):r.push(e[s]);return r}function Kz(e,t){let n=[0];for(let a=0;a<t;++a)n.push(e[a][0]);return n}function Xz(e,t,n){let a=e.slice(0,1);for(let r=0;r<n;++r)a.push(e[r+1]-t[r][0]-t[r][1]);return a}var w2=1.7580993408473768,k2=1.0507009873554805,Yz=.3275911,Qz=.254829592,Jz=-.284496736,Zz=1.421413741,eB=-1.453152027,tB=1.061405429;function nB(e,t){if(e.length!==t.length)throw new Error(`Cannot merge real and imag arrays of different lengths. real:${e.length}, imag: ${t.length}.`);let n=new Float32Array(e.length*2);for(let a=0;a<n.length;a+=2)n[a]=e[a/2],n[a+1]=t[a/2];return n}function aB(e){let t=new Float32Array(e.length/2),n=new Float32Array(e.length/2);for(let a=0;a<e.length;a+=2)t[a/2]=e[a],n[a/2]=e[a+1];return{real:t,imag:n}}function rB(e){let t=Math.ceil(e.length/4),n=new Float32Array(t),a=new Float32Array(t);for(let r=0;r<e.length;r+=4)n[Math.floor(r/4)]=e[r],a[Math.floor(r/4)]=e[r+1];return{real:n,imag:a}}function sB(e){let t=Math.floor(e.length/4),n=new Float32Array(t),a=new Float32Array(t);for(let r=2;r<e.length;r+=4)n[Math.floor(r/4)]=e[r],a[Math.floor(r/4)]=e[r+1];return{real:n,imag:a}}function iB(e,t){let n=e[t*2],a=e[t*2+1];return{real:n,imag:a}}function oB(e,t,n,a){e[a*2]=t,e[a*2+1]=n}function lB(e,t){let n=new Float32Array(e/2),a=new Float32Array(e/2);for(let r=0;r<Math.ceil(e/2);r++){let s=(t?2:-2)*Math.PI*(r/e);n[r]=Math.cos(s),a[r]=Math.sin(s)}return{real:n,imag:a}}function uB(e,t,n){let a=(n?2:-2)*Math.PI*(e/t),r=Math.cos(a),s=Math.sin(a);return{real:r,imag:s}}var wb="->",pB=/->/g,sk=",",ik="...";function cB(e,t){e=e.replace(/\s/g,"");let n=(e.length-e.replace(pB,"").length)/wb.length;if(n<1)throw new Error("Equations without an arrow are not supported.");if(n>1)throw new Error(`Equation must contain exactly one arrow ("${wb}").`);let[a,r]=e.split(wb);R(a.indexOf(ik)===-1,()=>`The ellipsis notation ("${ik}") is not supported yet.`);let s=a.split(sk),i=s.length;if(t!==i)throw new Error(`Expected ${i} input tensors, received ${t}`);if(i>2)throw new Error("Support for more than 2 input tensors is not implemented yet.");let o=[];for(let c=0;c<r.length;++c){let h=r[c];if(!s.some(m=>m.indexOf(h)!==-1))throw new Error(`Output subscripts contain the label ${h} not present in the input subscripts.`);o.indexOf(h)===-1&&o.push(h)}for(let c=0;c<a.length;++c){let h=a[c];o.indexOf(h)===-1&&h!==sk&&o.push(h)}let l=new Array(s.length);for(let c=0;c<i;++c){if(new Set(s[c].split("")).size!==s[c].length)throw new Error(`Found duplicate axes in input component ${s[c]}. Support for duplicate axes in input is not implemented yet.`);l[c]=[];for(let h=0;h<s[c].length;++h)l[c].push(o.indexOf(s[c][h]))}let u=o.length,p=r.length,d=[];for(let c=p;c<u;++c)d.push(c);return{allDims:o,summedDims:d,idDims:l}}function dB(e,t){let n=new Array(e);n.fill(-1);for(let r=0;r<t.length;++r)n[t[r]]=r;let a=[];for(let r=0;r<e;++r)n[r]===-1&&a.push(r);return n=n.filter(r=>r!==-1),{permutationIndices:n,expandDims:a}}function hB(e,t,n){let a=new Array(e);for(let r=0;r<n.length;++r){let s=n[r].shape;for(let i=0;i<t[r].length;++i)a[t[r][i]]===void 0?a[t[r][i]]=s[i]:R(a[t[r][i]]===s[i],()=>`Expected dimension ${a[t[r][i]]} at axis ${i} of input shaped ${JSON.stringify(s)}, but got dimension ${s[i]}`)}}function mB(e,t){let n=e,a=[],r=0;e.length===0&&n.push(-1),r=e.length+1;for(let i=0;i<r;++i)a.push([]);let s=[];for(let i=0;i<n.length;++i){let o=n[i],l=gB(t,o);for(let u of l)s.indexOf(u)===-1&&(a[i].push(u),s.push(u))}return{path:n,steps:a}}function fB(e){return e.every((t,n)=>t===n)}function gB(e,t){let n=[];for(let a=0;a<e.length;++a)(e[a].length===0||e[a].indexOf(t)!==-1||t===-1)&&n.push(a);return n}function yB(e,t,n=0){let a=[];if(typeof t=="number")R(e.shape[n]%t===0,()=>"Number of splits must evenly divide the axis."),a=new Array(t).fill(e.shape[n]/t);else{let r=t.reduce((i,o)=>(o===-1&&(i+=1),i),0);R(r<=1,()=>"There should be only one negative value in split array.");let s=t.indexOf(-1);if(s!==-1){let i=t.reduce((o,l)=>l>0?o+l:o);t[s]=e.shape[n]-i}R(e.shape[n]===t.reduce((i,o)=>i+o),()=>"The sum of sizes must match the size of the axis dimension."),a=t}return a}function bB(e){return`Received SparseTensor with denseShape[0] = 0 but
indices.shape[0] = ${e}`}function xB(e,t){return`indices(${e}, 0) is invalid: ${t} < 0`}function vB(e,t,n){return`indices(${e}, 0) is invalid: ${t} >= ${n}`}function wB(e,t){return`only one output dimension may be -1, not both ${e} and ${t}`}function kB(e,t){return`size ${e} must be non-negative, not ${t}`}function IB(){return"reshape cannot infer the missing input size for an empty tensor unless all specified input sizes are non-zero"}function SB(e,t){let n=bt(e),a=bt(t);return`Input to reshape is a SparseTensor with ${n}
dense values, but the requested shape requires a multiple of ${a}. inputShape=${e} outputShape= ${t}`}function NB(e,t){let n=bt(e),a=bt(t);return`Input to reshape is a tensor with ${n} dense values, but the requested shape has ${a}. inputShape=${e} outputShape=${t}`}function TB(){return"segment ids must be >= 0"}function _B(){return"segment ids are not increasing"}function CB(e,t){return`Segment id ${e} out of range [0, ${t}), possibly because segmentIds input is not sorted.`}function EB(e,t,n){return`Bad: indices[${e}] == ${t} out of range [0, ${n})`}var I2={};Me(I2,{collectGatherOpShapeInfo:()=>$B,computeOutShape:()=>AB,segOpComputeOptimalWindowSize:()=>FB});function FB(e,t){let n=!1,a;for(e<=Wv?(a=e,n=!0):a=Uh(e,Math.floor(Math.sqrt(e)));!n;)a>t||a===e?n=!0:a=Uh(e,a+1);return a}function AB(e,t,n){let a=[],r=e.length;for(let s=0;s<r;s++)s!==t?a.push(e[s]):a.push(n);return a}function $B(e,t,n,a){let r=t.shape.length,s=e.shape.length;if(a!==0&&(a<-r||a>r))throw new Error(`Expect batchDims in the range of [-${r}, ${r}], but got ${a}`);if(a<0&&(a+=r),a>s)throw new Error(`batchDims (${a}) must be less than rank(x) (
${s}).`);if(n<a)throw new Error(`batchDims (${a}) must be less than or equal to axis (${n}).`);for(let d=0;d<a;++d)if(e.shape[d]!==t.shape[d])throw new Error(`x.shape[${d}]: ${e.shape[d]} should be equal to indices.shape[${d}]: ${t.shape[d]}.`);let i=e.shape[n],o=[],l=1,u=1,p=1;for(let d=0;d<a;++d)o.push(e.shape[d]),l*=e.shape[d];for(let d=a;d<n;d++)o.push(e.shape[d]),u*=e.shape[d];for(let d=a;d<r;d++)o.push(t.shape[d]);for(let d=n+1;d<s;d++)o.push(e.shape[d]),p*=e.shape[d];return{batchSize:l,sliceSize:p,outerSize:u,dimSize:i,outputShape:o}}function DB(e){try{return e.map(t=>Xh(t))}catch(t){throw new Error(`Failed to decode encoded string bytes into utf-8, error: ${t}`)}}function RB(e){return e.map(t=>Pc(t))}var gr={};Me(gr,{nonMaxSuppressionV3Impl:()=>f2,nonMaxSuppressionV4Impl:()=>g2,nonMaxSuppressionV5Impl:()=>y2,whereImpl:()=>i2});var S2={kernelName:Ll,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>B(e,ju(oe(n,"float32"),-1))}}},MB={kernelName:zl,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>{let a=ut(oe(n,"float32")),r=pn(ce(we(1),a));return It(fe(e,r))}}}},PB={kernelName:Bl,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>{let a=pn(ce(ut(oe(n,"float32")),1));return fe(e,a)}}}},OB={kernelName:bs,inputsToSave:["a","b"],gradFunc:(e,t)=>{let[n,a]=t,r=pt(n.shape,a.shape);return{a:()=>{let s=e,i=Vt(n.shape,r);return i.length>0&&(s=be(s,i)),W(s,n.shape)},b:()=>{let s=e,i=Vt(a.shape,r);return i.length>0&&(s=be(s,i)),W(s,a.shape)}}}},LB={kernelName:wi,saveAllInputs:!0,gradFunc:(e,t)=>{let n={};return t.forEach((a,r)=>{n[r]=()=>e.clone()}),n}},zB={kernelName:ki,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>Ke(n)}}},BB={kernelName:yc,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>Ke(n)}}},WB={kernelName:Ul,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>fe(e,pn(ce(we(1),ut(oe(n,"float32")))))}}},VB={kernelName:Gl,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>{let a=pn(J(we(1),ut(oe(n,"float32"))));return fe(e,a)}}}},UB={kernelName:ql,inputsToSave:["a","b"],gradFunc:(e,t)=>{let[n,a]=t,r=pt(n.shape,a.shape);return{a:()=>{let s=J(ut(n),ut(a)),i=B(e,fe(a,s)),o=Vt(n.shape,r);return o.length>0&&(i=be(i,o)),W(i,n.shape)},b:()=>{let s=J(ut(n),ut(a)),i=It(B(e,fe(n,s))),o=Vt(a.shape,r);return o.length>0&&(i=be(i,o)),W(i,a.shape)}}}},GB={kernelName:Hl,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>fe(e,J(ut(oe(n,"float32")),1))}}},HB={kernelName:jl,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>fe(e,ce(we(1),ut(oe(n,"float32"))))}}};function jB(e,t,n,a,r,s){let i=F(e,"dy","avgPool3dGrad"),o=F(t,"input","avgPool3dGrad"),l=i,u=o,p=!1;o.rank===4&&(p=!0,l=W(i,[1,i.shape[0],i.shape[1],i.shape[2],i.shape[3]]),u=W(o,[1,o.shape[0],o.shape[1],o.shape[2],o.shape[3]])),R(l.rank===5,()=>`Error in avgPool3dGrad: dy must be rank 5 but got rank ${l.rank}.`),R(u.rank===5,()=>`Error in avgPool3dGrad: input must be rank 5 but got rank ${u.rank}.`),_n("avgPool3dGrad",r,s);let d={dy:l,input:u},c={filterSize:n,strides:a,pad:r,dimRoundingMode:s},h=L.runKernel(Im,d,c);return p?W(h,[h.shape[1],h.shape[2],h.shape[3],h.shape[4]]):h}var qB=z({avgPool3dGrad_:jB}),KB={kernelName:bc,inputsToSave:["x"],gradFunc:(e,t,n)=>{let[a]=t,{filterSize:r,strides:s,pad:i,dimRoundingMode:o}=n;return{x:()=>qB(e,a,r,s,i,o)}}};function XB(e,t,n,a,r){let s=F(e,"dy","avgPoolGrad"),i=F(t,"input","avgPoolGrad");R(i.rank===s.rank,()=>`Rank of input (${i.rank}) does not match rank of dy (${s.rank})`);let o=i,l=s,u=!1;i.rank===3&&(u=!0,o=W(i,[1,i.shape[0],i.shape[1],i.shape[2]]),l=W(s,[1,s.shape[0],s.shape[1],s.shape[2]])),R(l.rank===4,()=>`Error in avgPoolGrad: dy must be rank 4 but got rank ${l.rank}.`),R(o.rank===4,()=>`Error in avgPoolGrad: input must be rank 4 but got rank ${o.rank}.`);let p={dy:l,input:o},d={filterSize:n,strides:a,pad:r},c=L.runKernel(km,p,d);return u?W(c,[c.shape[1],c.shape[2],c.shape[3]]):c}var YB=z({avgPoolGrad_:XB}),QB={kernelName:Ii,inputsToSave:["x"],gradFunc:(e,t,n)=>{let[a]=t,{filterSize:r,strides:s,pad:i}=n;return{x:()=>YB(e,a,r,s,i)}}},JB={kernelName:Si,inputsToSave:["a","b"],gradFunc:(e,t,n)=>{let[a,r]=t,{transposeA:s,transposeB:i}=n;return!s&&!i?{a:()=>De(e,r,!1,!0),b:()=>De(a,e,!0,!1)}:!s&&i?{a:()=>De(e,r,!1,!1),b:()=>De(e,a,!0,!1)}:s&&!i?{a:()=>De(r,e,!1,!0),b:()=>De(a,e,!1,!1)}:{a:()=>De(r,e,!0,!0),b:()=>De(e,a,!0,!0)}}},ZB={kernelName:Kl,gradFunc:(e,t,n)=>{let{blockShape:a,crops:r}=n;return{x:()=>qc(e,a,r)}}},eW={kernelName:DI,gradFunc:(e,t,n)=>{let a=n,r=a.inputShape,s=a.shape,i=Array.from(s);for(let l=r.length-1;l>=0;l--)if(r[l]===s[l])i[l]=1;else if(r[l]!==1)throw new Error(`broadcastTo(): [${r}] cannot be broadcast to [${s}].`);let o=[];for(let l=0;l<i.length;l++)i[l]>1&&o.push(l);return{x:()=>be(e,o,!0)}}},tW={kernelName:Ni,gradFunc:e=>({x:()=>e.clone()})},nW={kernelName:Ti,gradFunc:e=>({x:()=>Ke(e)})},aW={kernelName:xs,inputsToSave:["x"],gradFunc:(e,t,n)=>{let[a]=t,{clipValueMin:r,clipValueMax:s}=n;return{x:()=>fn(Ca(Ns(a,r),Ts(a,s)),e,Ke(e))}}},rW={kernelName:xc,inputsToSave:["x"],gradFunc:S2.gradFunc},sW={kernelName:Xl,saveAllInputs:!0,gradFunc:(e,t,n)=>{let a=t.map(o=>o.shape),{axis:r}=n,s=Ea(r,t[0].shape)[0],i=a.map(o=>o[s]);return zn(e,i,s).map(o=>()=>o)}},iW={kernelName:_i,inputsToSave:["x","filter"],gradFunc:(e,t,n)=>{let[a,r]=t,{dilations:s,strides:i,pad:o,dataFormat:l}=n;return R(us(s),()=>`Error in gradient of conv2D: dilation rates greater than 1 are not yet supported in gradients. Got dilations '${s}'`),{x:()=>lv(a.shape,e,r,i,o,l),filter:()=>Lv(a,e,r.shape,i,o,l)}}},oW={kernelName:Ci,inputsToSave:["dy","filter"],gradFunc:(e,t,n)=>{let[a,r]=t,{strides:s,pad:i,dataFormat:o,dimRoundingMode:l}=n;return{dy:()=>Rt(e,r,s,i,o,1,l),filter:()=>Lv(e,a,r.shape,s,i,o,l)}}};function lW(e,t,n,a,r){let s=e;e.rank===4&&(s=W(e,[1,e.shape[0],e.shape[1],e.shape[2],e.shape[3]]));let i=t;i.rank===4&&(i=W(t,[1,t.shape[0],t.shape[1],t.shape[2],t.shape[3]])),R(s.rank===5,()=>`Error in conv3dDerFilter: input must be rank 5, but got shape ${s.shape}.`),R(i.rank===5,()=>`Error in conv3dDerFilter: dy must be rank 5, but got shape ${i.shape}.`),R(n.length===5,()=>`Error in conv3dDerFilter: filterShape must be length 5, but got ${n}.`),R(s.shape[4]===n[3],()=>`Error in conv3dDerFilter: depth of input ${s.shape[4]}) must match input depth in filter (${n[3]}.`),R(i.shape[4]===n[4],()=>`Error in conv3dDerFilter: depth of dy (${i.shape[4]}) must match output depth for filter (${n[4]}).`);let o={x:s,dy:i},l={strides:a,pad:r,filterShape:n};return L.runKernel(Cm,o,l)}var uW=z({conv3DBackpropFilter_:lW}),pW={kernelName:vc,inputsToSave:["x","filter"],gradFunc:(e,t,n)=>{let{dilations:a,strides:r,pad:s}=n;R(us(a),()=>`Error in gradient of conv3D: dilation rates greater than 1 are not yet supported in gradients. Got dilations '${a}'`);let[i,o]=t;return{x:()=>OS(i.shape,e,o,r,s),filter:()=>uW(i,e,o.shape,r,s)}}},cW={kernelName:Ei,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>B(It(wf(oe(n,"float32"))),e)}}},dW={kernelName:Fi,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>B(kf(oe(n,"float32")),e)}}},hW={kernelName:Ai,inputsToSave:["x"],gradFunc:(e,t,n)=>{let[a]=t,{axis:r,exclusive:s,reverse:i}=n;return{x:()=>{let o=GS([r],a.rank),l=pf(e,r,s,!i);return o!=null&&(l=Ae(l,o)),l}}}},mW={kernelName:$i,inputsToSave:["x","filter"],gradFunc:(e,t,n)=>{let{dilations:a,strides:r,pad:s,dimRoundingMode:i}=n,o=a==null?[1,1]:a;R(us(o),()=>`Error in gradient of depthwiseConv2dNative: dilation rates greater than 1 are not yet supported. Got dilations '${o}'`);let[l,u]=t;return R(l.rank===4,()=>`Error in gradient of depthwiseConv2dNative: input must be rank 4, but got rank ${l.rank}.`),R(u.rank===4,()=>`Error in gradient of depthwiseConv2dNative: filter must be rank 4, but got rank ${u.rank}.`),R(l.shape[3]===u.shape[2],()=>`Error in gradient of depthwiseConv2d: number of input channels (${l.shape[3]}) must match the inChannels dimension in filter ${u.shape[2]}.`),R(mr(r,o),()=>`Error in gradient of depthwiseConv2d: Either strides or dilations must be 1. Got strides ${r} and dilations '${o}'.`),_n("depthwiseConv2d",s,i),{x:()=>d2(l.shape,e,u,r,s,o,i),filter:()=>c2(l,e,u.shape,r,s,o,i)}}},fW={kernelName:wc,inputsToSave:["x","filter"],gradFunc:(e,t,n)=>{let[a,r]=t,s={x:a,filter:r,dy:e},i={x:a,filter:r,dy:e};return{x:()=>L.runKernel(Gh,s,n),filter:()=>L.runKernel(Hh,i,n)}}},gW={kernelName:Ri,outputsToSave:[!0],gradFunc:(e,t)=>{let[n]=t,a={dy:e,y:n};return{x:()=>L.runKernel(Mm,a)}}},yW={kernelName:Zl,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t,a=B(gn(It(ut(n))),2/Math.sqrt(Math.PI));return{x:()=>B(e,a)}}},bW={kernelName:Mi,outputsToSave:[!0],gradFunc:(e,t)=>{let[n]=t;return{x:()=>B(e,n)}}},xW={kernelName:tu,inputsToSave:["input"],gradFunc:(e,t)=>{let[n]=t;return{input:()=>W(e,n.shape)}}},vW={kernelName:nu,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>B(e,gn(n))}}},wW={kernelName:Pi,gradFunc:e=>({x:()=>Ke(e)})},kW={kernelName:Oi,inputsToSave:["a","b"],gradFunc:(e,t)=>{let[n,a]=t,r=pt(n.shape,a.shape);return{a:()=>{let s=fe(e,oe(a,"float32")),i=Vt(n.shape,r);return i.length>0?W(be(s,i),n.shape):s},b:()=>{let s=B(e,oe(n,"float32")),i=Vt(a.shape,r);i.length>0&&(s=W(be(s,i),a.shape));let o=ut(a);return It(fe(s,oe(o,"float32")))}}}},IW={kernelName:Li,inputsToSave:["x","mean","variance","scale"],gradFunc:(e,t,n)=>{let{varianceEpsilon:a}=n,[r,s,i,o]=t,l=o==null?we(1):o,u=Vt(s.shape,r.shape),p=[];if(s.rank===1){for(let f=0;f<r.shape.length-1;++f)p.push(r.shape[f]);p.push(1)}let d=ce(r,s),c=B(e,l),h=xf(J(i,we(a))),m=B(B(B(h,h),h),we(-.5));return{x:()=>s.rank===1?W(B(B(e,On(W(h,[1,1,1,s.shape[0]]),p)),l),r.shape):W(B(B(e,h),l),r.shape),mean:()=>{let f=B(B(h,we(-1)),c);return s.rank===1&&(f=be(f,u)),W(f,s.shape)},variance:()=>{let f=B(B(m,d),c);return s.rank===1&&(f=be(f,u)),W(f,s.shape)},scale:()=>{let f=B(d,h),g=B(e,f);return s.rank===1&&(g=be(g,u)),W(g,s.shape)},offset:()=>{let f=e;return s.rank===1&&(f=be(f,u)),W(f,s.shape)}}}},SW={kernelName:ru,inputsToSave:["x","indices"],gradFunc:(e,t,n)=>{let[a,r]=t,{axis:s}=n,i=Ea(s,a.shape)[0];return{x:()=>{let o=a.shape,l=r.size,u=o.slice(0,i),p=u.length,d=o.slice(s,o.length).slice(1),c=d.length,h=ok(0,p),m=ok(p+1,p+1+c),f=lk([u,[l],d]),g=W(e,f),y=W(r,[l]),b=lk([[p],h,m]),x=Ae(g,b),v=Rv(x,y,a.shape[i]),k=fv(b);return v=Ae(v,k),v},indices:()=>r}}};function ok(e,t){let n=[];for(let a=e;a<t;++a)n.push(a);return n}function lk(e){let t=[];for(let n=0;n<e.length;++n)for(let a=0;a<e[n].length;++a)t.push(e[n][a]);return t}var NW={kernelName:zi,inputsToSave:["a","b"],gradFunc:(e,t)=>{let[n,a]=t;return{a:()=>Ke(n),b:()=>Ke(a)}}},TW={kernelName:Bi,gradFunc:e=>({x:()=>oe(e,"float32")})},_W={kernelName:ou,gradFunc:e=>({x:()=>Ke(e)})},CW={kernelName:lu,gradFunc:e=>({x:()=>Ke(e)})},EW={kernelName:uu,gradFunc:e=>({x:()=>Ke(e)})},FW={kernelName:Wi,inputsToSave:["x"],gradFunc:(e,t,n)=>{let[a]=t,{alpha:r}=n,s=Un(a,0);return{x:()=>fn(s,e,B(e,r))}}},AW={kernelName:du,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>fe(e,J(n,1))}}},$W={kernelName:Vi,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>fe(e,oe(n,"float32"))}}},DW={kernelName:RI,inputsToSave:[],outputsToSave:[!0],gradFunc:(e,t,n)=>{let[a]=t,{axis:r}=n;return{logits:()=>{let s=gn(a);return ce(e,B(be(e,r,!0),s))}}}};function RW(e,t,n,a=5,r=1,s=1,i=.5){let o={x:e,y:t,dy:n},l={depthRadius:a,bias:r,alpha:s,beta:i};return L.runKernel(Bm,o,l)}var MW=z({localResponseNormalizationBackprop_:RW}),PW={kernelName:Nc,inputsToSave:["x"],outputsToSave:[!0],gradFunc:(e,t,n)=>{let[a,r]=t,{depthRadius:s,bias:i,alpha:o,beta:l}=n;return{x:()=>MW(a,r,e,s,i,o,l)}}};function N2(e,t,n,a){return t.rank<n.rank&&(t=W(t,pi(t.shape,a))),e.rank<n.rank&&(e=W(e,pi(e.shape,a))),{x:()=>B(e,oe(Zn(n,t),e.dtype))}}var uk={kernelName:Ui,inputsToSave:["x"],outputsToSave:[!0],gradFunc:(e,t,n)=>{let a=n,{reductionIndices:r}=a,s=t[0],i=t[1],o=Ea(r,s.shape),l=N2(e,i,s,o);return{x:()=>l.x()}}},OW={kernelName:Gi,inputsToSave:["a","b"],gradFunc:(e,t)=>{let[n,a]=t;return{a:()=>B(e,oe(Ns(n,a),"float32")),b:()=>B(e,oe(cf(n,a),"float32"))}}};function LW(e,t,n,a,r,s,i){let o=F(e,"dy","maxPool3dGrad"),l=F(t,"input","maxPool3dGrad"),u=F(n,"output","maxPool3dGrad"),p=o,d=l,c=u,h=!1;l.rank===4&&(h=!0,p=W(o,[1,o.shape[0],o.shape[1],o.shape[2],o.shape[3]]),d=W(l,[1,l.shape[0],l.shape[1],l.shape[2],l.shape[3]]),c=W(u,[1,u.shape[0],u.shape[1],u.shape[2],u.shape[3]])),R(p.rank===5,()=>`Error in maxPool3dGrad: dy must be rank 5 but got rank ${p.rank}.`),R(d.rank===5,()=>`Error in maxPool3dGrad: input must be rank 5 but got rank ${d.rank}.`),R(c.rank===5,()=>`Error in maxPool3dGrad: output must be rank 5 but got rank ${c.rank}.`),_n("maxPool3dGrad",s,i);let m={dy:p,input:d,output:c},f={filterSize:a,strides:r,pad:s,dimRoundingMode:i},g=L.runKernel(Vm,m,f);return h?W(g,[g.shape[1],g.shape[2],g.shape[3],g.shape[4]]):g}var zW=z({maxPool3dGrad_:LW}),BW={kernelName:Tc,inputsToSave:["x"],outputsToSave:[!0],gradFunc:(e,t,n)=>{let[a,r]=t,{filterSize:s,strides:i,pad:o,dimRoundingMode:l}=n;return{x:()=>zW(e,a,r,s,i,o,l)}}};function WW(e,t,n,a,r,s,i){let o=F(e,"dy","maxPoolGrad"),l=F(t,"input","maxPoolGrad"),u=F(n,"output","maxPoolGrad");R(l.rank===o.rank,()=>`Rank of input (${l.rank}) does not match rank of dy (${o.rank})`),R(o.rank===4,()=>`Error in maxPoolGrad: dy must be rank 4 but got rank ${o.rank}.`),R(l.rank===4,()=>`Error in maxPoolGrad: input must be rank 4 but got rank ${l.rank}.`),_n("maxPoolGrad",s,i);let p={dy:o,input:l,output:u},d={filterSize:a,strides:r,pad:s,dimRoundingMode:i};return L.runKernel(Wm,p,d)}var VW=z({maxPoolGrad_:WW}),UW={kernelName:Hi,inputsToSave:["x"],outputsToSave:[!0],gradFunc:(e,t,n)=>{let[a,r]=t,{filterSize:s,strides:i,pad:o}=n;return{x:()=>VW(e,a,r,s,i,o)}}},GW={kernelName:ji,inputsToSave:["x"],gradFunc:(e,t,n)=>{let[a]=t,{axis:r}=n,s=Ea(r,a.shape),i=US(a.shape,s)[1],o=bt(i);return{x:()=>{let l=a.shape.slice();s.forEach(p=>{l[p]=1});let u=W(e,l);return fe(B(u,Qn(a.shape,"float32")),o)}}}},HW={kernelName:qi,inputsToSave:["x"],outputsToSave:[!0],gradFunc:(e,t,n)=>{let a=n,{axis:r}=a,[s,i]=t,o=Ea(r,s.shape),l=N2(e,i,s,o);return{x:()=>l.x()}}},jW={kernelName:Ki,inputsToSave:["a","b"],gradFunc:(e,t)=>{let[n,a]=t;return{a:()=>B(e,oe(Ts(n,a),"float32")),b:()=>B(e,oe(Un(n,a),"float32"))}}},qW={kernelName:Xi,inputsToSave:["x"],gradFunc:(e,t,n)=>{let a=t[0],{paddings:r}=n,s=r.map(i=>i[0]);return{x:()=>He(e,s,a.shape)}}},KW={kernelName:mu,inputsToSave:["a","b"],gradFunc:(e,t)=>{let[n,a]=t,r=pt(n.shape,a.shape);return{a:()=>{let s=Vt(n.shape,r);return s.length>0?W(be(e,s),n.shape):e},b:()=>{let s=B(e,It(Vu(fe(n,a)))),i=Vt(a.shape,r);return i.length>0?W(be(s,i),a.shape):s}}}},XW={kernelName:Yi,inputsToSave:["a","b"],gradFunc:(e,t)=>{let[n,a]=t,r=pt(n.shape,a.shape);return{a:()=>{let s=B(e,oe(a,"float32")),i=Vt(n.shape,r);return i.length>0?W(be(s,i),n.shape):s},b:()=>{let s=B(e,oe(n,"float32")),i=Vt(a.shape,r);return i.length>0?W(be(s,i),a.shape):s}}}},YW={kernelName:fu,gradFunc:e=>({x:()=>It(e)})},QW={kernelName:Qi,inputsToSave:["indices"],gradFunc:(e,t)=>{let n=t[0];return{indices:()=>St(n.shape,"float32")}}},JW={kernelName:vu,gradFunc:e=>({x:()=>Ke(e)})},ZW={kernelName:wu,saveAllInputs:!0,gradFunc:(e,t,n)=>{let{axis:a}=n;return ht(e,a).map(r=>()=>r)}},pk={kernelName:Ji,inputsToSave:["x"],gradFunc:(e,t,n)=>{let a=t[0],{paddings:r}=n,s=r.map(i=>i[0]);return{x:()=>He(e,s,a.shape)}}},e4={kernelName:Zi,inputsToSave:["a","b"],outputsToSave:[!0],gradFunc:(e,t)=>{let[n,a,r]=t,s=n,i=a,o=pt(s.shape,i.shape);return{a:()=>{let l=oe(i,"float32"),u=B(e,B(l,Dr(s,ce(l,we(1))))),p=Vt(s.shape,o);return p.length>0&&(u=be(u,p)),W(u,s.shape)},b:()=>{let l=Un(s,0),u=fn(l,ea(s),Ke(s)),p=B(e,B(r,u)),d=Vt(i.shape,o);return d.length>0&&(p=be(p,d)),W(p,i.shape)}}}},t4={kernelName:eo,inputsToSave:["x","alpha"],gradFunc:(e,t)=>{let[n,a]=t,r=Un(n,0);return{x:()=>fn(r,e,B(e,a)),alpha:()=>{let s=fn(r,Ke(e),B(e,n)),i=Vt(a.shape,e.shape);return i.length>0&&(s=be(s,i)),W(s,a.shape)}}}};function n4(e,t,n){let a=e.shape.slice();a[n]=1;let r=W(t,a),s=sc(e,n,!0,!1),i=sc(e,n,!0,!0),o=B(s,i);return B(r,o)}function a4(e,t,n){let a=e.shape.length,r=a-n.length,s=C.getAxesPermutation(n,a),i=e;s!=null&&(i=Ae(e,s));let o=i.shape.slice(),l=o.splice(a-n.length,n.length).reduce((d,c)=>d*c,1);o.push(l);let u=i.reshape(o),p=n4(u,t,r);if(p=p.reshape(i.shape),s!=null){let d=C.getUndoAxesPermutation(s);p=Ae(p,d)}return p}var r4={kernelName:to,inputsToSave:["x"],gradFunc:(e,t,n)=>{let[a]=t,{axis:r}=n,s=[];return r==null?s=a.shape.map((i,o)=>o):typeof r=="number"?s=[r]:s=r,{x:()=>a4(a,e,s)}}},s4={kernelName:Di,inputsToSave:["a","b"],gradFunc:(e,t)=>{let[n,a]=t,r=pt(n.shape,a.shape);return{a:()=>{let s=fe(e,oe(a,"float32")),i=Vt(n.shape,r);return i.length>0?W(be(s,i),n.shape):s},b:()=>{let s=B(e,oe(n,"float32")),i=Vt(a.shape,r);i.length>0&&(s=W(be(s,i),a.shape));let o=ut(a);return It(fe(s,oe(o,"float32")))}}}},i4={kernelName:ku,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>fe(e,It(ut(n)))}}},o4={kernelName:ro,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t,a=B(Ts(n,6),ju(n));return{x:()=>B(e,oe(a,"float32"))}}},l4={kernelName:no,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>B(e,oe(ju(n),"float32"))}}},u4={kernelName:Iu,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>W(e,n.shape)}}},p4={kernelName:ao,inputsToSave:["images"],gradFunc:(e,t,n)=>{let[a]=t,r={dy:e,images:a};return{images:()=>L.runKernel(qm,r,n)}}},c4={kernelName:Cc,inputsToSave:["images"],gradFunc:(e,t,n)=>{let[a]=t,r={dy:e,images:a};return{images:()=>L.runKernel(jm,r,n)}}},d4={kernelName:so,gradFunc:(e,t,n)=>{let{dims:a}=n,r=Ea(a,e.shape);return{x:()=>na(e,r)}}},h4={kernelName:io,gradFunc:e=>({x:()=>Ke(e)})},m4={kernelName:oo,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>It(fe(e,B(Dr(n,1.5),2)))}}},f4={kernelName:Nu,inputsToSave:["condition"],gradFunc:(e,t)=>{let[n]=t;return{condition:()=>oe(Ke(n),"float32"),t:()=>B(e,oe(n,e.dtype)),e:()=>B(e,oe(jc(n),e.dtype))}}},g4={kernelName:Tu,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>{let a=Un(n,we(0)),r=we(w2),s=we(k2),i=B(e,s),o=B(B(e,r),gn(oe(n,"float32")));return fn(a,i,o)}}}},y4={kernelName:uo,outputsToSave:[!0],gradFunc:(e,t)=>{let[n]=t;return{x:()=>B(e,B(n,ce(we(1),n)))}}},b4={kernelName:Eu,gradFunc:e=>({x:()=>Ke(e)})},x4={kernelName:lo,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>B(Vc(oe(n,"float32")),e)}}},v4={kernelName:Cu,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>B(uf(oe(n,"float32")),e)}}},w4={kernelName:_u,inputsToSave:["x"],gradFunc:(e,t,n)=>{let[a]=t,{begin:r,size:s}=n,i=a.shape,[o,l]=vS(a,r,s),u=[];for(let p=0;p<e.rank;p++)u.push([o[p],i[p]-o[p]-l[p]]);return{x:()=>ya(e,u)}}},k4={kernelName:ho,outputsToSave:[!0],gradFunc:(e,t,n)=>{let[a]=t,{dim:r}=n,s=!0,i=B(e,a);return{logits:()=>ce(i,B(be(i,[r],s),a))}}},I4={kernelName:Fu,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>B(e,ma(n))}}},ck={kernelName:Au,gradFunc:(e,t,n)=>{let{blockShape:a,paddings:r}=n;return{x:()=>Wc(e,a,r)}}},dk={kernelName:$u,gradFunc:(e,t,n)=>{let{axis:a}=n;return{x:()=>Ze(e,a)}}},S4={kernelName:po,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>fe(e,B(pn(oe(n,"float32")),2))}}},N4={kernelName:$c,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>B(e,B(oe(n,"float32"),2))}}},T4={kernelName:mo,inputsToSave:["a","b"],gradFunc:(e,t)=>{let[n,a]=t,r=we(2);return{a:()=>B(e,B(r,ce(n,a))),b:()=>B(e,B(r,ce(a,n)))}}},_4={kernelName:ws,gradFunc:e=>({x:()=>Ke(e)})},C4={kernelName:fo,inputsToSave:["a","b"],gradFunc:(e,t)=>{let[n,a]=t,r=pt(n.shape,a.shape);return{a:()=>{let s=e,i=Vt(n.shape,r);return i.length>0&&(s=be(s,i)),W(s,n.shape)},b:()=>{let s=e,i=Vt(a.shape,r);return i.length>0&&(s=be(s,i)),W(It(s),a.shape)}}}},E4={kernelName:co,inputsToSave:["x"],gradFunc:(e,t,n)=>{let[a]=t,r=a.shape.slice(),{axis:s}=n;Ea(s,a.shape).forEach(l=>{r[l]=1});let i=W(e,r),o=B(i,Qn(a.shape,"float32"));return{x:()=>o}}},F4={kernelName:go,inputsToSave:["x"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>fe(e,ut(Vc(n)))}}},A4={kernelName:yo,outputsToSave:[!0],gradFunc:(e,t)=>{let[n]=t;return{x:()=>B(ce(we(1),ut(n)),e)}}},$4={kernelName:vs,inputsToSave:["x"],gradFunc:(e,t,n)=>{let[a]=t,{reps:r}=n;return{x:()=>{let s=Ke(a);if(a.rank===1)for(let i=0;i<r[0];++i)s=J(s,He(e,[i*a.shape[0]],[a.shape[0]]));else if(a.rank===2)for(let i=0;i<r[0];++i)for(let o=0;o<r[1];++o)s=J(s,He(e,[i*a.shape[0],o*a.shape[1]],[a.shape[0],a.shape[1]]));else if(a.rank===3)for(let i=0;i<r[0];++i)for(let o=0;o<r[1];++o)for(let l=0;l<r[2];++l)s=J(s,He(e,[i*a.shape[0],o*a.shape[1],l*a.shape[2]],[a.shape[0],a.shape[1],a.shape[2]]));else if(a.rank===4)for(let i=0;i<r[0];++i)for(let o=0;o<r[1];++o)for(let l=0;l<r[2];++l)for(let u=0;u<r[3];++u)s=J(s,He(e,[i*a.shape[0],o*a.shape[1],l*a.shape[2],u*a.shape[3]],[a.shape[0],a.shape[1],a.shape[2],a.shape[3]]));else throw new Error(`Gradient for tile operation is not implemented for rank-${a.rank} tensors yet.`);return s}}}},D4={kernelName:Cr,gradFunc:(e,t,n)=>{let a=n,{perm:r}=a,s=fv(r);return{x:()=>Ae(e,s)}}},R4={kernelName:Ou,gradFunc:(e,t,n)=>{let a=n,{axis:r}=a;return{value:()=>Mt(e,r)}}},M4={kernelName:Dc,inputsToSave:["segmentIds"],gradFunc:(e,t)=>{let[n]=t;return{x:()=>P4(e,n)}}};function P4(e,t){let n=fr(t,Ke(t)),a=ci(e,n),r=Ns(t,we(0,"int32")),s=a.rank-r.rank;for(let o=0;o<s;++o)r=mn(r,o+1);r=Ca(r,Qn(a.shape,"bool"));let i=Ke(a);return fn(r,a,i)}var O4={kernelName:Lu,gradFunc:e=>({x:()=>Ke(e)})},L4=[S2,MB,PB,OB,LB,zB,BB,WB,VB,UB,GB,HB,KB,QB,JB,ZB,eW,tW,nW,aW,rW,sW,oW,iW,pW,cW,dW,hW,mW,fW,s4,gW,yW,bW,xW,vW,kW,wW,IW,SW,NW,TW,_W,CW,EW,FW,AW,$W,DW,PW,uk,uk,OW,BW,UW,GW,HW,jW,qW,KW,XW,YW,QW,JW,ZW,pk,pk,e4,t4,r4,i4,o4,l4,u4,p4,c4,d4,h4,m4,f4,g4,y4,b4,x4,v4,w4,k4,I4,ck,ck,dk,dk,S4,T4,N4,_4,C4,E4,F4,A4,$4,D4,R4,M4,O4];for(let e of L4)MI(e);ne().prototype.abs=function(){return this.throwIfDisposed(),zt(this)};ne().prototype.acos=function(){return this.throwIfDisposed(),Yx(this)};ne().prototype.acosh=function(){return this.throwIfDisposed(),Qx(this)};ne().prototype.add=function(e){return this.throwIfDisposed(),J(this,e)};ne().prototype.all=function(e,t){return this.throwIfDisposed(),sf(this,e,t)};ne().prototype.any=function(e,t){return this.throwIfDisposed(),rc(this,e,t)};ne().prototype.argMax=function(e){return this.throwIfDisposed(),li(this,e)};ne().prototype.argMin=function(e){return this.throwIfDisposed(),Jx(this,e)};ne().prototype.asScalar=function(){return this.throwIfDisposed(),R(this.size===1,()=>"The array must have only 1 element."),W(this,[])};ne().prototype.asType=function(e){return this.throwIfDisposed(),oe(this,e)};ne().prototype.as1D=function(){return this.throwIfDisposed(),W(this,[this.size])};ne().prototype.as2D=function(e,t){return this.throwIfDisposed(),W(this,[e,t])};ne().prototype.as3D=function(e,t,n){return this.throwIfDisposed(),W(this,[e,t,n])};ne().prototype.as4D=function(e,t,n,a){return this.throwIfDisposed(),W(this,[e,t,n,a])};ne().prototype.as5D=function(e,t,n,a,r){return this.throwIfDisposed(),W(this,[e,t,n,a,r])};ne().prototype.asin=function(){return this.throwIfDisposed(),Zx(this)};ne().prototype.asinh=function(){return this.throwIfDisposed(),ev(this)};ne().prototype.atan=function(){return this.throwIfDisposed(),tv(this)};ne().prototype.atan2=function(e){return this.throwIfDisposed(),nv(this,e)};ne().prototype.atanh=function(){return this.throwIfDisposed(),av(this)};ne().prototype.avgPool=function(e,t,n,a){return this.throwIfDisposed(),ga(this,e,t,n,a)};ne().prototype.batchToSpaceND=function(e,t){return this.throwIfDisposed(),Wc(this,e,t)};ne().prototype.batchNorm=function(e,t,n,a,r){return this.throwIfDisposed(),$r(this,e,t,n,a,r)};ne().prototype.broadcastTo=function(e){return this.throwIfDisposed(),xl(this,e)};ne().prototype.cast=function(e){return this.throwIfDisposed(),oe(this,e)};ne().prototype.ceil=function(){return this.throwIfDisposed(),ov(this)};ne().prototype.clipByValue=function(e,t){return this.throwIfDisposed(),an(this,e,t)};ne().prototype.concat=function(e,t){return this.throwIfDisposed(),e instanceof Fe&&(e=[e]),Ze([this,...e],t)};ne().prototype.conv1d=function(e,t,n,a,r,s){return this.throwIfDisposed(),of(this,e,t,n,a,r,s)};ne().prototype.conv2dTranspose=function(e,t,n,a,r){return this.throwIfDisposed(),lf(this,e,t,n,a,r)};ne().prototype.conv2d=function(e,t,n,a,r,s){return this.throwIfDisposed(),Rt(this,e,t,n,a,r,s)};ne().prototype.cos=function(){return this.throwIfDisposed(),Vc(this)};ne().prototype.cosh=function(){return this.throwIfDisposed(),uf(this)};ne().prototype.cumprod=function(e,t,n){return this.throwIfDisposed(),sc(this,e,t,n)};ne().prototype.cumsum=function(e,t,n){return this.throwIfDisposed(),pf(this,e,t,n)};ne().prototype.depthToSpace=function(e,t){return this.throwIfDisposed(),pv(this,e,t)};ne().prototype.depthwiseConv2d=function(e,t,n,a,r,s){return this.throwIfDisposed(),Ss(this,e,t,n,a,r,s)};ne().prototype.dilation2d=function(e,t,n,a,r){return this.throwIfDisposed(),cv(this,e,t,n,a,r)};ne().prototype.divNoNan=function(e){return this.throwIfDisposed(),dv(this,e)};ne().prototype.div=function(e){return this.throwIfDisposed(),fe(this,e)};ne().prototype.dot=function(e){return this.throwIfDisposed(),BS(this,e)};ne().prototype.elu=function(){return this.throwIfDisposed(),Wu(this)};ne().prototype.equal=function(e){return this.throwIfDisposed(),Zn(this,e)};ne().prototype.erf=function(){return this.throwIfDisposed(),hv(this)};ne().prototype.euclideanNorm=function(e,t){return this.throwIfDisposed(),gv(this,e,t)};ne().prototype.exp=function(){return this.throwIfDisposed(),gn(this)};ne().prototype.expandDims=function(e){return this.throwIfDisposed(),mn(this,e)};ne().prototype.expm1=function(){return this.throwIfDisposed(),yv(this)};ne().prototype.fft=function(){return this.throwIfDisposed(),Xc(this)};ne().prototype.flatten=function(){return this.throwIfDisposed(),W(this,[this.size])};ne().prototype.floor=function(){return this.throwIfDisposed(),Vu(this)};ne().prototype.floorDiv=function(e){return this.throwIfDisposed(),rf(this,e)};ne().prototype.gather=function(e,t){return this.throwIfDisposed(),ci(this,e,t)};ne().prototype.greaterEqual=function(e){return this.throwIfDisposed(),Ns(this,e)};ne().prototype.greater=function(e){return this.throwIfDisposed(),Un(this,e)};ne().prototype.ifft=function(){return this.throwIfDisposed(),_l(this)};ne().prototype.irfft=function(){return this.throwIfDisposed(),Sf(this)};ne().prototype.isFinite=function(){return this.throwIfDisposed(),jS(this)};ne().prototype.isInf=function(){return this.throwIfDisposed(),qS(this)};ne().prototype.isNaN=function(){return this.throwIfDisposed(),xv(this)};ne().prototype.leakyRelu=function(e){return this.throwIfDisposed(),Gc(this,e)};ne().prototype.lessEqual=function(e){return this.throwIfDisposed(),Ts(this,e)};ne().prototype.less=function(e){return this.throwIfDisposed(),cf(this,e)};ne().prototype.localResponseNormalization=function(e,t,n,a){return this.throwIfDisposed(),vv(this,e,t,n,a)};ne().prototype.logSigmoid=function(){return this.throwIfDisposed(),YS(this)};ne().prototype.logSoftmax=function(e){return this.throwIfDisposed(),hf(this,e)};ne().prototype.logSumExp=function(e,t){return this.throwIfDisposed(),wv(this,e,t)};ne().prototype.log=function(){return this.throwIfDisposed(),ea(this)};ne().prototype.log1p=function(){return this.throwIfDisposed(),Hc(this)};ne().prototype.logicalAnd=function(e){return this.throwIfDisposed(),Ca(this,e)};ne().prototype.logicalNot=function(){return this.throwIfDisposed(),jc(this)};ne().prototype.logicalOr=function(e){return this.throwIfDisposed(),mf(this,e)};ne().prototype.logicalXor=function(e){return this.throwIfDisposed(),QS(this,e)};ne().prototype.matMul=function(e,t,n){return this.throwIfDisposed(),De(this,e,t,n)};ne().prototype.maxPool=function(e,t,n,a){return this.throwIfDisposed(),Pt(this,e,t,n,a)};ne().prototype.max=function(e,t){return this.throwIfDisposed(),Ta(this,e,t)};ne().prototype.maximum=function(e){return this.throwIfDisposed(),fr(this,e)};ne().prototype.mean=function(e,t){return this.throwIfDisposed(),Et(this,e,t)};ne().prototype.min=function(e,t){return this.throwIfDisposed(),ic(this,e,t)};ne().prototype.minimum=function(e){return this.throwIfDisposed(),Uu(this,e)};ne().prototype.mirrorPad=function(e,t){return this.throwIfDisposed(),Sv(this,e,t)};ne().prototype.mod=function(e){return this.throwIfDisposed(),Nv(this,e)};ne().prototype.mul=function(e){return this.throwIfDisposed(),B(this,e)};ne().prototype.neg=function(){return this.throwIfDisposed(),It(this)};ne().prototype.norm=function(e,t,n){return this.throwIfDisposed(),Uc(this,e,t,n)};ne().prototype.notEqual=function(e){return this.throwIfDisposed(),di(this,e)};ne().prototype.oneHot=function(e,t=1,n=0){return this.throwIfDisposed(),Sl(this,e,t,n)};ne().prototype.onesLike=function(){return this.throwIfDisposed(),ta(this)};ne().prototype.pad=function(e,t){return this.throwIfDisposed(),ya(this,e,t)};ne().prototype.pool=function(e,t,n,a,r,s){return this.throwIfDisposed(),t2(this,e,t,n,a,r,s)};ne().prototype.pow=function(e){return this.throwIfDisposed(),Dr(this,e)};ne().prototype.prelu=function(e){return this.throwIfDisposed(),Kc(this,e)};ne().prototype.prod=function(e,t){return this.throwIfDisposed(),gf(this,e,t)};ne().prototype.reciprocal=function(){return this.throwIfDisposed(),Cv(this)};ne().prototype.relu=function(){return this.throwIfDisposed(),Xe(this)};ne().prototype.relu6=function(){return this.throwIfDisposed(),yf(this)};ne().prototype.reshapeAs=function(e){return this.throwIfDisposed(),W(this,e.shape)};ne().prototype.reshape=function(e){return this.throwIfDisposed(),W(this,e)};ne().prototype.resizeBilinear=function(e,t,n){return this.throwIfDisposed(),b2(this,e,t,n)};ne().prototype.resizeNearestNeighbor=function(e,t,n){return this.throwIfDisposed(),x2(this,e,t,n)};ne().prototype.reverse=function(e){return this.throwIfDisposed(),na(this,e)};ne().prototype.rfft=function(){return this.throwIfDisposed(),Yc(this)};ne().prototype.round=function(){return this.throwIfDisposed(),bf(this)};ne().prototype.rsqrt=function(){return this.throwIfDisposed(),xf(this)};ne().prototype.selu=function(){return this.throwIfDisposed(),vf(this)};ne().prototype.separableConv2d=function(e,t,n,a,r,s){return this.throwIfDisposed(),vo(this,e,t,n,a,r,s)};ne().prototype.sigmoid=function(){return this.throwIfDisposed(),ma(this)};ne().prototype.sign=function(){return this.throwIfDisposed(),Ev(this)};ne().prototype.sin=function(){return this.throwIfDisposed(),wf(this)};ne().prototype.sinh=function(){return this.throwIfDisposed(),kf(this)};ne().prototype.slice=function(e,t){return this.throwIfDisposed(),He(this,e,t)};ne().prototype.softmax=function(e){return this.throwIfDisposed(),Qa(this,e)};ne().prototype.softplus=function(){return this.throwIfDisposed(),xo(this)};ne().prototype.spaceToBatchND=function(e,t){return this.throwIfDisposed(),qc(this,e,t)};ne().prototype.split=function(e,t){return this.throwIfDisposed(),zn(this,e,t)};ne().prototype.sqrt=function(){return this.throwIfDisposed(),pn(this)};ne().prototype.square=function(){return this.throwIfDisposed(),ut(this)};ne().prototype.squaredDifference=function(e){return this.throwIfDisposed(),Nf(this,e)};ne().prototype.squeeze=function(e){return this.throwIfDisposed(),dr(this,e)};ne().prototype.stack=function(e,t){this.throwIfDisposed();let n=e instanceof Fe?[this,e]:[this,...e];return Mt(n,t)};ne().prototype.step=function(e){return this.throwIfDisposed(),ju(this,e)};ne().prototype.stridedSlice=function(e,t,n,a,r,s,i,o){return this.throwIfDisposed(),Av(this,e,t,n,a,r,s,i,o)};ne().prototype.sub=function(e){return this.throwIfDisposed(),ce(this,e)};ne().prototype.sum=function(e,t){return this.throwIfDisposed(),be(this,e,t)};ne().prototype.tan=function(){return this.throwIfDisposed(),$v(this)};ne().prototype.tanh=function(){return this.throwIfDisposed(),ui(this)};ne().prototype.tile=function(e){return this.throwIfDisposed(),On(this,e)};ne().prototype.toBool=function(){return this.throwIfDisposed(),oe(this,"bool")};ne().prototype.toFloat=function(){return this.throwIfDisposed(),oe(this,"float32")};ne().prototype.toInt=function(){return this.throwIfDisposed(),oe(this,"int32")};ne().prototype.topk=function(e,t){return this.throwIfDisposed(),Dv(this,e,t)};ne().prototype.transpose=function(e){return this.throwIfDisposed(),Ae(this,e)};ne().prototype.unique=function(e){return this.throwIfDisposed(),Zh(this,e)};ne().prototype.unsortedSegmentSum=function(e,t){return this.throwIfDisposed(),Rv(this,e,t)};ne().prototype.unstack=function(e){return this.throwIfDisposed(),ht(this,e)};ne().prototype.where=function(e,t){return this.throwIfDisposed(),fn(e,this,t)};ne().prototype.zerosLike=function(){return this.throwIfDisposed(),Ke(this)};var Ir=class extends Error{constructor(e){super(e),Object.setPrototypeOf(this,Ir.prototype)}},Va=class extends Error{constructor(e){super(e),Object.setPrototypeOf(this,Va.prototype)}},H=class extends Error{constructor(e){super(e),Object.setPrototypeOf(this,H.prototype)}},Oe=class extends Error{constructor(e){super(e),Object.setPrototypeOf(this,Oe.prototype)}},T2=class extends Error{constructor(e){super(e),Object.setPrototypeOf(this,T2.prototype)}},_2=class{constructor(e){this.maxEntries=e||100,this.cache=new Map}get(e){let t;return this.cache.has(e)&&(t=this.cache.get(e),this.cache.delete(e),this.cache.set(e,t)),t}put(e,t){if(this.cache.has(e))this.cache.delete(e);else if(this.cache.size>=this.maxEntries){let n=this.cache.keys().next().value;this.cache.delete(n)}this.cache.set(e,t)}getMaxEntries(){return this.maxEntries}setMaxEntries(e){if(e<0)throw new Error(`The maxEntries of LRU caches must be at least 0, but got ${e}.`);if(this.maxEntries>e)for(let t=0;t<this.maxEntries-e;t++){let n=this.cache.keys().next().value;this.cache.delete(n)}this.maxEntries=e}};function hi(e,t){if(Array.isArray(e)){let n=[];for(let a=0;a<t;a++)n=n.concat(e);return n}else{let n=new Array(t);return n.fill(e),n}}function ir(e,t){if(!e)throw new T2(t)}function hk(e,t){let n=0;for(let a of e)a===t&&n++;return n}function Pn(e){return e.length===1?e[0]:e}function kt(e){return Array.isArray(e)?e:[e]}function Sr(e){let t=e.replace(/(.)([A-Z][a-z0-9]+)/g,"$1_$2").replace(/([a-z])([A-Z])/g,"$1_$2").toLowerCase();return t[0]!=="_"?t:"private"+t}function Ks(e){return e.length<=1||e.indexOf("_")===-1?e:e.replace(/[_]+(\w|$)/g,(t,n)=>n.toUpperCase())}var ka={};function Vv(e){if(e==null)return null;let t={};return t.className=e.getClassName(),t.config=e.getConfig(),t}function qb(e){if(!(e==null||typeof e!="object"))if(Array.isArray(e))e.forEach(t=>qb(t));else{let t=Object.keys(e);for(let n of t){let a=e[n];a!=null&&typeof a=="object"&&(!Array.isArray(a)&&a.type==="ndarray"&&typeof a.value=="number"?e[n]=a.value:qb(a))}}}function Jc(e,t={},n={},a="object",r=!1){if(typeof e=="string"){let s=e,i;if(s in n)i=n[s];else if(s in ka)i=ka[s];else if(i=t[s],i==null)throw new H(`Unknown ${a}: ${e}. This may be due to one of the following reasons:
1. The ${a} is defined in Python, in which case it needs to be ported to TensorFlow.js or your JavaScript code.
2. The custom ${a} is defined in JavaScript, but is not registered properly with tf.serialization.registerClass().`);return i}else{let s=e;if(s.className==null||s.config==null)throw new H(`${a}: Improper config format: ${JSON.stringify(s)}.
'className' and 'config' must set.`);let i=s.className,o,l;if(i in n?[o,l]=n[i]:i in ka?[o,l]=ka.className:i in t&&([o,l]=t[i]),o==null)throw new H(`Unknown ${a}: ${i}. This may be due to one of the following reasons:
1. The ${a} is defined in Python, in which case it needs to be ported to TensorFlow.js or your JavaScript code.
2. The custom ${a} is defined in JavaScript, but is not registered properly with tf.serialization.registerClass().`);if(l!=null){let u={};for(let h of Object.keys(ka))u[h]=ka[h];for(let h of Object.keys(n))u[h]=n[h];let p=s.config;p.customObjects=u;let d=Object.assign({},ka);for(let h of Object.keys(n))ka[h]=n[h];qb(s.config);let c=l(o,s.config,n,r);return ka=Object.assign({},d),c}else{let u=Object.assign({},ka);for(let d of Object.keys(n))ka[d]=n[d];let p=new o(s.config);return ka=Object.assign({},u),p}}}function z4(e,t){return e<t?-1:e>t?1:0}function vh(e,t){return-1*z4(e,t)}function ss(e){if(e==null)return e;let t=[];for(let n of e)t.indexOf(n)===-1&&t.push(n);return t}function B4(e){if(e==null)throw new H(`Invalid value in obj: ${JSON.stringify(e)}`);for(let t in e)if(e.hasOwnProperty(t))return!1;return!0}function wo(e,t,n){if(n!=null&&e.indexOf(n)<0)throw new H(`${n} is not a valid ${t}. Valid values are ${e} or null/undefined.`)}function Uv(e,t,n=0,a=1/0){return ir(n>=0),ir(a>=n),Array.isArray(e)&&e.length>=n&&e.length<=a&&e.every(r=>typeof r===t)}function nn(e,t){Array.isArray(e)?(w.assert(e.length>0,()=>`${t} is unexpectedly an empty array.`),e.forEach((n,a)=>nn(n,`element ${a+1} of ${t}`))):w.assert(Number.isInteger(e)&&e>0,()=>`Expected ${t} to be a positive integer, but got ${C2(e)}.`)}function C2(e){return e===null?"null":Array.isArray(e)?"["+e.map(t=>C2(t)).join(",")+"]":typeof e=="string"?`"${e}"`:`${e}`}function W4(e,t,n){let a=n!=null?n():w.now(),r;return(...s)=>{let i=n!=null?n():w.now();return i-a<t||(a=i,r=e(...s)),r}}function E2(e){return e==="relu"?"relu":e==="linear"?"linear":e==="elu"?"elu":null}var V4=0;function F2(){return V4++}var wh={};function Of(e=""){return e in wh||(wh[e]=0),wh[e]+=1,e+wh[e].toString()}var U4=["channelsFirst","channelsLast"],G4=["nearest","bilinear"],H4=["valid","same","causal"],j4=["max","avg"],q4=["sum","mul","concat","ave"],ul=new Map;function Ot(e){wo(U4,"DataFormat",e)}function K4(e){wo(G4,"InterpolationFormat",e)}function ba(e){wo(H4,"PaddingMode",e)}function A2(e){wo(j4,"PoolMode",e)}var Yp=[],mk="/";function ei(e,t){Yp.push(e);try{let n=t();return Yp.pop(),n}catch(n){throw Yp.pop(),n}}function X4(){return Yp.length===0?"":Yp.join(mk)+mk}function $2(e){if(!R2(e))throw new Error("Not a valid tensor name: '"+e+"'");return X4()+e}function D2(e){if(!R2(e))throw new Error("Not a valid tensor name: '"+e+"'");ul.has(e)||ul.set(e,0);let t=ul.get(e);if(ul.set(e,ul.get(e)+1),t>0){let n=`${e}_${t}`;return ul.set(n,1),n}else return e}var Y4=new RegExp(/^[A-Za-z0-9][-A-Za-z0-9\._\/]*$/);function R2(e){return!!e.match(Y4)}function Q4(e){return e===parseInt(e.toString(),10)}function is(e,t,n){t==null&&(t=0),n==null&&(n=e.length);let a=1;for(let r=t;r<n;++r)a*=e[r];return a}function Cl(e){if(e.length===0)return Number.NaN;let t=Number.POSITIVE_INFINITY;for(let n=0;n<e.length;n++){let a=e[n];a<t&&(t=a)}return t}function cs(e){if(e.length===0)return Number.NaN;let t=Number.NEGATIVE_INFINITY;for(let n=0;n<e.length;n++){let a=e[n];a>t&&(t=a)}return t}function Ka(e,t){if(t<e)throw new H(`end (${t}) < begin (${e}) is forbidden.`);let n=[];for(let a=e;a<t;++a)n.push(a);return n}var kb;function jt(){return kb==null&&(kb=oS().epsilon()),kb}function Xa(){return"channelsLast"}function Lf(e,t){return oe(e,t)}function Zc(e,t=-1){let n=e.shape.slice();return t<0&&(t=n.length+t+1),n.splice(t,0,1),W(e,n)}function J4(e,t){return O(()=>{if(e.shape.length!==2)throw new H(`repeat() expects a rank-2 tensor, but received a rank-${e.shape.length} tensor.`);let n=Zc(e,1);return Kb(n,[1,t,1])})}function Z4(e){let t=[is(e.shape)];return W(e,t)}function eV(e){if(e.rank<=1)throw new H(`batchFlatten requires a minimum rank of 2. Got rank: ${e.rank}.`);let t=[e.shape[0],is(e.shape,1)];return W(e,t)}function ti(e,t,n){return O(()=>{switch(e.rank){case 1:return If(e,t,n);case 2:return Fv(e,[t,0],[n,e.shape[1]]);case 3:return Hu(e,[t,0,0],[n,e.shape[1],e.shape[2]]);case 4:return oc(e,[t,0,0,0],[n,e.shape[1],e.shape[2],e.shape[3]]);case 5:return He(e,[t,0,0,0,0],[n,e.shape[1],e.shape[2],e.shape[3],e.shape[4]]);case 6:return He(e,[t,0,0,0,0,0],[n,e.shape[1],e.shape[2],e.shape[3],e.shape[4],e.shape[5]]);default:throw new H(`sliceAlongFirstAxis() received an unsupported tensor rank: ${e.rank}`)}})}function Ib(e,t,n){return O(()=>{switch(e.rank){case 1:return If(e,t,n);case 2:return Fv(e,[0,t],[e.shape[0],n]);case 3:return Hu(e,[0,0,t],[e.shape[0],e.shape[1],n]);case 4:return oc(e,[0,0,0,t],[e.shape[0],e.shape[1],e.shape[2],n]);default:throw new H(`sliceAlongLastAxis() received an unsupported tensor rank: ${e.rank}`)}})}function kh(e,t,n,a){return O(()=>{switch(e.rank){case 1:return If(e,t,n);case 2:switch(a){case 1:return ti(e,t,n);case 2:return Ib(e,t,n);default:throw new H(`The axis is not within the rank of the tensor ${a}`)}case 3:switch(a){case 1:return ti(e,t,n);case 2:return Hu(e,[0,t,0],[e.shape[0],n,e.shape[2]]);case 3:return Ib(e,t,n);default:throw new H(`The axis is not within the rank of the tensor ${a}`)}case 4:switch(a){case 1:return ti(e,t,n);case 2:return oc(e,[0,t,0,0],[e.shape[0],n,e.shape[2],e.shape[3]]);case 3:return oc(e,[0,0,t,0],[e.shape[0],e.shape[1],n,e.shape[3]]);case 4:return Ib(e,t,n);default:throw new H(`The axis is not within the rank of the tensor ${a}`)}default:throw new H(`sliceAlongLastAxis() received an unsupported tensor rank: ${e.rank}`)}})}function Gv(e,t=-1){let n;return t<0&&(n=e[0].rank,n!==0?t=n:t=0),t===e[0].rank&&(t=-1),Ze(e,t)}function fk(e,t){switch(e.rank){case 1:return DS([e,t]);case 2:return RS([e,t],0);case 3:return MS([e,t],0);case 4:return PS([e,t],0);default:throw new H(`concatAlongFirstAxis() received an unsupported tensor rank: ${e.rank}`)}}function Kb(e,t){if(Array.isArray(t)||(t=[t]),e.rank!==t.length)throw new H(`The length of input n (${t.length}) does not match the number of dimensions in input x (${e.rank})`);return On(e,t)}function zf(e,t=0,n=1,a,r){return n2(e,t,n,a,r)}function ur(e,t,n,a){if(e.rank<2||t.rank<2)throw new Oe(`dot requires both inputs to be rank >= 2 but got x shape = ${e.shape} and y shape = ${t.shape}`);if(t.rank>=3){let r=e.shape.slice(-1)[0],s=t.shape.slice(-2)[0];if(r!==s)throw new Oe(`If rank y >= 3, then the second last dim of y must equal the last dim of x but got x shape = ${e.shape} and y shape = ${t.shape}`)}if(e.rank===2&&t.rank===2)return ps.matMul({a:e,b:t,transposeA:!1,transposeB:!1,bias:a?Xb(e.rank,a,Xa()):null,activation:n});{let r=e.shape.slice(),s=r.pop();e=W(e,[-1,s]);let i=t.shape.slice(),o=i.pop(),l=i.pop(),u=[...i,o],p=Array.from({length:t.rank},(m,f)=>f===0?t.rank-2:f<=t.rank-2?f-1:f);t=W(Ae(t,p),[l,-1]);let d=[...r,...u],c=!1,h=!1;return W(ps.matMul({a:e,b:t,transposeA:c,transposeB:h,bias:a?Xb(e.rank,a,Xa()):null,activation:n}),d)}}function M2(e,t,n){return O(()=>(Array.isArray(t)?t=qe(t,"int32"):t=oe(t,"int32"),ci(e,t,n)))}function ed(e){return B(e,e)}function Xb(e,t,n){let a=t.shape;if(t.rank!==1&&t.rank!==e)throw new H(`Unexpected bias dimensions: ${t.rank}; expected it to be 1 or ${e}`);if(e===5){if(n==="channelsFirst")return a.length===1?W(t,[1,a[0],1,1,1]):W(t,[1,a[3],a[0],a[1],a[2]]);if(n==="channelsLast")return a.length===1?W(t,[1,1,1,1,a[0]]):W(t,[1].concat(a))}else if(e===4){if(n==="channelsFirst")return a.length===1?W(t,[1,a[0],1,1]):W(t,[1,a[2],a[0],a[1]]);if(n==="channelsLast")return a.length===1?W(t,[1,1,1,a[0]]):W(t,[1].concat(a))}else if(e===3){if(n==="channelsFirst")return a.length===1?W(t,[1,a[0],1]):W(t,[1,a[1],a[0]]);if(n==="channelsLast")return a.length===1?W(t,[1,1,a[0]]):W(t,[1].concat(a))}else if(e<3)return t;throw new H(`Unsupported input rank by biasAdd: ${t.rank}`)}function Za(e,t,n){return O(()=>(n==null&&(n=Xa()),Ot(n),J(e,Xb(e.rank,t,n))))}function tV(e,t=1){if(t!==1)throw new Oe(`Support for alpha values other than 1 (${t}) is not implemented yet.`);return Wu(e)}function nV(e){return O(()=>fe(e,J(zt(e),1)))}function P2(e,t,n,a){return O(()=>u2(e,t,n,a))}function aV(e){return O(()=>{let t=J(.5,B(.2,e));return an(t,0,1)})}function td(e,t,n=!1){return n?e():t()}var rV=["fanIn","fanOut","fanAvg"],sV=["normal","uniform","truncatedNormal"];function iV(e){wo(rV,"FanMode",e)}function oV(e){wo(sV,"Distribution",e)}var Fa=class extends se.Serializable{fromConfigUsesCustomObjects(){return!1}getConfig(){return{}}},Hv=class extends Fa{apply(e,t){return St(e,t)}};Hv.className="Zeros";se.registerClass(Hv);var Bf=class extends Fa{apply(e,t){return Qn(e,t)}};Bf.className="Ones";se.registerClass(Bf);var jv=class extends Fa{constructor(e){if(super(),typeof e!="object")throw new H(`Expected argument of type ConstantConfig but got ${e}`);if(e.value===void 0)throw new H(`config must have value set but got ${e}`);this.value=e.value}apply(e,t){return O(()=>B(we(this.value),Qn(e,t)))}getConfig(){return{value:this.value}}};jv.className="Constant";se.registerClass(jv);var qv=class extends Fa{constructor(e){super(),this.DEFAULT_MINVAL=-.05,this.DEFAULT_MAXVAL=.05,this.minval=e.minval||this.DEFAULT_MINVAL,this.maxval=e.maxval||this.DEFAULT_MAXVAL,this.seed=e.seed}apply(e,t){return Gu(e,this.minval,this.maxval,t)}getConfig(){return{minval:this.minval,maxval:this.maxval,seed:this.seed}}};qv.className="RandomUniform";se.registerClass(qv);var Kv=class extends Fa{constructor(e){super(),this.DEFAULT_MEAN=0,this.DEFAULT_STDDEV=.05,this.mean=e.mean||this.DEFAULT_MEAN,this.stddev=e.stddev||this.DEFAULT_STDDEV,this.seed=e.seed}apply(e,t){if(t=t||"float32",t!=="float32"&&t!=="int32")throw new Oe(`randomNormal does not support dType ${t}.`);return zf(e,this.mean,this.stddev,t,this.seed)}getConfig(){return{mean:this.mean,stddev:this.stddev,seed:this.seed}}};Kv.className="RandomNormal";se.registerClass(Kv);var Xv=class extends Fa{constructor(e){super(),this.DEFAULT_MEAN=0,this.DEFAULT_STDDEV=.05,this.mean=e.mean||this.DEFAULT_MEAN,this.stddev=e.stddev||this.DEFAULT_STDDEV,this.seed=e.seed}apply(e,t){if(t=t||"float32",t!=="float32"&&t!=="int32")throw new Oe(`truncatedNormal does not support dType ${t}.`);return Tf(e,this.mean,this.stddev,t,this.seed)}getConfig(){return{mean:this.mean,stddev:this.stddev,seed:this.seed}}};Xv.className="TruncatedNormal";se.registerClass(Xv);var Yv=class extends Fa{constructor(e){super(),this.gain=e.gain!=null?e.gain:1}apply(e,t){return O(()=>{if(e.length!==2||e[0]!==e[1])throw new H("Identity matrix initializer can only be used for 2D square matrices.");return B(this.gain,bv(e[0]))})}getConfig(){return{gain:this.gain}}};Yv.className="Identity";se.registerClass(Yv);function lV(e,t="channelsLast"){let n,a;if(Ot(t),e.length===2)n=e[0],a=e[1];else if([3,4,5].indexOf(e.length)!==-1){if(t==="channelsFirst"){let r=is(e,2);n=e[1]*r,a=e[0]*r}else if(t==="channelsLast"){let r=is(e,0,e.length-2);n=e[e.length-2]*r,a=e[e.length-1]*r}}else{let r=is(e);n=Math.sqrt(r),a=Math.sqrt(r)}return[n,a]}var Wn=class extends Fa{constructor(e){if(super(),e.scale<0)throw new H(`scale must be a positive float. Got: ${e.scale}`);this.scale=e.scale==null?1:e.scale,this.mode=e.mode==null?"fanIn":e.mode,iV(this.mode),this.distribution=e.distribution==null?"normal":e.distribution,oV(this.distribution),this.seed=e.seed}apply(e,t){let n=lV(e),a=n[0],r=n[1],s=this.scale;if(this.mode==="fanIn"?s/=Math.max(1,a):this.mode==="fanOut"?s/=Math.max(1,r):s/=Math.max(1,(a+r)/2),this.distribution==="normal"){let i=Math.sqrt(s);if(t=t||"float32",t!=="float32"&&t!=="int32")throw new Oe(`${this.getClassName()} does not support dType ${t}.`);return Tf(e,0,i,t,this.seed)}else{let i=Math.sqrt(3*s);return Gu(e,-i,i,t)}}getConfig(){return{scale:this.scale,mode:this.mode,distribution:this.distribution,seed:this.seed}}};Wn.className="VarianceScaling";se.registerClass(Wn);var Wf=class extends Wn{constructor(e){super({scale:1,mode:"fanAvg",distribution:"uniform",seed:e==null?null:e.seed})}getClassName(){return Wn.className}};Wf.className="GlorotUniform";se.registerClass(Wf);var Vf=class extends Wn{constructor(e){super({scale:1,mode:"fanAvg",distribution:"normal",seed:e==null?null:e.seed})}getClassName(){return Wn.className}};Vf.className="GlorotNormal";se.registerClass(Vf);var Uf=class extends Wn{constructor(e){super({scale:2,mode:"fanIn",distribution:"normal",seed:e==null?null:e.seed})}getClassName(){return Wn.className}};Uf.className="HeNormal";se.registerClass(Uf);var Gf=class extends Wn{constructor(e){super({scale:2,mode:"fanIn",distribution:"uniform",seed:e==null?null:e.seed})}getClassName(){return Wn.className}};Gf.className="HeUniform";se.registerClass(Gf);var Hf=class extends Wn{constructor(e){super({scale:1,mode:"fanIn",distribution:"normal",seed:e==null?null:e.seed})}getClassName(){return Wn.className}};Hf.className="LeCunNormal";se.registerClass(Hf);var jf=class extends Wn{constructor(e){super({scale:1,mode:"fanIn",distribution:"uniform",seed:e==null?null:e.seed})}getClassName(){return Wn.className}};jf.className="LeCunNormal";se.registerClass(jf);var Qv=class extends Fa{constructor(e){if(super(),this.DEFAULT_GAIN=1,this.gain=e.gain==null?this.DEFAULT_GAIN:e.gain,this.seed=e.seed,this.seed!=null)throw new Oe("Random seed is not implemented for Orthogonal Initializer yet.")}apply(e,t){return O(()=>{if(e.length<2)throw new Oe("Shape must be at least 2D.");e[0]*e[1]>2e3&&console.warn(`Orthogonal initializer is being called on a matrix with more than 2000 (${e[0]*e[1]}) elements: Slowness may result.`);let n=e[0]>e[1]?[e[1],e[0]]:e,a=zf(n,0,1,"float32"),r=v2.gramSchmidt(a);return e[0]>e[1]&&(r=Ae(r)),B(this.gain,r)})}getConfig(){return{gain:this.gain,seed:this.seed}}};Qv.className="Orthogonal";se.registerClass(Qv);var gk={constant:"Constant",glorotNormal:"GlorotNormal",glorotUniform:"GlorotUniform",heNormal:"HeNormal",heUniform:"HeUniform",identity:"Identity",leCunNormal:"LeCunNormal",leCunUniform:"LeCunUniform",ones:"Ones",orthogonal:"Orthogonal",randomNormal:"RandomNormal",randomUniform:"RandomUniform",truncatedNormal:"TruncatedNormal",varianceScaling:"VarianceScaling",zeros:"Zeros"};function yk(e,t={}){return Jc(e,se.SerializationMap.getMap().classNameMap,t,"initializer")}function Ft(e){return Vv(e)}function Nt(e){if(typeof e=="string"){let t=e in gk?gk[e]:e;if(t==="GlorotNormal")return new Vf;if(t==="GlorotUniform")return new Wf;if(t==="HeNormal")return new Uf;if(t==="HeUniform")return new Gf;if(t==="LeCunNormal")return new Hf;if(t==="LeCunUniform")return new jf;{let n={};return n.className=t,n.config={},yk(n)}}else return e instanceof Fa?e:yk(e)}function Yb(e){return Array.isArray(e)&&Array.isArray(e[0])}function em(e){return e.length===0?[]:Array.isArray(e[0])?e:[e]}function ze(e){let t;if(Array.isArray(e)){if(e.length!==1)throw new H(`Expected Tensor length to be 1; got ${e.length}`);t=e[0]}else t=e;return t}function it(e){if(Array.isArray(e)&&Array.isArray(e[0])){if(e.length===1)return e=e,e[0];throw new H(`Expected exactly 1 Shape; got ${e.length}`)}else return e}function tm(e){let t=0;for(let n of e)n.shape.length===0?t+=1:t+=n.shape.reduce((a,r)=>a*r);return t}var bk="Variable",O2=class{constructor(e,t="float32",n=bk,a=!0,r=null){this.dtype=t==null?"float32":t,this.shape=e.shape,this.id=F2(),n=n==null?bk:n,this.originalName=$2(n),this.name=D2(this.originalName),this.trainable_=a,this.constraint=r,this.val=s2(e,this.trainable_,this.name,this.dtype)}read(){return this.assertNotDisposed(),this.val}write(e){return this.assertNotDisposed(),uV(this.val,e),this.val.id!==e.id&&(this.val.assign(e),this.constraint!=null&&this.val.assign(this.constraint.apply(this.val))),this}dispose(){this.assertNotDisposed(),this.val.dispose()}assertNotDisposed(){if(this.val.isDisposed)throw new Error(`LayersVariable ${this.name} is already disposed.`)}get trainable(){return this.trainable_}set trainable(e){this.trainable_=e,this.val.trainable=e}};function uV(e,t){if(e.shape.toString()!==t.shape.toString())throw new Error("Shape mismatch: "+JSON.stringify(e.shape)+" vs. "+JSON.stringify(t.shape))}function Qb(e){return e.map(t=>t.read())}function Jv(e){e.forEach(t=>{t[0].write(t[1])})}var Bt=class{constructor(e){this.dtype=e.dtype,this.shape=e.shape,e.shape!=null?this.ndim=e.shape.length:this.ndim=e.ndim,this.maxNDim=e.maxNDim,this.minNDim=e.minNDim,this.axes=e.axes||{}}},Ua=class{constructor(e,t,n,a,r,s,i){this.dtype=e,this.shape=t,this.sourceLayer=n,this.inputs=a,this.callArgs=r,this.outputTensorIndex=i,this.id=F2(),s!=null&&(this.originalName=$2(s),this.name=D2(this.originalName)),this.rank=t.length}},pV=0,qf=class{constructor(e,t){this.callArgs=t,this.id=pV++,this.outboundLayer=e.outboundLayer,this.inboundLayers=e.inboundLayers,this.nodeIndices=e.nodeIndices,this.tensorIndices=e.tensorIndices,this.inputTensors=e.inputTensors,this.outputTensors=e.outputTensors,this.inputMasks=e.inputMasks,this.outputMasks=e.outputMasks,this.inputShapes=e.inputShapes,this.outputShapes=e.outputShapes;for(let n of e.inboundLayers)n!=null&&n.outboundNodes.push(this);e.outboundLayer.inboundNodes.push(this)}getConfig(){let e=[];for(let t of this.inboundLayers)t!=null?e.push(t.name):e.push(null);return{outboundLayer:this.outboundLayer?this.outboundLayer.name:null,inboundLayers:e,nodeIndices:this.nodeIndices,tensorIndices:this.tensorIndices}}},cV=0,Ye=class extends se.Serializable{constructor(e={}){super(),this._callHook=null,this._addedWeightNames=[],this._stateful=!1,this.id=cV++,this.activityRegularizer=null,this.inputSpec=null,this.supportsMasking=!1,this._trainableWeights=[],this._nonTrainableWeights=[],this._losses=[],this._updates=[],this._built=!1,this.inboundNodes=[],this.outboundNodes=[];let t=e.name;if(!t){let n=this.getClassName();t=Sr(n)+"_"+Of(n)}if(this.name=t,this.trainable_=e.trainable==null?!0:e.trainable,e.inputShape!=null||e.batchInputShape!=null){let n;if(e.batchInputShape!=null)n=e.batchInputShape;else if(e.inputShape!=null){let r=null;e.batchSize!=null&&(r=e.batchSize),n=[r].concat(e.inputShape)}this.batchInputShape=n;let a=e.dtype;a==null&&(a=e.inputDType),a==null&&(a="float32"),this.dtype=a}e.weights!=null?this.initialWeights=e.weights:this.initialWeights=null,this._refCount=null,this.fastWeightInitDuringBuild=!1}static nodeKey(e,t){return e.name+"_ib-"+t.toString()}getNodeAtIndex(e,t){if(this.inboundNodes.length===0)throw new Va(`The layer has never been called and thus has no defined ${t}.`);if(this.inboundNodes.length<=e)throw new H(`Asked to get ${t} at node ${e}, but the layer has only ${this.inboundNodes.length} inbound nodes.`);return this.inboundNodes[e]}getInputAt(e){return Pn(this.getNodeAtIndex(e,"input").inputTensors)}getOutputAt(e){return Pn(this.getNodeAtIndex(e,"output").outputTensors)}get input(){if(this.inboundNodes.length>1)throw new Ir(`Layer ${this.name} has multiple inbound nodes, hence the notion of "layer input" is ill-defined. Use \`getInputAt(nodeIndex)\` instead.`);if(this.inboundNodes.length===0)throw new Ir(`Layer ${this.name} is not connected, no input to return.`);return Pn(this.getNodeAtIndex(0,"input").inputTensors)}get output(){if(this.inboundNodes.length===0)throw new Ir(`Layer ${this.name} has no inbound nodes.`);if(this.inboundNodes.length>1)throw new Ir(`Layer ${this.name} has multiple inbound nodes, hence the notion of "layer output" is ill-defined. Use \`getOutputAt(nodeIndex)\` instead.`);return Pn(this.getNodeAtIndex(0,"output").outputTensors)}get losses(){return this._losses}calculateLosses(){return this.losses.map(e=>e())}get updates(){return this._updates}get built(){return this._built}set built(e){this._built=e}get trainable(){return this.trainable_}set trainable(e){this._trainableWeights.forEach(t=>t.trainable=e),this.trainable_=e}get trainableWeights(){return this.trainable_?this._trainableWeights.filter(e=>e.trainable):[]}set trainableWeights(e){this._trainableWeights=e}get nonTrainableWeights(){return this.trainable?this._trainableWeights.filter(e=>!e.trainable).concat(this._nonTrainableWeights):this._trainableWeights.concat(this._nonTrainableWeights)}set nonTrainableWeights(e){this._nonTrainableWeights=e}get weights(){return this.trainableWeights.concat(this.nonTrainableWeights)}get stateful(){return this._stateful}resetStates(){if(!this.stateful)throw new Error("Cannot call the resetStates() method of a non-stateful Layer object.")}assertInputCompatibility(e){if(e=kt(e),this.inputSpec==null||this.inputSpec.length===0)return;let t=kt(this.inputSpec);if(e.length!==t.length)throw new H(`Layer ${this.name} expects ${t.length} inputs, but it received ${e.length} input tensors. Input received: ${e}`);for(let n=0;n<e.length;n++){let a=e[n],r=t[n];if(r==null)continue;let s=a.rank;if(r.ndim!=null&&s!==r.ndim)throw new H(`Input ${n} is incompatible with layer ${this.name}: expected ndim=${r.ndim}, found ndim=${s}`);if(r.maxNDim!=null&&s>r.maxNDim)throw new H(`Input ${n} is incompatible with layer ${this.name}: expected max_ndim=${r.maxNDim}, found ndim=${s}`);if(r.minNDim!=null&&s<r.minNDim)throw new H(`Input ${n} is incompatible with layer ${this.name}: expected min_ndim=${r.minNDim}, found ndim=${s}.`);if(r.dtype!=null&&a.dtype!==r.dtype)throw new H(`Input ${n} is incompatible with layer ${this.name} : expected dtype=${r.dtype}, found dtype=${a.dtype}.`);if(r.axes){let i=a.shape;for(let o in r.axes){let l=Number(o),u=r.axes[o],p=l>=0?i[l]:i[i.length+l];if(u!=null&&[u,null].indexOf(p)===-1)throw new H(`Input ${n} is incompatible with layer ${this.name}: expected axis ${l} of input shape to have value ${u} but got shape ${i}.`)}}if(r.shape!=null)for(let i=0;i<r.shape.length;++i){let o=r.shape[i],l=a.shape[i];if(o!=null&&l!=null&&o!==l)throw new H(`Input ${n} is incompatible with layer ${this.name}: expected shape=${r.shape}, found shape=${a.shape}.`)}}}call(e,t){return e}invokeCallHook(e,t){this._callHook!=null&&this._callHook(e,t)}setCallHook(e){this._callHook=e}clearCallHook(){this._callHook=null}apply(e,t){t=t||{},this.assertNotDisposed();let n=kt(e),a=!0;for(let s of n)if(!(s instanceof Ua)){a=!1;break}let r=!0;for(let s of n)if(s instanceof Ua){r=!1;break}if(a===r)throw new H("Arguments to apply() must be all SymbolicTensors or all Tensors");return ei(this.name,()=>{if(!this.built){this.assertInputCompatibility(e);let s=[];for(let i of kt(e))s.push(i.shape);this.build(Pn(s)),this.built=!0,this.initialWeights&&this.setWeights(this.initialWeights),this._refCount===null&&r&&(this._refCount=1)}if(this.assertInputCompatibility(e),r){let s=this.call(e,t),i=kt(s),o=[];for(let l of i)n.indexOf(l)!==-1&&(l=l.clone()),o.push(l);if(s=Pn(o),this.activityRegularizer!=null)throw new Oe("Layer invocation in the presence of activity regularizer(s) is not supported yet.");return s}else{let s=dV(e),i=this.computeOutputShape(s),o,l=hV(e);if(this.warnOnIncompatibleInputShape(Array.isArray(e)?s[0]:s),i!=null&&i.length>0&&Array.isArray(i[0])?o=i.map((u,p)=>new Ua(l,u,this,kt(e),t,this.name,p)):o=new Ua(l,i,this,kt(e),t,this.name),this.addInboundNode(e,o,null,null,s,i,t),this._refCount++,this.activityRegularizer!=null)throw new Oe("Layer invocation in the presence of activity regularizer(s) is not supported yet.");return o}})}warnOnIncompatibleInputShape(e){if(this.batchInputShape!=null)if(e.length!==this.batchInputShape.length)console.warn(`The rank of the input tensor provided (shape: ${JSON.stringify(e)}) does not match that of the batchInputShape (${JSON.stringify(this.batchInputShape)}) of the layer ${this.name}`);else{let t=!1;this.batchInputShape.forEach((n,a)=>{n!=null&&e[a]!=null&&e[a]!==n&&(t=!0)}),t&&console.warn(`The shape of the input tensor (${JSON.stringify(e)}) does not match the expectation of layer ${this.name}: ${JSON.stringify(this.batchInputShape)}`)}}get outputShape(){if(this.inboundNodes==null||this.inboundNodes.length===0)throw new Ir(`The layer ${this.name} has never been called and thus has no defined output shape.`);let e=[];for(let t of this.inboundNodes){let n=JSON.stringify(t.outputShapes);e.indexOf(n)===-1&&e.push(n)}if(e.length===1){let t=this.inboundNodes[0].outputShapes;return Array.isArray(t)&&Array.isArray(t[0])&&t.length===1?t[0]:t}else throw new Ir(`The layer ${this.name} has multiple inbound nodes with different output shapes. Hence the notion of "output shape" is ill-defined for the layer.`)}countParams(){if(!this.built)throw new Va(`You tried to call countParams() on ${this.name}, but the layer is not built yet. Build it first by calling build(batchInputShape).`);return tm(this.weights)}build(e){this.built=!0}getWeights(e=!1){return Qb(e?this.trainableWeights:this.weights)}setWeights(e){O(()=>{let t=this.weights;if(t.length!==e.length)throw new H(`You called setWeights(weights) on layer "${this.name}" with a weight list of length ${e.length}, but the layer was expecting ${t.length} weights. Provided weights: ${e}...`);if(t.length===0)return;let n=[],a=Qb(t);for(let r=0;r<a.length;++r){let s=a[r],i=t[r],o=e[r];if(!w.arraysEqual(s.shape,o.shape))throw new H(`Layer weight shape ${s.shape} not compatible with provided weight shape ${o.shape}`);n.push([i,o])}Jv(n)})}addWeight(e,t,n,a,r,s,i,o){if(this._addedWeightNames.indexOf(e)!==-1)throw new H(`Duplicate weight name ${e} for layer ${this.name}`);this._addedWeightNames.push(e),n==null&&(n="float32"),this.fastWeightInitDuringBuild&&(a=o!=null?o():Nt("zeros"));let l=a.apply(t,n),u=new O2(l,n,e,s,i);return l.dispose(),r!=null&&this.addLoss(()=>r.apply(u.read())),s==null&&(s=!0),s?this._trainableWeights.push(u):this._nonTrainableWeights.push(u),u}setFastWeightInitDuringBuild(e){this.fastWeightInitDuringBuild=e}addLoss(e){e==null||Array.isArray(e)&&e.length===0||(e=kt(e),this._losses!==void 0&&this._losses!==null&&this.losses.push(...e))}computeOutputShape(e){return e}computeMask(e,t){if(!this.supportsMasking){if(t!=null)if(Array.isArray(t))t.forEach(n=>{if(n!=null)throw new TypeError(`Layer ${this.name} does not support masking, but was passed an inputMask.`)});else throw new TypeError(`Layer ${this.name} does not support masking, but was passed an inputMask.`);return null}return t}addInboundNode(e,t,n,a,r,s,i=null){let o=kt(e);t=kt(t),n=kt(n),a=kt(a),r=em(r),s=em(s);let l=[],u=[],p=[];for(let d of o)l.push(d.sourceLayer),u.push(d.nodeIndex),p.push(d.tensorIndex);new qf({outboundLayer:this,inboundLayers:l,nodeIndices:u,tensorIndices:p,inputTensors:o,outputTensors:t,inputMasks:n,outputMasks:a,inputShapes:r,outputShapes:s},i);for(let d=0;d<t.length;d++)t[d].sourceLayer=this,t[d].nodeIndex=this.inboundNodes.length-1,t[d].tensorIndex=d}getConfig(){let e={name:this.name,trainable:this.trainable};return this.batchInputShape!=null&&(e.batchInputShape=this.batchInputShape),this.dtype!=null&&(e.dtype=this.dtype),e}disposeWeights(){return this.weights.forEach(e=>e.dispose()),this.weights.length}assertNotDisposed(){if(this._refCount===0)throw new Error(`Layer '${this.name}' is already disposed.`)}dispose(){if(!this.built)throw new Error(`Cannot dispose Layer ${this.name} because it has not been built yet.`);if(this._refCount===null)throw new Error(`Cannot dispose Layer ${this.name} because it has not been used yet.`);this.assertNotDisposed();let e=0;return--this._refCount===0&&(e=this.disposeWeights()),{refCountAfterDispose:this._refCount,numDisposedVariables:e}}};function dV(e){e=kt(e);let t=[];for(let n of e)t.push(n.shape);return Pn(t)}function hV(e){return"float32"}function L2(e,t,n){if((t==null||n!=null&&n>0)&&(t=e.sourceLayer,n=e.nodeIndex),t.inboundNodes.length===0)return[e];{let a=t.inboundNodes[n];if(a.inboundLayers.length===0)return a.inputTensors;{let r=[];for(let s=0;s<a.inboundLayers.length;s++){let i=a.inputTensors[s],o=a.inboundLayers[s],l=a.nodeIndices[s],u=L2(i,o,l);for(let p of u)r.indexOf(p)===-1&&r.push(p)}return r}}}var Ku=class extends Ye{constructor(e){if(super({dtype:e.dtype,name:e.name!=null?e.name:Of("input").toString()}),e.batchSize==null&&(e.batchSize=null),e.sparse==null&&(e.sparse=!1),this.trainable=!1,this.built=!0,this.sparse=e.sparse,e.inputShape!=null&&e.batchInputShape!=null)throw new H("Only provide the inputShape OR batchInputShape argument to inputLayer, not both at the same time.");let t=e.batchInputShape;if(t==null){if(e.inputShape==null)throw new H("An InputLayer should be passed either a `batchInputShape` or an `inputShape`.");t=[e.batchSize].concat(e.inputShape)}else if(e.batchSize!=null)throw new H("Cannot specify batchSize if batchInputShape is specified when creating an InputLayer.");let n=e.dtype||"float32";this.batchInputShape=t,this.dtype=n,this.inputSpec=[{shape:t}];let a=new Ua(this.dtype,this.batchInputShape,this,[],{},this.name);a.nodeIndex=0,a.tensorIndex=0,new qf({outboundLayer:this,inboundLayers:[],nodeIndices:[],tensorIndices:[],inputTensors:[a],outputTensors:[a],inputMasks:[null],outputMasks:[null],inputShapes:[t],outputShapes:[t]})}apply(e,t){throw new H(`Cannot pass any input to an InputLayer's apply() method. InputLayer name: ${this.name}`)}dispose(){return{refCountAfterDispose:this._refCount,numDisposedVariables:0}}getConfig(){return{batchInputShape:this.batchInputShape,dtype:this.dtype,sparse:this.sparse,name:this.name}}};Ku.className="InputLayer";se.registerClass(Ku);function z2(e){if(e.batchShape==null&&e.shape==null)throw new Error("Please provide to Input either a `shape` or a `batchShape` argument. Note that `shape` does not include the batch dimension.");if(e.batchShape!=null&&e.shape!=null)throw new H("Please provide either a `shape` or `batchShape` argument to Input, but not both.");let t=e.batchShape;e.shape!=null&&t==null&&(t=[null].concat(e.shape));let n=e.dtype;return n==null&&(n="float32"),new Ku({batchInputShape:t,name:e.name,dtype:n,sparse:e.sparse}).inboundNodes[0].outputTensors[0]}function mV(e,t){if(e.dtype==null||e.dtype===t.dtype)return t;try{return oe(t,e.dtype)}catch(n){throw new H(`The dtype of the feed (${t.dtype}) can not be cast to the dtype of the key '${e.name}' (${e.dtype}).`)}}var Qs=class{constructor(e){if(this.id2Value={},this.id2Mask={},this.name2Id={},e instanceof Qs)for(let t in e.id2Value)this.id2Value[t]=e.id2Value[t],t in e.id2Mask&&(this.id2Mask[t]=e.id2Mask[t]);else{if(e==null)return;for(let t of e)this.add(t.key,t.value)}}add(e,t,n){if(this.id2Value[e.id]==null)this.id2Value[e.id]=mV(e,t),this.name2Id[e.name]=e.id,n!=null&&(this.id2Mask[e.id]=n);else throw new H(`Duplicate key: name=${e.name}, id=${e.id}`);return this}addFeed(e){this.add(e.key,e.value)}hasKey(e){return this.id2Value[e.id]!=null}names(){return Object.keys(this.name2Id)}getValue(e){if(e instanceof Ua){if(this.id2Value[e.id]==null)throw new H(`Nonexistent key: ${e.name}`);return this.id2Value[e.id]}else{let t=this.name2Id[e];if(t==null)throw new H(`Feed dict has no SymbolicTensor name: ${e}`);return this.id2Value[t]}}getMask(e){if(e instanceof Ua){if(this.id2Value[e.id]==null)throw new H(`Nonexistent key: ${e.name}`);return this.id2Mask[e.id]}else{let t=this.name2Id[e];if(t==null)throw new H(`Feed dict has no SymbolicTensor name: ${e}`);return this.id2Mask[t]}}disposeMasks(){this.id2Mask!=null&&Re(this.id2Mask)}},nm=new _2,am=new _2;function fV(e){nm!=null&&nm.setMaxEntries(e),am!=null&&am.setMaxEntries(e)}function Vp(e,t,n,a){let r=n==null?!1:n.training,s=Array.isArray(e),i=s?e:[e],o=i.map(m=>m.name),l=[],u=t.names();for(let m of o)u.indexOf(m)!==-1?l.push(t.getValue(m)):l.push(null);a!=null&&(a.maxNumTensors=-1/0,a.minNumTensors=1/0);let p=o.join(",")+"|"+t.names().sort().join(","),d=nm.get(p),c;if(d==null){let m=gV(i,t);d=m.sorted,c=m.recipientCounts,nm.put(p,d),am.put(p,c)}c={},r||Object.assign(c,am.get(p));let h=new Qs(t);for(let m=0;m<d.length;++m){if(a!=null){let A=Qh().numTensors;A>a.maxNumTensors&&(a.maxNumTensors=A),A<a.minNumTensors&&(a.minNumTensors=A)}let f=d[m],g=f.sourceLayer;if(g instanceof Ku)continue;let y=[],b=[],x=[],v=!1;for(let A of f.inputs){let M=h.getValue(A),$=h.getMask(A);y.push(M),b.push($),$!=null&&(v=!0),r||(c[A.name]--,c[A.name]===0&&!t.hasKey(A)&&o.indexOf(A.name)===-1&&!M.isDisposed&&A.sourceLayer.stateful!==!0&&x.push(M))}v&&(n=n||{},n.mask=b[0]);let k=kt(g.apply(y,n)),T=null;g.supportsMasking&&(T=g.computeMask(y,b));let _=bV(f),E=Array.isArray(_)?_:[_];for(let A=0;A<E.length;++A){h.hasKey(E[A])||h.add(E[A],k[A],Array.isArray(T)?T[0]:T);let M=o.indexOf(E[A].name);M!==-1&&(l[M]=k[A])}r||Re(x)}return h.disposeMasks(),s?l:l[0]}function gV(e,t){w.assert(e!=null&&e.length>0,()=>"Expected at least one fetch, got none");let n=[],a={};if(e.length===1){let r=xk(e[0],t);n=r.sorted,a=r.recipientMap}else{let r=new Set;for(let s of e){let{sorted:i,recipientMap:o}=xk(s,t);for(let l of i)r.has(l.name)||(n.push(l),r.add(l.name));for(let l in o)a[l]==null&&(a[l]=new Set),o[l].forEach(u=>a[l].add(u))}}return{sorted:n,recipientCounts:yV(a)}}function yV(e){let t={};for(let n in e)t[n]=e[n].size;return t}function xk(e,t){let n=new Set,a=[],r={};for(let o of t.names())n.add(o);let s=[],i=[];for(s.push(e);s.length>0;){let o=s[s.length-1];if(n.has(o.name)){s.pop();continue}let l=i[i.length-1]===s.length-1;if(o.inputs.length===0||l)s.pop(),a.push(o),n.add(o.name),l&&i.pop();else{i.push(s.length-1);for(let u of o.inputs)r[u.name]==null&&(r[u.name]=new Set),r[u.name].add(o.name),!n.has(u.name)&&s.push(u)}}return{sorted:a,recipientMap:r}}function bV(e){let t;if(e.sourceLayer.inboundNodes.length===1)t=e.sourceLayer.output;else{let n=null;for(let a=0;a<e.sourceLayer.inboundNodes.length;++a)for(let r of e.sourceLayer.inboundNodes[a].outputTensors)if(r.id===e.id){n=a;break}t=e.sourceLayer.getOutputAt(n)}return t}var xV=X();xV.registerFlag("TOPOLOGICAL_SORT_CACHE_MAX_ENTRIES",()=>100,fV);var B2={};Me(B2,{maxNorm:()=>vV,minMaxNorm:()=>IV,nonNeg:()=>kV,unitNorm:()=>wV});function Zv(e,t){return O(()=>pn(be(B(e,e),t,!0)))}var nd=class extends se.Serializable{getConfig(){return{}}},ew=class extends nd{constructor(e){super(),this.defaultMaxValue=2,this.defaultAxis=0,this.maxValue=e.maxValue!=null?e.maxValue:this.defaultMaxValue,this.axis=e.axis!=null?e.axis:this.defaultAxis}apply(e){return O(()=>{let t=Zv(e,this.axis),n=an(t,0,this.maxValue);return B(e,fe(n,J(jt(),t)))})}getConfig(){return{maxValue:this.maxValue,axis:this.axis}}};ew.className="MaxNorm";se.registerClass(ew);var tw=class extends nd{constructor(e){super(),this.defaultAxis=0,this.axis=e.axis!=null?e.axis:this.defaultAxis}apply(e){return O(()=>fe(e,J(jt(),Zv(e,this.axis))))}getConfig(){return{axis:this.axis}}};tw.className="UnitNorm";se.registerClass(tw);var nw=class extends nd{apply(e){return Xe(e)}};nw.className="NonNeg";se.registerClass(nw);var aw=class extends nd{constructor(e){super(),this.defaultMinValue=0,this.defaultMaxValue=1,this.defaultRate=1,this.defaultAxis=0,this.minValue=e.minValue!=null?e.minValue:this.defaultMinValue,this.maxValue=e.maxValue!=null?e.maxValue:this.defaultMaxValue,this.rate=e.rate!=null?e.rate:this.defaultRate,this.axis=e.axis!=null?e.axis:this.defaultAxis}apply(e){return O(()=>{let t=Zv(e,this.axis),n=J(B(this.rate,an(t,this.minValue,this.maxValue)),B(1-this.rate,t));return B(e,fe(n,J(jt(),t)))})}getConfig(){return{minValue:this.minValue,maxValue:this.maxValue,rate:this.rate,axis:this.axis}}};aw.className="MinMaxNorm";se.registerClass(aw);var vk={maxNorm:"MaxNorm",minMaxNorm:"MinMaxNorm",nonNeg:"NonNeg",unitNorm:"UnitNorm"};function Xt(e){return Vv(e)}function wk(e,t={}){return Jc(e,se.SerializationMap.getMap().classNameMap,t,"constraint")}function Yt(e){if(e==null)return null;if(typeof e=="string"){let t={className:e in vk?vk[e]:e,config:{}};return wk(t)}else return e instanceof nd?e:wk(e)}function vV(e){return new ew(e)}function wV(e){return new tw(e)}function kV(){return new nw}function IV(e){return new aw(e)}var W2={};Me(W2,{constant:()=>TV,glorotNormal:()=>DV,glorotUniform:()=>$V,heNormal:()=>RV,heUniform:()=>MV,identity:()=>FV,leCunNormal:()=>PV,leCunUniform:()=>OV,ones:()=>NV,orthogonal:()=>LV,randomNormal:()=>CV,randomUniform:()=>_V,truncatedNormal:()=>EV,varianceScaling:()=>AV,zeros:()=>SV});function SV(){return new Hv}function NV(){return new Bf}function TV(e){return new jv(e)}function _V(e){return new qv(e)}function CV(e){return new Kv(e)}function EV(e){return new Xv(e)}function FV(e){return new Yv(e)}function AV(e){return new Wn(e)}function $V(e){return new Wf(e)}function DV(e){return new Vf(e)}function RV(e){return new Uf(e)}function MV(e){return new Gf(e)}function PV(e){return new Hf(e)}function OV(e){return new jf(e)}function LV(e){return new Qv(e)}var V2={};Me(V2,{Layer:()=>Ye,RNN:()=>yr,RNNCell:()=>od,activation:()=>pG,add:()=>xG,alphaDropout:()=>n6,average:()=>vG,averagePooling1d:()=>p0,averagePooling2d:()=>c0,averagePooling3d:()=>d0,avgPool1d:()=>EG,avgPool2d:()=>AG,avgPool3d:()=>DG,avgPooling1d:()=>FG,avgPooling2d:()=>$G,avgPooling3d:()=>RG,batchNormalization:()=>TG,bidirectional:()=>KG,concatenate:()=>wG,conv1d:()=>tG,conv2d:()=>nG,conv2dTranspose:()=>aG,conv3d:()=>rG,conv3dTranspose:()=>sG,convLstm2d:()=>GG,convLstm2dCell:()=>HG,cropping2D:()=>oG,dense:()=>cG,depthwiseConv2d:()=>uG,dot:()=>NG,dropout:()=>dG,elu:()=>XU,embedding:()=>bG,flatten:()=>mG,gaussianDropout:()=>t6,gaussianNoise:()=>e6,globalAveragePooling1d:()=>MG,globalAveragePooling2d:()=>PG,globalMaxPool1d:()=>YG,globalMaxPool2d:()=>QG,globalMaxPooling1d:()=>DN,globalMaxPooling2d:()=>RN,gru:()=>LG,gruCell:()=>zG,input:()=>iN,inputLayer:()=>KU,layerNormalization:()=>_G,leakyReLU:()=>QU,lstm:()=>BG,lstmCell:()=>WG,masking:()=>a6,maxPool1d:()=>JG,maxPool2d:()=>ZG,maxPooling1d:()=>MN,maxPooling2d:()=>PN,maxPooling3d:()=>OG,maximum:()=>kG,minimum:()=>IG,multiply:()=>SG,permute:()=>yG,prelu:()=>JU,reLU:()=>YU,repeatVector:()=>fG,reshape:()=>gG,rnn:()=>jG,separableConv2d:()=>iG,simpleRNN:()=>VG,simpleRNNCell:()=>UG,softmax:()=>ZU,spatialDropout1d:()=>hG,stackedRNNCells:()=>qG,thresholdedReLU:()=>eG,timeDistributed:()=>XG,upSampling2d:()=>lG,zeroPadding2d:()=>CG});async function Zr(e){if(e==null)return;let t=[],n=[],a=[];for(let r in e){let s=e[r];if(typeof s!="number"){let i=s;t.push(i.data()),n.push(r),a.push(i)}}if(t.length>0){let r=await Promise.all(t);for(let s=0;s<r.length;++s)e[n[s]]=r[s][0];Re(a)}}function U2(e){if(e!=null)for(let t in e){let n=e[t];typeof n!="number"&&n.dispose()}}var kk;(function(e){e[e.SILENT=0]="SILENT",e[e.VERBOSE=1]="VERBOSE"})(kk||(kk={}));var zV=125,El=class{constructor(){this.validationData=null}setParams(e){this.params=e}async onEpochBegin(e,t){}async onEpochEnd(e,t){}async onBatchBegin(e,t){}async onBatchEnd(e,t){}async onTrainBegin(e){}async onTrainEnd(e){}setModel(e){}},G2=class{constructor(e,t=10){e==null&&(e=[]),this.callbacks=e,this.queueLength=t}append(e){this.callbacks.push(e)}setParams(e){for(let t of this.callbacks)t.setParams(e)}setModel(e){for(let t of this.callbacks)t.setModel(e)}async onEpochBegin(e,t){t==null&&(t={});for(let n of this.callbacks)await n.onEpochBegin(e,t)}async onEpochEnd(e,t){t==null&&(t={});for(let n of this.callbacks)await n.onEpochEnd(e,t)}async onBatchBegin(e,t){t==null&&(t={});for(let n of this.callbacks)await n.onBatchBegin(e,t)}async onBatchEnd(e,t){t==null&&(t={});for(let n of this.callbacks)await n.onBatchEnd(e,t)}async onTrainBegin(e){e==null&&(e={});for(let t of this.callbacks)await t.onTrainBegin(e)}async onTrainEnd(e){e==null&&(e={});for(let t of this.callbacks)await t.onTrainEnd(e)}},BV=class extends El{constructor(){super()}async onEpochBegin(e){this.seen=0,this.totals={}}async onBatchEnd(e,t){t==null&&(t={});let n=t.size==null?0:t.size;this.seen+=n;for(let a in t){let r=t[a];if(typeof r=="number")this.totals.hasOwnProperty(a)||(this.totals[a]=0),this.totals[a]=this.totals[a]+r*n;else{let s;a in this.totals?s=this.totals[a]:this.totals[a]=0;let i=O(()=>J(this.totals[a],B(r,n)));this.totals[a]=i,s!=null&&s.dispose()}}}async onEpochEnd(e,t){if(t!=null)for(let n of this.params.metrics)this.totals[n]!=null&&(typeof this.totals[n]=="number"?t[n]=this.totals[n]/this.seen:O(()=>{let a=B(fe(1,this.seen),this.totals[n]);t[n]=a,this.totals[n].dispose(),tn(t[n])}))}},H2=class extends El{async onTrainBegin(e){this.epoch=[],this.history={}}async onEpochEnd(e,t){t==null&&(t={}),this.epoch.push(e);for(let n in t)this.history[n]==null&&(this.history[n]=[]),this.history[n].push(t[n])}async syncData(){let e=[],t=[],n=[];for(let r in this.history){let s=this.history[r];for(let i=0;i<s.length;++i)if(typeof s[i]!="number"){let o=s[i];e.push(o.data()),t.push(r),n.push(i)}}let a=await Promise.all(e);for(let r=0;r<a.length;++r)this.history[t[r]][n[r]].dispose(),this.history[t[r]][n[r]]=a[r][0]}},j2=class extends El{constructor(e,t){if(super(),this.currentEpoch=0,this.nowFunc=e.nowFunc,this.nextFrameFunc=e.nextFrameFunc||Bv,this.yieldEvery=t||"auto",this.yieldEvery==="auto"&&(this.yieldEvery=zV),this.yieldEvery==="never"&&e.onYield!=null)throw new Error("yieldEvery is `never` but you provided an `onYield` callback. Either change `yieldEvery` or remove the callback");w.isNumber(this.yieldEvery)&&(this.maybeWait=W4(this.maybeWait.bind(this),this.yieldEvery,this.nowFunc)),this.trainBegin=e.onTrainBegin,this.trainEnd=e.onTrainEnd,this.epochBegin=e.onEpochBegin,this.epochEnd=e.onEpochEnd,this.batchBegin=e.onBatchBegin,this.batchEnd=e.onBatchEnd,this.yield=e.onYield}async maybeWait(e,t,n){let a=[];this.yield!=null&&(await Zr(n),a.push(this.yield(e,t,n))),a.push(this.nextFrameFunc()),await Promise.all(a)}async onEpochBegin(e,t){this.currentEpoch=e,this.epochBegin!=null&&(await Zr(t),await this.epochBegin(e,t))}async onEpochEnd(e,t){let n=[];this.epochEnd!=null&&(await Zr(t),n.push(this.epochEnd(e,t))),this.yieldEvery==="epoch"&&n.push(this.nextFrameFunc()),await Promise.all(n)}async onBatchBegin(e,t){this.batchBegin!=null&&(await Zr(t),await this.batchBegin(e,t))}async onBatchEnd(e,t){let n=[];this.batchEnd!=null&&(await Zr(t),n.push(this.batchEnd(e,t))),this.yieldEvery==="batch"?n.push(this.nextFrameFunc()):w.isNumber(this.yieldEvery)&&n.push(this.maybeWait(this.currentEpoch,e,t)),await Promise.all(n)}async onTrainBegin(e){this.trainBegin!=null&&(await Zr(e),await this.trainBegin(e))}async onTrainEnd(e){this.trainEnd!=null&&(await Zr(e),await this.trainEnd(e))}};function q2(e,t){return e==null&&(e={}),e instanceof El?[e]:Array.isArray(e)&&e[0]instanceof El?e:kt(e).map(n=>new j2(n,t))}var Sa=class{constructor(){}static registerCallbackConstructor(e,t){w.assert(e>=0&&Number.isInteger(e),()=>`Verbosity level is expected to be an integer >= 0, but got ${e}`),Sa.checkForDuplicate(t),Sa.constructors[e]==null&&(Sa.constructors[e]=[]),Sa.constructors[e].push(t)}static checkForDuplicate(e){for(let t in Sa.constructors)Sa.constructors[+t].forEach(n=>{if(n===e)throw new H("Duplicate callback constructor.")})}static clear(){Sa.constructors={}}static createCallbacks(e){let t=[];for(let n in Sa.constructors){let a=+n;e>=a&&t.push(...Sa.constructors[a])}return t.map(n=>new n)}};Sa.constructors={};function K2(e,t,n,a,r,s,i,o,l){let u=new H2,p=[new BV,...Sa.createCallbacks(t)];e!=null&&p.push(...e),p.push(u);let d=new G2(p);return d.setParams({epochs:n,initialEpoch:a,samples:r,steps:s,batchSize:i,verbose:t,doValidation:o,metrics:l}),{callbackList:d,history:u}}function ja(e,t={},n=!1){return Jc(e,se.SerializationMap.getMap().classNameMap,t,"layer",n)}function rm(e,t){return O(()=>{e.dtype!=="float32"&&(e=oe(e,"float32"));let n=be(ed(e),t,!0),a=Cn(n.shape,jt()),r=pn(fr(n,a));return fe(e,r)})}function ko(e,t){return O(()=>Et(ed(ce(t,e)),-1))}function Kf(e,t){return O(()=>Et(zt(ce(t,e)),-1))}function Xu(e,t){return O(()=>{let n=ce(e,t),a=an(zt(e),jt(),Number.MAX_VALUE),r=zt(fe(n,a));return B(100,Et(r,-1))})}function WV(e,t){return O(()=>{let n=an(t,jt(),Number.MAX_VALUE),a=ea(J(1,n)),r=an(e,jt(),Number.MAX_VALUE),s=ea(J(1,r));return Et(ed(ce(a,s)),-1)})}function VV(e,t){return O(()=>{let n=fr(0,ce(1,B(e,t)));return Et(ed(n),-1)})}function UV(e,t){return O(()=>{let n=fr(0,ce(1,B(e,t)));return Et(n,-1)})}function GV(e,t){return O(()=>{let n=be(B(e,t),-1),a=Ta(B(ce(1,e),t),-1);return fr(0,J(1,ce(a,n)))})}function HV(e,t){return O(()=>{let n=Math.log(2),a=ce(t,e),r=ce(J(a,xo(B(-2,a))),n);return Et(r,-1)})}function lc(e,t,n=!1){return O(()=>{if(n)t=Qa(t);else{let a=be(t,t.shape.length-1,!0);t=fe(t,a)}return t=an(t,jt(),1-jt()),It(be(B(oe(e,"float32"),ea(t)),t.shape.length-1))})}function sm(e,t,n=!1){return O(()=>{let a=oe(Vu(Z4(e)),"int32");t=an(t,jt(),1-jt());let r=t.shape,s=W(Sl(a,r[r.length-1]),r);return lc(s,t,n)})}function jV(e,t){if(!w.arraysEqual(e.shape,t.shape))throw new H(`logits and labels must have the same shape, but got shapes ${JSON.stringify(e.shape)} and ${JSON.stringify(t.shape)}`);return O(()=>{let n=Xe(t),a=It(zt(t));return J(ce(n,B(t,e)),Hc(gn(a)))})}function Xf(e,t){return O(()=>{let n;return n=an(t,jt(),1-jt()),n=ea(fe(n,ce(1,n))),Et(jV(e,n),-1)})}function qV(e,t){return O(()=>{let n=an(e,jt(),1),a=an(t,jt(),1);return be(B(e,ea(fe(n,a))),-1)})}function KV(e,t){return O(()=>{let n=ea(J(jt(),t));return Et(ce(t,B(e,n)),-1)})}function rw(e,t){return O(()=>{let n=rm(e,-1),a=rm(t,-1),r=B(n,a);return It(be(r,-1))})}var im={meanSquaredError:ko,meanAbsoluteError:Kf,meanAbsolutePercentageError:Xu,meanSquaredLogarithmicError:WV,squaredHinge:VV,hinge:UV,categoricalHinge:GV,logcosh:HV,categoricalCrossentropy:lc,sparseCategoricalCrossentropy:sm,binaryCrossentropy:Xf,kullbackLeiblerDivergence:qV,poisson:KV,cosineProximity:rw};function Sb(e){if(typeof e=="string"){if(e in im)return im[e];let t=`Unknown loss ${e}`;throw e.toLowerCase().includes("softmaxcrossentropy")&&(t=`Unknown loss ${e}. Use "categoricalCrossentropy" as the string name for tf.losses.softmaxCrossEntropy`),new H(t)}else return e}function sw(e,t){return O(()=>{let n=B(.5,ta(t)),a=Lf(Un(t,n),e.dtype);return Et(Zn(e,a),-1)})}function iw(e,t){return O(()=>Lf(Zn(li(e,-1),li(t,-1)),"float32"))}function X2(e,t){return O(()=>oe(be(Ca(Zn(e,1),Zn(t,1))),"float32"))}function XV(e,t){return O(()=>oe(be(Ca(Zn(e,1),Zn(t,0))),"float32"))}function YV(e,t){return O(()=>oe(be(Ca(Zn(e,0),Zn(t,1))),"float32"))}function Y2(e,t){return O(()=>{let n=X2(e,t),a=YV(e,t),r=J(n,a);return oe(fn(Un(r,0),fe(n,r),0),"float32")})}function QV(e,t){return O(()=>{let n=X2(e,t),a=XV(e,t),r=J(n,a);return oe(fn(Un(r,0),fe(n,r),0),"float32")})}function Q2(e,t){return Xf(e,t)}function J2(e,t){return e.rank===t.rank&&(e=dr(e,[e.rank-1])),t=li(t,-1),t.dtype!==e.dtype&&(t=oe(t,e.dtype)),oe(Zn(e,t),"float32")}var JV=ko,ZV=ko,eU=Kf,tU=Kf,nU=Xu,aU=Xu,ow=lc,rU=rw,Z2=sm,om={binaryAccuracy:sw,categoricalAccuracy:iw,precision:Y2,categoricalCrossentropy:ow,sparseCategoricalCrossentropy:Z2,mse:JV,MSE:ZV,mae:eU,MAE:tU,mape:nU,MAPE:aU,cosine:rU};function sU(e){if(typeof e=="string"&&e in om)return om[e];if(typeof e!="string"&&e!=null)return e;throw new H(`Unknown metric ${e}`)}function Ih(e){if(ir(e!==null,`Unknown LossOrMetricFn ${e}`),typeof e=="string")return e;{let t;for(let n of Object.keys(im))if(im[n]===e){t=n;break}if(t!==void 0)return t;for(let n of Object.keys(om))if(om[n]===e){t=n;break}return t!==void 0?t:e.name}}function iU(e){let t={Adagrad:()=>Hs.adagrad(.01),Adadelta:()=>Hs.adadelta(1,.95,jt()),Adam:()=>Hs.adam(.001,.9,.999,jt()),Adamax:()=>Hs.adamax(.002,.9,.999,jt(),0),RMSProp:()=>Hs.rmsprop(.001,.9,0,jt()),SGD:()=>Hs.sgd(.01)};if(t.adagrad=t.Adagrad,t.adadelta=t.Adadelta,t.adam=t.Adam,t.adamax=t.Adamax,t.rmsprop=t.RMSProp,t.sgd=t.SGD,e in t)return t[e]();throw new H(`Unknown Optimizer ${e}`)}function Ik(e,t,n=!1){if(e==null||typeof e!="object"||Object.getPrototypeOf(e)!==Object.prototype||!Jb(e))throw new Error("User-defined metadata is expected to be a JSON object, but is not.");if(n){let a=JSON.stringify(e);a.length>1048576&&console.warn(`User-defined metadata of model "${t}" is too large in size (length=${a.length} when serialized). It is not recommended to store such large objects in user-defined metadata. Please make sure its serialized length is <= ${1048576}.`)}}function Jb(e){if(e===null)return!0;if(typeof e=="object")if(Object.getPrototypeOf(e)===Object.prototype){let t=Object.keys(e);for(let n of t)if(typeof n!="string"||!Jb(e[n]))return!1;return!0}else if(Array.isArray(e)){for(let t of e)if(!Jb(t))return!1;return!0}else return!1;else{let t=typeof e;return t==="string"||t==="number"||t==="boolean"}}function oU(e,t,n,a=console.log){let r=uU(e),s=["Layer (type)","Input Shape","Output shape","Param #"];r?(t=t||90,n=n||[.32,.61,.89,1]):(t=t||115,n=n||[.24,.48,.7,.8,1]),n[n.length-1]<=1&&(n=n.map(p=>Math.floor(t*p)));let i;if(!r){s.push("Receives inputs"),i=[];for(let p in e.nodesByDepth)i.push(...e.nodesByDepth[p])}a("_".repeat(t)),lm(s,n,a),a("=".repeat(t));let o=e.layers;for(let p=0;p<o.length;++p)r?pU(o[p],n,a):cU(o[p],n,i,a),a((p===o.length-1?"=":"_").repeat(t));e.checkTrainableWeightsConsistency();let l=lU(e),u=tm(e.nonTrainableWeights);a(`Total params: ${l+u}`),a(`Trainable params: ${l}`),a(`Non-trainable params: ${u}`),a("_".repeat(t))}function lU(e){let t;return e.collectedTrainableWeights!=null?t=tm(e.collectedTrainableWeights):t=tm(e.trainableWeights),t}function uU(e){let t=!0,n=[],a=[];for(let r in e.nodesByDepth)n.push(e.nodesByDepth[r]);for(let r of n){if(r.length>1||r.length===1&&r[0].inboundLayers.length>1){t=!1;break}a.push(...r)}if(t)for(let r of e.layers){let s=!1;for(let i of r.inboundNodes)if(a.indexOf(i)!==-1)if(s){t=!1;break}else s=!0;if(!t)break}return t}function lm(e,t,n=console.log){let a="";for(let r=0;r<e.length;++r)r>0&&(a=a.slice(0,a.length-1)+" "),a+=e[r],a=a.slice(0,t[r]),a+=" ".repeat(t[r]-a.length);n(a)}function pU(e,t,n){let a,r;try{r=e.inboundNodes.map(l=>JSON.stringify(l.inputShapes)).join(",")}catch(l){r="multiple"}try{a=JSON.stringify(e.outputShape)}catch(l){a="multiple"}let s=e.name,i=e.getClassName(),o=[`${s} (${i})`,r,a,e.countParams().toString()];lm(o,t,n)}function cU(e,t,n,a){let r,s;try{s=e.inboundNodes.map(d=>JSON.stringify(d.inputShapes)).join(",")}catch(d){s="multiple"}try{r=JSON.stringify(e.outputShape)}catch(d){r="multiple"}let i=[];for(let d of e.inboundNodes)if(!(n!=null&&n.length>0&&n.indexOf(d)===-1))for(let c=0;c<d.inboundLayers.length;++c){let h=d.inboundLayers[c].name,m=d.nodeIndices[c],f=d.tensorIndices[c];i.push(`${h}[${m}][${f}]`)}let o=e.name,l=e.getClassName(),u=i.length===0?"":i[0],p=[`${o} (${l})`,s,r,e.countParams().toString(),u];lm(p,t,a);for(let d=1;d<i.length;++d)lm(["","","","",i[d]],t,a)}function eN(e,t,n){return(e==="inboundNodes"||e==="outputLayers"||e==="inputLayers")&&t===0&&typeof n=="string"}function uc(e,t){if(e===null)return null;if(typeof e=="string")return Ks(e);if(typeof e=="number"||typeof e=="boolean")return e;if(e instanceof Array){let n=[],a=e.length;for(let r=0;r<a;++r){let s=e[r];eN(t,r,s)?n.push(s):n.push(uc(s,t))}return n}else{let n={};for(let a of Object.keys(e)){let r=e[a];if(a==="name"&&typeof r=="string")n[a]=r;else{let s=Ks(a);n[s]=uc(r,s)}}return n}}function Zb(e,t){if(e==null)return null;if(typeof e=="string")return Sr(e);if(typeof e=="number"||typeof e=="boolean")return e;if(e instanceof Array){let n=[],a=e.length;for(let r=0;r<a;++r){let s=e[r];eN(t,r,s)?n.push(s):n.push(Zb(s,t))}return n}else{let n={};for(let a of Object.keys(e)){let r=e[a],s=Sr(a);(a==="name"||a==="className")&&typeof r=="string"?n[s]=r:n[s]=Zb(r,a)}return n}}var lw="3.18.0",rr=class extends Ye{constructor(e){if(super({}),this.containerNodes=new Set,this.name=e.name,this.name==null){let y=this.getClassName().toLowerCase();this.name=Of(y)}if(this.supportsMasking=!1,this.trainable_=!0,Array.isArray(e.inputs)?this.inputs=e.inputs.slice():this.inputs=[e.inputs],Array.isArray(e.outputs)?this.outputs=e.outputs.slice():this.outputs=[e.outputs],ss(this.inputs).length!==this.inputs.length)throw new H(`The list of inputs passed to the model is redundant. All inputs should only appear once. Found: ${this.inputs.map(y=>y.name)}`);ss(this.outputs).length!==this.outputs.length&&console.warn(`The list of outputs passed to the model is redundant. All outputs should only appear once. Found: ${this.outputs.map(y=>y.name)}`),this.inputLayers=[],this.inputLayersNodeIndices=[],this.inputLayersTensorIndices=[],this.outputLayers=[],this.outputLayersNodeIndices=[],this.outputLayersTensorIndices=[],this.layers=[],this.internalContainerRefs=[];for(let y of this.outputs){let b=y.sourceLayer,x=y.nodeIndex,v=y.tensorIndex;this.outputLayers.push(b),this.outputLayersNodeIndices.push(x),this.outputLayersTensorIndices.push(v)}for(let y of this.inputs){let b=y.sourceLayer,x=y.nodeIndex,v=y.tensorIndex;ir(x===0,"input layer has >1 nodes"),ir(v===0,"input layer has >1 tensors"),this.inputLayers.push(b),this.inputLayersNodeIndices.push(x),this.inputLayersTensorIndices.push(v)}this.inputNames=[],this.outputNames=[],this.feedInputShapes=[],this.feedInputNames=[],this.feedOutputNames=[];for(let y=0;y<this.inputLayers.length;y++){let b=this.inputLayers[y];if(!(b instanceof Ku))throw new TypeError(`Input layers to a LayersModel must be InputLayer objects. Received inputs: ${e.inputs}. Input ${y} (0-based) originates from layer type ${b.getClassName()}.`);this.inputNames.push(b.name),this.feedInputShapes.push(b.batchInputShape),this.feedInputNames.push(b.name)}for(let y of this.outputLayers)this.outputNames.push(y.name);this.internalInputShapes=this.inputs.map(y=>y.shape),this.internalOutputShapes=this.outputs.map(y=>y.shape);let t={},n={},a={},r={},s={},i=[],o=(y,b,x,v,k,T)=>{(v==null||k==null||T==null)&&(v=y.sourceLayer,k=y.nodeIndex,T=y.tensorIndex);let _=v.inboundNodes[k];if(x.indexOf(_)!==-1)throw new Va(`The tensor ${y.name} at layer "${v.name}" is part of a cycle.`);if(b.indexOf(_)!==-1)return;this.containerNodes.add(rr.nodeKey(v,k)),v.id in s||(s[v.id]=Object.keys(s).length),x.indexOf(_)===-1&&x.push(_);let E=_.inboundLayers.length;for(let A=0;A<E;A++){let M=_.inputTensors[A],$=_.inboundLayers[A],S=_.nodeIndices[A],P=_.tensorIndices[A];o(M,b,x,$,S,P)}for(b.push(_);x.indexOf(_)>=0;)x.splice(x.indexOf(_),1);i.push(_)},l=[],u=[];for(let y of this.outputs)o(y,l,u);let p=i.slice().reverse();for(let y of p){n[y.id]=y,y.id in t||(t[y.id]=0);let b=t[y.id],x=a[y.outboundLayer.id]==null?0:a[y.outboundLayer.id];b=Math.max(b,x),a[y.outboundLayer.id]=b,r[y.outboundLayer.id]=y.outboundLayer,t[y.id]=b;for(let v=0;v<y.inboundLayers.length;v++){let k=y.inboundLayers[v],T=y.nodeIndices[v],_=k.inboundNodes[T],E=t[_.id]==null?0:t[_.id];t[_.id]=Math.max(b+1,E),n[_.id]=_}}let d={};for(let y in t){let b=t[y];b in d||(d[b]=[]),d[b].push(n[y])}let c={};for(let y in a){let b=a[y];b in c||(c[b]=[]),c[b].push(r[y])}let h=Object.keys(c).map(y=>parseInt(y,10)).sort(vh);this.layers=[];for(let y of h){let b=c[y];b.sort((x,v)=>{let k=s[x.id],T=s[v.id];return k<T?-1:k>T?1:0});for(let x of b)x instanceof rr&&this.internalContainerRefs.push(x),this.layers.push(x)}this.layersByDepth=c,h=Object.keys(d).map(y=>parseInt(y,10)).sort(vh);let m=this.inputs.slice(),f=[];for(let y of h)for(let b of d[y]){let x=b.outboundLayer;if(x!=null){for(let v of b.inputTensors)if(m.indexOf(v)===-1)throw new Va(`Graph disconnected: cannot obtain value for tensor ${v} at layer "${x.name}". The following previous layers were accessed without issue: ${f}`);for(let v of b.outputTensors)m.push(v);f.push(x.name)}}this.nodesByDepth=d;let g=this.layers.map(y=>y.name);for(let y of g){let b=g.filter(x=>x===y).length;if(b!==1)throw new Va(`The name "${y}" is used ${b} times in the model. All layer names should be unique. Layer names: `+JSON.stringify(g))}this.outboundNodes=[],this.inboundNodes=[],new qf({outboundLayer:this,inboundLayers:[],nodeIndices:[],tensorIndices:[],inputTensors:this.inputs,outputTensors:this.outputs,inputMasks:this.inputs.map(y=>null),outputMasks:this.outputs.map(y=>null),inputShapes:this.inputs.map(y=>y.shape),outputShapes:this.outputs.map(y=>y.shape)}),this.built=!0,this._refCount=1}assertNotDisposed(){if(this._refCount===0)throw new Error(`Container '${this.name}' is already disposed.`)}dispose(){this.assertNotDisposed();let e={refCountAfterDispose:null,numDisposedVariables:0};if(--this._refCount===0){for(let t of this.layers)e.numDisposedVariables+=t.dispose().numDisposedVariables;for(let t of this.internalContainerRefs)e.numDisposedVariables+=t.dispose().numDisposedVariables}return e.refCountAfterDispose=this._refCount,e}get trainable(){return this.trainable_}set trainable(e){this.layers.forEach(t=>{t._trainableWeights.forEach(n=>n.trainable=e)}),this.trainable_=e}get trainableWeights(){if(this._trainableWeights.length>0)throw new H("Container instance unexpectedly contains _trainableWeights.The trainable weights of a Container are a union of the trainable weights of its consituent Layers. Its own _trainableWeights must remain an empty Array.");if(!this.trainable)return[];let e=[];for(let t of this.layers)e=e.concat(t.trainableWeights);return e}get nonTrainableWeights(){let e=[];for(let t of this.layers)e.push(...t.nonTrainableWeights);if(!this.trainable){let t=[];for(let n of this.layers)t.push(...n.trainableWeights);return t.concat(e)}return e}get weights(){return this.trainableWeights.concat(this.nonTrainableWeights)}loadWeights(e,t=!0){let n={},a=0;for(let s of this.layers)for(let i of s.weights){if(n[i.originalName]!=null)throw new H(`Duplicate weight name: ${i.originalName}`);n[i.originalName]=i,a++}let r=[];for(let s in e){let i=s;if(n[s]==null){let o=s.split("/");i=o.slice(0,-2).concat([o[o.length-1]]).join("/")}if(n[i]!=null)r.push([n[i],e[s]]);else if(t)throw new H(`Provided weight data has no target variable: ${s}`);delete n[i]}if(t){let s=[];for(let i in n)s.push(i);if(s.length>0)throw new H(`${s.length} of ${a} weights are not set: ${s}`)}Jv(r)}updatedConfig(){let e=this.getConfig(),t={};return t.className=this.getClassName(),t.config=e,t.kerasVersion=`tfjs-layers ${lw}`,t.backend="TensorFlow.js",t}toJSON(e,t=!0){let n=Zb(this.updatedConfig());return t?JSON.stringify(n):n}call(e,t){return O(()=>{e=kt(e);let n=new Qs;for(let a=0;a<this.inputs.length;++a)n.add(this.inputs[a],e[a]);return Vp(this.outputs,n,t)})}computeMask(e,t){return O(()=>{e=kt(e);let n;return t==null?n=hi(null,e.length):n=kt(t),this.runInternalGraph(e,n)[1]})}computeOutputShape(e){let t=em(e);if(t.length!==this.inputLayers.length)throw new H(`Invalid inputShape argument ${e}: model has ${this.inputLayers.length} tensor inputs.`);let n={};for(let i=0;i<t.length;i++){let o=this.inputLayers[i],l=t[i],u=o.name+"_0_0";n[u]=l}let a=Object.keys(this.nodesByDepth).map(i=>parseInt(i,10)).sort(vh);if(a.length>1)for(let i of a){let o=this.nodesByDepth[i];for(let l of o){let u=l.outboundLayer;if(this.inputLayers.map(m=>m.id).indexOf(u.id)!==-1)continue;let p=[];for(let m=0;m<l.inboundLayers.length;m++){let f=l.inboundLayers[m],g=l.nodeIndices[m],y=l.tensorIndices[m],b=`${f.name}_${g}_${y}`,x=n[b];p.push(x)}let d=u.computeOutputShape(Pn(p)),c=em(d),h=u.inboundNodes.indexOf(l);for(let m=0;m<c.length;m++){let f=`${u.name}_${h}_${m}`;n[f]=c[m]}}}let r=[],s=[];for(let i=0;i<this.outputLayers.length;i++){let o=this.outputLayers[i],l=this.outputLayersNodeIndices[i],u=this.outputLayersTensorIndices[i],p=`${o.name}_${l}_${u}`;s.push(p)}for(let i=0;i<s.length;i++){let o=s[i];ir(o in n),r.push(n[o])}return Pn(r)}runInternalGraph(e,t){t==null&&(t=hi(null,e.length));let n={};for(let o=0;o<this.inputs.length;++o){let l=this.inputs[o],u=e[o],p=t[o];n[l.id]=[u,p]}let a=Object.keys(this.nodesByDepth).map(o=>parseInt(o,10)).sort(vh);for(let o of a){let l=this.nodesByDepth[o];for(let u of l){let p=u.outboundLayer,d=u.inputTensors,c=u.outputTensors,h=new Array;for(let m of d)m.id in n&&h.push(n[m.id]);if(h.length===d.length){let m={},f,g,y,b;if(u.callArgs!=null&&(m=u.callArgs),h.length===1){let[x,v]=h[0];m.mask==null&&(m.mask=v),y=kt(p.call(x,m)),b=kt(p.computeMask(x,v)),f=[x],g=[v]}else f=h.map(x=>x[0]),g=h.map(x=>x[1]),m.mask==null&&(m.mask=g),y=kt(p.call(f,m)),b=kt(p.computeMask(f,g));if(p.activityRegularizer)throw new Oe("LayersModel invocation with concrete Tensor value(s) in the presence of activity regularizer(s) is not supported yet.");for(let x=0;x<c.length;++x){let v=c[x],k=y[x],T=b[x];n[v.id]=[k,T]}}}}let r=[],s=[],i=[];for(let o of this.outputs){ir(o.id in n,`Could not compute output ${o.name} : ${o.id}`);let[l,u]=n[o.id];i.push(l.shape),r.push(l),s.push(u)}return[r,s,i]}buildNodeConversionMap(e){let t={},n;for(let a of this.layers){n=a instanceof rr?1:0;for(let r=0;r<a.inboundNodes.length;r++){let s=rr.nodeKey(a,r);this.containerNodes.has(s)&&(t[s]=n,n+=1)}}return t}getLayer(e,t){if(t!=null){if(this.layers.length<=t)throw new H(`Was asked to retrieve layer at index ${t}, but model only has ${this.layers.length} layer(s).`);return this.layers[t]}else if(e==null)throw new H("Provide either a layer name or layer index");for(let n of this.layers)if(n.name===e)return n;throw new H(`No such layer: ${e}`)}calculateLosses(){return O(()=>{let e=[];for(let t of this.layers)for(let n=0;n<t.inboundNodes.length;++n){let a=rr.nodeKey(t,n);this.containerNodes.has(a)&&e.push(...t.calculateLosses())}return e})}getConfig(){let e={name:this.name},t=this.buildNodeConversionMap(this.layers),n=[];for(let s of this.layers){let i=s.getClassName(),o=s.getConfig(),l=[];for(let p=0;p<s.inboundNodes.length;p++){let d=s.inboundNodes[p],c=rr.nodeKey(s,p),h={};if(this.containerNodes.has(c)){if(d.callArgs)try{JSON.stringify(d.callArgs),h=d.callArgs}catch(m){console.warn(`Layer ${s.name} was passed non-serializable keyword arguments: ${d.callArgs}. They will not be included in the serialized model (and thus will be missing at deserialization time).`),h={}}if(d.inboundLayers.length>0){let m=[];for(let f=0;f<d.inboundLayers.length;f++){let g=d.inboundLayers[f],y=d.nodeIndices[f],b=d.tensorIndices[f],x=rr.nodeKey(g,y),v=t[x];v==null&&(v=0),m.push([g.name,v,b,h])}l.push(m)}}}let u={};u.name=s.name,u.className=i,u.config=o,u.inboundNodes=l,n.push(u)}e.layers=n;let a=[];for(let s=0;s<this.inputLayers.length;s++){let i=this.inputLayers[s],o=this.inputLayersNodeIndices[s],l=rr.nodeKey(i,o);if(!this.containerNodes.has(l))continue;let u=t[l];u==null&&(u=0);let p=this.inputLayersTensorIndices[s];a.push([i.name,u,p])}e.inputLayers=a;let r=[];for(let s=0;s<this.outputLayers.length;s++){let i=this.outputLayers[s],o=this.outputLayersNodeIndices[s],l=rr.nodeKey(i,o);if(!this.containerNodes.has(l))continue;let u=t[l];u==null&&(u=0);let p=this.outputLayersTensorIndices[s];r.push([i.name,u,p])}return e.outputLayers=r,e}static fromConfig(e,t,n={},a=!1){let r={},s={};function i(f,g){f.name in s?s[f.name].push(g):s[f.name]=[g]}function o(f,g){let y=[],b;for(let x of g){let v=x[0],k=x[1],T=x[2];if(b=x[3]==null?{}:x[3],!(v in r)){i(f,g);return}let _=r[v];if(_.inboundNodes.length<=k){i(f,g);return}let E=_.inboundNodes[k];y.push(E.outputTensors[T])}y.length>0&&f.apply(Pn(y),b)}function l(f){let g=f.name,y=ja(f,t.customObjects!=null?t.customObjects:{});y.setFastWeightInitDuringBuild(a),r[g]=y,f.inboundNodes.forEach(b=>{if(!(b instanceof Array))throw new H(`Corrupted configuration, expected array for nodeData: ${b}`);i(y,b)})}let u=t.name,p=t.layers;for(let f of p)l(f);for(;!B4(s);)for(let f of p){let g=r[f.name];if(g.name in s){let y=s[g.name];delete s[g.name];for(let b of y)o(g,b)}}let d=[],c=[],h=t.inputLayers;for(let f of h){let g=f[0],y=f[1],b=f[2];ir(g in r);let x=r[g].inboundNodes[y].outputTensors;d.push(x[b])}let m=t.outputLayers;for(let f of m){let g=f[0],y=f[1],b=f[2];ir(g in r);let x=r[g].inboundNodes[y].outputTensors;c.push(x[b])}return new e({inputs:d,outputs:c,name:u})}get stateful(){if(this._stateful)throw new H("Container instance unexpectedly has _stateful = true. The statefulness of a Container is determined by the Layers it contains. Its _stateful property must remain the default false.");for(let e of this.layers)if(e.stateful)return!0;return!1}resetStates(){O(()=>{this.layers.forEach(e=>{e.stateful&&e.resetStates()})})}};function dU(e,t,n){let a=t.length;if(e==null||Array.isArray(e)&&e.length===0)return t.map(r=>null);if(a===1)return Array.isArray(e)&&e.length===1?e:typeof e=="object"&&t[0]in e?[e[t[0]]]:[e];if(Array.isArray(e)){if(e.length!==a)throw new Error(`Provided ${n} is an array of ${e.length} element(s), but the model has ${a} outputs. Make sure a set of weights is provided for each model output.`);return e}else if(typeof e=="object"&&Object.keys(e).length>0&&typeof e[Object.keys(e)[0]]=="object"){let r=[];return t.forEach(s=>{s in e?r.push(e[s]):r.push(null)}),r}else throw new Error(`The model has multiple (${a}) outputs, so ${n} must be either an array with ${a} elements or an object with ${t} keys. Provided ${n} not understood: ${JSON.stringify(e)}`)}function tN(e,t){return dU(e,t,"classWeight")}async function nN(e,t,n,a){if(t!=null||a!=null)throw new Error("Support sampleWeight is not implemented yet");if(n!=null){let r=O(()=>{if(e.shape.length===1)return Er(e);if(e.shape.length===2){if(e.shape[1]>1)return li(e,1);if(e.shape[1]===1)return W(e,[e.shape[0]]);throw new Error(`Encountered unexpected last-dimension size (${e.shape[1]}) during handling of class weights. The size is expected to be >= 1.`)}else throw new Error(`Unexpected rank of target (y) tensor (${e.rank}) during handling of class weights. The rank is expected to be 1 or 2.`)}),s=Array.from(await r.data());Re(r);let i=[];return s.forEach(o=>{if(n[o]==null)throw new Error(`classWeight must contain all classes in the training data. The class ${o} exists in the data but not in classWeight`);i.push(n[o])}),qe(i,"float32")}else return null}function hU(e,t){return B(e,t)}var mU=32;function aN(e,t){let n,a,r=t;n=r.xs,a=r.ys,w.assert(n!=null&&a!=null,()=>`A Dataset iterator for fitDataset() is expected to generate objects of the form \`{xs: xVal, ys: yVal}\`, where the two values may be \`tf.Tensor\`, an array of Tensors, or a map of string to Tensor. The provided Dataset instead generates ${t}`);let s=Sk("input",e.inputNames,n),i=Sk("output",e.outputNames,a),o=s[0].shape[0];w.assert(s.length===e.inputs.length,()=>`LayersModel has ${e.inputs.length} inputs, but the dataset provides ${s.length} inputs. (Expected input keys: ${JSON.stringify(e.inputNames)})`),w.assert(i.length===e.outputs.length,()=>`LayersModel has ${e.outputs.length} outputs, but the dataset provides ${i.length} outputs. (Expected output keys: ${JSON.stringify(e.outputNames)})`);for(let l=0;l<s.length;l++)w.assert(s[l].shape[0]===o,()=>`Batch size mismatch: input ${e.inputNames[l]} has ${s[l].shape[0]}; expected ${o} based on input ${e.inputNames[0]}.`);for(let l=0;l<i.length;l++)w.assert(i[l].shape[0]===o,()=>`Batch size mismatch: output ${e.outputNames[l]} has ${i[l].shape[0]}; expected ${o} based on input ${e.inputNames[0]}.`);return{xs:s,ys:i}}function Sk(e,t,n){if(n instanceof Fe)return[n];if(Array.isArray(n))return w.assert(n.length===t.length,()=>`Received an array of ${n.length} Tensors, but expected ${t.length} to match the ${e} keys ${t}.`),n;{let a=[];for(let r of t){if(n[r]==null)throw new H(`The feature data generated by the dataset lacks the required ${e} key '${r}'.`);a.push(n[r])}return a}}function fU(e){if(e.length===3)throw new Oe("Validation with sample weights is not implemented yet.");return{xs:e[0],ys:e[1]}}async function gU(e,t,n){let a=n.batchesPerEpoch!=null;if(w.assert(e.optimizer!=null,()=>"You must compile a model before training/testing. Use LayersModel.compile(modelCompileConfig)."),w.assert(n!=null,()=>"For fitDataset(), the 2nd argument (config) is required, but it is not provided in this call."),w.assert(n.epochs!=null&&n.epochs>0&&Number.isInteger(n.epochs),()=>`For fitDataset(), config.epochs is expected to be a positive integer, but got ${n.epochs}`),w.assert(!a||n.batchesPerEpoch>0&&Number.isInteger(n.batchesPerEpoch),()=>`For fitDataset(), config.batchesPerEpoch is expected to be a positive integer if specified, but got ${n.batchesPerEpoch}`),w.assert(n.validationSplit==null,()=>"`validationSplit` is not supported by `fitDataset()`. Use validationData instead."),e.isTraining)throw new Error("Cannot start training because another fit() call is ongoing.");e.isTraining=!0;try{let r=n.validationData!=null,s,i;if(r)if(Nk(n.validationData))w.assert(n.validationBatches==null||n.validationBatches>0&&Number.isInteger(n.validationBatches),()=>`For fitDataset() with dataset-based validation, config.validationBatches is expected not to be provided, or to be a positive integer, but got ${n.validationBatches}`);else{let g=fU(n.validationData);s=g.xs,i=g.ys}let o=e.makeTrainFunction(),l=e.getDedupedMetricsNames(),u;r?u=l.slice().concat(l.map(g=>"val_"+g)):u=l.slice();let p=q2(n.callbacks,n.yieldEvery),d=n.verbose==null?1:n.verbose,{callbackList:c,history:h}=K2(p,d,n.epochs,null,null,yU(t,n),null,r,u);c.setModel(e),e.history=h,await c.onTrainBegin(),e.stopTraining_=!1;let m=n.initialEpoch==null?0:n.initialEpoch,f=await t.iterator();for(;m<n.epochs;){let g={};await c.onEpochBegin(m);let y=0,b=0;for(a||(f=await t.iterator());!a||y<n.batchesPerEpoch;){let x=await f.next();if(a&&x.done){console.warn(`You provided \`batchesPerEpoch\` as ${n.batchesPerEpoch}, but your dataset iterator ran out of data after ${y} batches; interrupting training. Make sure that your dataset can generate at least \`batchesPerEpoch * epochs\` batches (in this case, ${n.batchesPerEpoch*n.epochs} batches). You may need to use the repeat() function when building your dataset.`);break}if(x.value!=null){let{xs:v,ys:k}=aN(e,x.value),T={};T.batch=b,T.size=v[0].shape[0],await c.onBatchBegin(b,T);let _=[];if(n.classWeight!=null){let M=tN(n.classWeight,e.outputNames);for(let $=0;$<M.length;++$)_.push(await nN(k[$],null,M[$]))}let E=v.concat(k).concat(_),A=o(E);Re(E);for(let M=0;M<l.length;++M){let $=l[M],S=A[M];T[$]=S,tn(S)}await c.onBatchEnd(b,T),U2(T),b++,y++}if(a?y>=n.batchesPerEpoch:x.done){if(r){let v;Nk(n.validationData)?v=kt(await e.evaluateDataset(n.validationData,{batches:n.validationBatches})):v=kt(e.evaluate(s,i,{batchSize:n.validationBatchSize==null?mU:n.validationBatchSize,verbose:0}));for(let k=0;k<e.metricsNames.length;++k)g[`val_${e.metricsNames[k]}`]=v[k]}break}if(e.stopTraining_)break}if(await c.onEpochEnd(m,g),m++,e.stopTraining_)break}return await c.onTrainEnd(),await e.history.syncData(),e.history}finally{e.isTraining=!1}}function yU(e,t){let n=null;return t.batchesPerEpoch!=null?n=t.batchesPerEpoch:Number.isFinite(e.size)&&(n=e.size),n}function Nk(e){return typeof e.iterator=="function"}function bU(e){return typeof e.next=="function"}async function xU(e,t,n){n=n||{};let a=n.batches!=null,r=e.testFunction,s=[];if(n.verbose>0)throw new Oe("Verbose mode is not implemented yet.");w.assert(!a||n.batches>0&&Number.isInteger(n.batches),()=>`Test loop expects \`batches\` to be a positive integer, but received ${JSON.stringify(n.batches)}`);let i=bU(t)?t:await t.iterator(),o=0,l=0;for(;!a||l<n.batches;){let u=await i.next();if(s=O(()=>{if(u.value){let{xs:p,ys:d}=aN(e,u.value),c=p.concat(d),h=O(()=>r(c));if(Re(c),l===0)for(let f=0;f<h.length;++f)s.push(we(0));let m=c[0].shape[0];for(let f=0;f<h.length;++f){let g=h[f],y=s[f];s[f]=O(()=>J(s[f],B(m,g))),l>0&&Re(y)}Re(h),o+=m,++l}return s}),u.done){a&&console.warn(`Your dataset iterator ran out of data during evaluateDataset(). Interrupting evalution. Make sure that your dataset can generate at least \`batches\` batches (in this case, ${n.batches} batches). You may need to use the repeat() function when building your dataset.`);break}}for(let u=0;u<s.length;++u){let p=s[u];s[u]=fe(s[u],o),Re(p)}return Pn(s)}function ex(e){w.assert(e>0&&Number.isInteger(e),()=>`batchSize is required to be a positive integer, but got ${e}`)}function Up(e,t,n){return e==null?[null]:Array.isArray(e)?e.map(a=>ti(a,t,n-t)):ti(e,t,n-t)}function uw(e,t){return O(()=>e==null?null:Array.isArray(e)?e.map(n=>uw(n,t)):M2(e,t.dtype==="int32"?t:oe(t,"int32")))}function tx(e,t){let n=[],a=0,r=null;for(;a<e;)r=a+t,r>=e&&(r=e),n.push([a,r]),a=r;return n}async function vU(e,t,n,a,r,s,i,o,l,u,p,d,c,h,m){r==null&&(r=32),s==null&&(s=1),p==null&&(p=!0),c==null&&(c=0);let f=!1;if(l!=null&&u!=null&&(f=!0),m!=null&&(f=!0,h==null))throw new H("Can only use `validationSteps` when doing step-wise training, i.e., `stepsPerEpoch` must be set.");let g=e.checkNumSamples(n,r,h,"steps_per_epoch"),y;g!=null&&(y=Ka(0,g)),i==null&&(i=1);let{callbackList:b,history:x}=K2(o,i,s,c,g,h,r,f,d);b.setModel(e),e.history=x,await b.onTrainBegin(),e.stopTraining_=!1;for(let v=c;v<s;++v){await b.onEpochBegin(v);let k={};if(h!=null)throw new Oe("stepsPerEpoch mode is not implemented yet.");{if(p==="batch")throw new Oe("batch shuffling is not implemneted yet");p&&w.shuffle(y);let T=qe(y),_=tx(g,r);for(let E=0;E<_.length;++E){let A={};if(await b.onBatchBegin(E,A),O(()=>{let M=_[E][0],$=_[E][1],S=ti(T,M,$-M);A.batch=E,A.size=$-M;let P=uw(n,S),V=t(P);for(let j=0;j<a.length;++j){let q=a[j],K=V[j];A[q]=K,tn(K)}if(E===_.length-1&&f){let j=e.testLoop(l,u,r);for(let q=0;q<a.length;++q){let K=a[q],Z=j[q];tn(Z),k["val_"+K]=Z}}}),await b.onBatchEnd(E,A),U2(A),e.stopTraining_)break}T.dispose()}if(await b.onEpochEnd(v,k),e.stopTraining_)break}return await b.onTrainEnd(),await e.history.syncData(),e.history}async function wU(e,t,n,a={}){if(e.isTraining)throw new Error("Cannot start training because another fit() call is ongoing.");e.isTraining=!0;let r,s,i,o,l,u,p,d,c;try{let h=a.batchSize==null?32:a.batchSize;ex(h);let m=!1,f=await e.standardizeUserData(t,n,a.sampleWeight,a.classWeight,m,h);r=f[0],s=f[1],c=f[2];let g=!1,y;if(a.validationData!=null&&a.validationData.length>0){if(g=!0,a.validationData.length===2)l=a.validationData[0],u=a.validationData[1];else throw a.validationData.length===3?new Oe("validationData including sample weights is not supported yet."):new H(`When passing validation data, it must contain 2 (valX, valY) or 3 (valX, valY, valSampleWeight) items; ${a.validationData} is invalid.`);let E=!0,A=await e.standardizeUserData(l,u,null,null,E,h);p=A[0],d=A[1],y=p.concat(d)}else if(a.validationSplit!=null&&a.validationSplit>0&&a.validationSplit<1){g=!0;let E=Math.floor(r[0].shape[0]*(1-a.validationSplit)),A=r[0].shape[0];p=Up(r,E,A),i=r,r=Up(r,0,E),d=Up(s,E,A),o=s,s=Up(s,0,E),y=p.concat(d)}else a.validationSteps!=null&&(g=!0);let b=r.concat(s).concat(c);e.checkTrainableWeightsConsistency();let x=e.makeTrainFunction(),v=e.getDedupedMetricsNames(),k,T;g?(e.makeTestFunction(),k=e.testFunction,T=v.slice().concat(v.map(E=>"val_"+E))):(k=null,y=[],T=v.slice());let _=q2(a.callbacks,a.yieldEvery);return await vU(e,x,b,v,h,a.epochs,a.verbose,_,k,y,a.shuffle,T,a.initialEpoch,null,null)}finally{e.isTraining=!1,Wa(r,t),Wa(s,n),Wa(i,t),Wa(o,n),Wa(p,l),Wa(d,u),c!=null&&Re(c)}}function rN(e){let t=[];e instanceof Fe&&(e=[e]);for(let n=0;n<e.length;++n){let a=e[n];if(a.rank===1)t.push(Zc(a,1));else{if(a.rank===0)throw new Error("Expected tensor to be at least 1D, but received a 0D tensor (scalar).");t.push(a)}}return t}function Wa(e,t){if(e==null)return;let n=[];if(t instanceof Fe)n.push(t.id);else if(Array.isArray(t))t.forEach(r=>n.push(r.id));else if(t!=null)for(let r in t){let s=t[r];n.push(s.id)}let a=[];if(e instanceof Fe)n.indexOf(e.id)===-1&&a.push(e);else if(Array.isArray(e))e.forEach(r=>{n.indexOf(r.id)===-1&&a.push(r)});else if(e!=null)for(let r in e){let s=e[r];n.indexOf(s.id)===-1&&a.push(s)}a.forEach(r=>{r.isDisposed||r.dispose()})}function kU(e){return e instanceof Fe}function nx(e){return Array.isArray(e)}function Tk(e){return!kU(e)&&!nx(e)}function _k(e,t,n,a=!0,r=""){if(t==null||t.length===0){if(e!=null){let i=!1;if(nx(e)&&e.length>0)i=!0;else if(Tk(e)){for(let o in e)if(e.hasOwnProperty(o)){i=!0;break}}else i=!0;if(i)throw new H(`Error when checking model ${r} expected no data, but got ${e}`)}return[]}if(e==null)return t.map(i=>null);let s;if(Tk(e)){e=e,s=[];for(let i of t){if(e[i]==null)throw new H(`No data provided for "${i}". Need data for each key in: ${t}`);s.push(e[i])}}else if(nx(e)){if(e=e,e.length!==t.length)throw new H(`Error when checking model ${r}: the Array of Tensors that you are passing to your model is not the size the model expected. Expected to see ${t.length} Tensor(s), but instead got the following list of Tensor(s): ${e}`);s=e}else{if(e=e,t.length>1)throw new H(`The model ${r} expects ${t.length} Tensor(s), but only received one Tensor. Found: Tensor with shape ${e.shape}`);s=[e]}if(s=rN(s),n!=null)for(let i=0;i<t.length;++i){if(n[i]==null)continue;let o=s[i];if(o.shape.length!==n[i].length)throw new H(`Error when checking ${r}: expected ${t[i]} to have ${n[i].length} dimension(s). but got array with shape ${o.shape}`);for(let l=0;l<n[i].length;++l){if(l===0&&!a)continue;let u=o.shape[l],p=n[i][l];if(p!=null&&p>=0&&u!==p)throw new H(`${r} expected a batch of elements where each example has shape [${n[i].slice(1,n[i].length)}] (i.e.,tensor shape [*,${n[i].slice(1,n[i].length)}]) but the ${r} received an input with ${o.shape[0]} examples, each with shape [${o.shape.slice(1,o.shape.length)}] (tensor shape [${o.shape}])`)}}return s}function IU(e,t,n){let a=ss(e.map(s=>s.shape[0]));a.sort();let r=ss(t.map(s=>s.shape[0]));if(r.sort(),a.length>1)throw new H(`All input Tensors (x) should have the same number of samples. Got array shapes: ${JSON.stringify(e.map(s=>s.shape))}`);if(r.length>1)throw new H(`All target Tensors (y) should have the same number of samples. Got array shapes: ${JSON.stringify(t.map(s=>s.shape))}`);if(a.length>0&&r.length>0&&!w.arraysEqual(a,r))throw new H(`Input Tensors should have the same number of samples as target Tensors. Found ${a[0]} input sample(s) and ${r[0]} target sample(s).`)}function SU(e,t,n){let a=[ko,Xf,lc];for(let r=0;r<e.length;++r){let s=e[r],i=t[r],o=n[r];if(i!=null){if(i===lc&&s.shape[s.shape.length-1]===1)throw new H(`You are passing a target array of shape ${s.shape} while using a loss 'categorical_crossentropy'. 'categorical_crossentropy'expects targets to be binary matrices (1s and 0s) of shape [samples, classes].`);if(a.indexOf(i)!==-1){let l=s.shape.slice(1),u=o.slice(1);for(let p=0;p<l.length;++p){let d=l[p],c=u[p];if(c!=null&&d!==c)throw new H(`A target Tensor with shape ${s.shape} was passed for an output of shape ${o}, while using a loss function that expects targets to have the same shape as the output.`)}}}}}function Ck(e,t,n,a=!0,r=""){let s;if(Array.isArray(e)){if(e.length!==t.length)throw new H(`Error when checking model ${r}: the Array of Tensors that you are passing to your model is not the size the the model expected. Expected to see ${t.length} Tensor(s), but instead got ${e.length} Tensors(s).`);s=e}else{if(t.length>1)throw new H(`The model expects ${t.length} ${r} Tensors, but only received one Tensor. Found: array with shape ${JSON.stringify(e.shape)}.`);s=[e]}if(n!=null)for(let i=0;i<t.length;++i){if(n[i]==null)continue;let o=s[i];if(o.shape.length!==n[i].length)throw new H(`Error when checking ${r}: expected ${t[i]} to have ${n[i].length} dimension(s), but got array with shape ${JSON.stringify(o.shape)}`);for(let l=0;l<n[i].length;++l){if(l===0&&!a)continue;let u=o.shape[l],p=n[i][l];if(p!=null&&p!==u)throw new H(`Error when checking ${r}: expected ${t[i]} to have shape ${JSON.stringify(n[i])} but got array with shape ${JSON.stringify(o.shape)}.`)}}}function NU(e,t){if(e==null||Array.isArray(e)&&e.length===0)return t.map(a=>[]);let n;if(typeof e=="string"||typeof e=="function")n=[e];else if(Array.isArray(e)||typeof e=="object")n=e;else throw new TypeError(`Type of metrics argument not understood. Expected an string,function, Array, or Object, found: ${e}`);if(Array.isArray(n))return t.map(a=>n);{let a=[];for(let r of t){let s=n.hasOwnProperty(r)?n[r]:[];Array.isArray(s)||(s=[s]),a.push(s)}return a}}var TU="layers-model",Fr=class extends rr{constructor(e){super(e),this.isTraining=!1}summary(e,t,n=console.log){if(!this.built)throw new H("This model has never been called, thus its weights have not been created yet. So no summary can be displayed. Build the model first (e.g., by calling it on some test data).");oU(this,e,t,n)}compile(e){if(e.loss==null&&(e.loss=[]),this.loss=e.loss,typeof e.optimizer=="string")this.optimizer_=iU(e.optimizer),this.isOptimizerOwned=!0;else{if(!(e.optimizer instanceof Mr))throw new H("User-defined optimizer must be an instance of tf.Optimizer.");this.optimizer_=e.optimizer,this.isOptimizerOwned=!1}let t=[];if(!Array.isArray(e.loss)&&typeof e.loss!="string"&&typeof e.loss!="function"){e.loss=e.loss;for(let s in e.loss)if(this.outputNames.indexOf(s)===-1)throw new H(`Unknown entry in loss dictionary: "${s}". Only expected the following keys: ${this.outputNames}`);for(let s of this.outputNames)e.loss[s]==null&&console.warn(`Output "${s}" is missing from loss dictionary. We assume this was done on purpose, and we will not be expecting data to be passed to ${s} during training`),t.push(Sb(e.loss[s]))}else if(Array.isArray(e.loss)){if(e.loss.length!==this.outputs.length)throw new H(`When passing an Array as loss, it should have one entry per model output. The model has ${this.outputs.length} output(s), but you passed loss=${e.loss}.`);t=e.loss.map(s=>Sb(s))}else{let s=Sb(e.loss);this.outputs.forEach(i=>{t.push(s)})}this.lossFunctions=t,this.feedOutputNames=[],this.feedOutputShapes=[],this.feedLossFns=[];for(let s=0;s<this.outputs.length;++s){let i=this.internalOutputShapes[s],o=this.outputNames[s];this.feedOutputNames.push(o),this.feedOutputShapes.push(i),this.feedLossFns.push(this.lossFunctions[s])}let n=[];this.metrics=e.metrics,this.metricsNames=["loss"],this.metricsTensors=[],ei("loss",()=>{for(let s=0;s<this.outputs.length;++s){if(n.indexOf(s)!==-1)continue;let i=this.lossFunctions[s];this.outputs.length>1&&(this.metricsTensors.push([i,s]),this.metricsNames.push(this.outputNames[s]+"_loss"))}});let a=NU(e.metrics,this.outputNames),r=(s,i,o)=>{this.outputNames.length>1&&(i=this.outputNames[s]+"_"+i),this.metricsNames.push(i),this.metricsTensors.push([o,s])};ei("metric",()=>{for(let s=0;s<this.outputs.length;++s){if(n.indexOf(s)!==-1)continue;let i=a[s];(o=>{let l="",u,p,d;for(let c of o){if(typeof c=="string"&&["accuracy","acc","crossentropy","ce"].indexOf(c)!==-1){let m=this.internalOutputShapes[s];m[m.length-1]===1||this.lossFunctions[s]===Xf?["accuracy","acc"].indexOf(c)!==-1?p=sw:["crossentropy","ce"].indexOf(c)!==-1&&(p=Q2):this.lossFunctions[s]===sm?["accuracy","acc"].indexOf(c)!==-1?p=J2:["crossentropy","ce"].indexOf(c)!==-1&&(p=Z2):["accuracy","acc"].indexOf(c)!==-1?p=iw:["crossentropy","ce"].indexOf(c)!==-1&&(p=ow);let f;["accuracy","acc"].indexOf(c)!==-1?f="acc":["crossentropy","ce"].indexOf(c)!==-1&&(f="ce"),d=p,u=l+f}else d=sU(c),u=l+Ih(c);let h;ei(u,()=>{h=d}),r(s,u,h)}})(i)}}),this.collectedTrainableWeights=this.trainableWeights}checkTrainableWeightsConsistency(){this.collectedTrainableWeights!=null&&this.trainableWeights.length!==this.collectedTrainableWeights.length&&console.warn("Discrepancy between trainableweights and collected trainable weights. Did you set `model.trainable` without calling `model.compile()` afterwards?")}evaluate(e,t,n={}){let a=n.batchSize==null?32:n.batchSize;ex(a);let r=!0,s=this.standardizeUserDataXY(e,t,r,a);try{let i=s[0].concat(s[1]);this.makeTestFunction();let o=this.testFunction,l=this.testLoop(o,i,a,n.verbose,n.steps);return Pn(l)}finally{Wa(s[0],e),Wa(s[1],t)}}async evaluateDataset(e,t){return this.makeTestFunction(),xU(this,e,t)}checkNumSamples(e,t,n,a="steps"){let r;if(n!=null){if(r=null,t!=null)throw new H(`If ${a} is set, batchSize must be null or undefined.Got batchSize = ${t}`)}else if(e!=null)Array.isArray(e)?r=e[0].shape[0]:r=e.shape[0];else throw new H(`Either the input data should have a defined shape, or ${a} shoud be specified.`);return r}execute(e,t){if(Array.isArray(t)&&t.length===0)throw new H("`outputs` is an empty Array, which is not allowed.");let n=Array.isArray(t),a=n?t:[t],r=this.retrieveSymbolicTensors(a),s=new Qs;if(e instanceof Fe&&(e=[e]),Array.isArray(e)){if(e.length!==this.inputs.length)throw new H(`The number of inputs provided (${e.length}) does not match the number of inputs of this model (${this.inputs.length}).`);for(let o=0;o<this.inputs.length;++o)s.add(this.inputs[o],e[o])}else for(let o of this.inputs){let l=e[o.name];if(l==null)throw new H(`No value is provided for the model's input ${o.name}`);s.add(o,l)}let i=Vp(r,s);return n?i:i[0]}retrieveSymbolicTensors(e){let t=hi(null,e.length),n=e.length;for(let a of this.layers){let r=Array.isArray(a.output)?a.output:[a.output],s=r.map(i=>i.name);for(let i=0;i<e.length;++i){let o=s.indexOf(e[i]);if(o!==-1&&(t[i]=r[o],n--),n===0)break}if(n===0)break}if(n>0){let a=[];throw t.forEach((r,s)=>{r==null&&a.push(e[s])}),new H(`Cannot find SymbolicTensors for output name(s): ${JSON.stringify(a)}`)}return t}predictLoop(e,t=32,n=!1){return O(()=>{let a=this.checkNumSamples(e);if(n)throw new Oe("Verbose predictLoop() is not implemented yet.");let r=tx(a,t),s=this.outputs.map(i=>[]);for(let i=0;i<r.length;++i)O(()=>{let o=r[i][0],l=r[i][1],u=Up(e,o,l),p=[];if(Array.isArray(u))for(let c=0;c<u.length;++c)p.push({key:this.inputs[c],value:u[c]});else p.push({key:this.inputs[0],value:u});let d=new Qs(p);return Vp(this.outputs,d)}).forEach((o,l)=>s[l].push(o));return Pn(s.map(i=>Ze(i,0)))})}predict(e,t={}){let n=rN(e);Ck(n,this.inputNames,this.feedInputShapes,!1);try{let a=t.batchSize==null?32:t.batchSize;return ex(a),this.predictLoop(n,a)}finally{Wa(n,e)}}predictOnBatch(e){Ck(e,this.inputNames,this.feedInputShapes,!0);let t=(Array.isArray(e)?e[0]:e).shape[0];return this.predictLoop(e,t)}standardizeUserDataXY(e,t,n=!0,a){if(this.optimizer_==null)throw new Va("You must compile a model before training/testing. Use LayersModel.compile(modelCompileArgs).");let r=[];for(let s=0;s<this.feedOutputShapes.length;++s){let i=this.feedOutputShapes[s];this.feedLossFns[s]===sm?r.push(i.slice(0,i.length-1).concat([1])):r.push(i)}if(e=_k(e,this.feedInputNames,this.feedInputShapes,!1,"input"),t=_k(t,this.feedOutputNames,r,!1,"target"),IU(e,t,null),SU(t,this.feedLossFns,this.feedOutputShapes),this.stateful&&a!=null&&a>0&&e[0].shape[0]%a!==0)throw new H(`In a stateful network, you should only pass inputs with a number of samples that is divisible by the batch size ${a}. Found: ${e[0].shape[0]} sample(s).`);return[e,t]}async standardizeUserData(e,t,n,a,r=!0,s){let[i,o]=this.standardizeUserDataXY(e,t,r,s);if(n!=null)throw new Error("sample weight is not supported yet.");let l=null;if(a!=null){let u=tN(a,this.outputNames);l=[];for(let p=0;p<u.length;++p)l.push(await nN(o[p],null,u[p]))}return[i,o,l]}testLoop(e,t,n,a=0,r){return O(()=>{let s=this.checkNumSamples(t,n,r,"steps"),i=[];if(a>0)throw new Oe("Verbose mode is not implemented yet.");if(r!=null)throw new Oe("steps mode in testLoop() is not implemented yet");{let o=tx(s,n),l=qe(Ka(0,s));for(let u=0;u<o.length;++u){let p=o[u][0],d=o[u][1],c=ti(l,p,d-p),h=uw(t,c),m=e(h);if(u===0)for(let f=0;f<m.length;++f)i.push(we(0));for(let f=0;f<m.length;++f){let g=m[f];i[f]=J(i[f],B(d-p,g))}}for(let u=0;u<i.length;++u)i[u]=fe(i[u],s)}return i})}getDedupedMetricsNames(){let e=this.metricsNames,t=[];for(let n=0;n<e.length;++n){let a=e[n],r=a;hk(e,a)>1&&(r+=`_${hk(e.slice(0,n),a)}`),t.push(r)}return t}makeTrainFunction(){return e=>{let t=[],n=e.slice(0,this.inputs.length),a=e.slice(this.inputs.length,this.inputs.length+this.outputs.length),r=e.slice(this.inputs.length+this.outputs.length,this.inputs.length+this.outputs.length*2),s=[],i=()=>{let u=[];for(let h=0;h<this.inputs.length;++h)u.push({key:this.inputs[h],value:n[h]});let p=new Qs(u),d=Vp(this.outputs,p,{training:!0}),c;for(let h=0;h<this.lossFunctions.length;++h){let m=this.lossFunctions[h](a[h],d[h]);r[h]!=null&&(m=hU(m,r[h]));let f=Et(m);t.push(f),h===0?c=m:c=J(c,m)}for(let h=0;h<this.metricsTensors.length;++h){let m;if(this.outputs.length>1&&h<this.outputs.length)m=t[h];else{let f=this.metricsTensors[h][0],g=this.metricsTensors[h][1];m=Et(f(a[g],d[g]))}tn(m),s.push(m)}return c=Et(c),this.calculateLosses().forEach(h=>{c=J(c,h)}),c},o=this.collectedTrainableWeights.map(u=>u.read()),l=!0;return[this.optimizer_.minimize(i,l,o)].concat(s)}}makeTestFunction(){this.testFunction=e=>O(()=>{let t=[],n,a=e.slice(0,this.inputs.length),r=e.slice(this.inputs.length,this.inputs.length+this.outputs.length),s=[];for(let l=0;l<this.inputs.length;++l)s.push({key:this.inputs[l],value:a[l]});let i=new Qs(s),o=Vp(this.outputs,i);for(let l=0;l<this.lossFunctions.length;++l){let u=this.lossFunctions[l],p=Et(u(r[l],o[l]));l===0?n=p:n=J(n,p),t.push(n)}for(let l=0;l<this.metricsTensors.length;++l){let u=this.metricsTensors[l][0],p=this.metricsTensors[l][1],d=Et(u(r[p],o[p]));t.push(d)}return t})}async fit(e,t,n={}){return wU(this,e,t,n)}async fitDataset(e,t){return gU(this,e,t)}async trainOnBatch(e,t){let n=await this.standardizeUserData(e,t),a=n[0],r=n[1],s=this.makeTrainFunction()(a.concat(r)),i=[];for(let o of s){let l=await o.data();i.push(l[0])}return Re(s),Wa(n[0],e),Wa(n[1],t),Pn(i)}getNamedWeights(e){let t=[],n=e!=null&&e.trainableOnly,a=n?this.trainableWeights:this.weights,r=this.getWeights(n);for(let s=0;s<a.length;++s)n&&!a[s].trainable||t.push({name:a[s].originalName,tensor:r[s]});return t}set stopTraining(e){this.stopTraining_=e}get stopTraining(){return this.stopTraining_}get optimizer(){return this.optimizer_}set optimizer(e){this.optimizer_!==e&&(this.optimizer_=e,this.isOptimizerOwned=!1)}dispose(){let e=super.dispose();if(e.refCountAfterDispose===0&&this.optimizer!=null&&this.isOptimizerOwned){let t=Qh().numTensors;this.optimizer_.dispose(),e.numDisposedVariables+=t-Qh().numTensors}return e}getLossIdentifiers(){let e;if(typeof this.loss=="string")e=Sr(this.loss);else if(Array.isArray(this.loss)){for(let t of this.loss)if(typeof t!="string")throw new Error("Serialization of non-string loss is not supported.");e=this.loss.map(t=>Sr(t))}else{let t=Object.keys(this.loss);e={};let n=this.loss;for(let a of t)if(typeof n[a]=="string")e[a]=Sr(n[a]);else throw new Error("Serialization of non-string loss is not supported.")}return e}getMetricIdentifiers(){if(typeof this.metrics=="string"||typeof this.metrics=="function")return[Sr(Ih(this.metrics))];if(Array.isArray(this.metrics))return this.metrics.map(e=>Sr(Ih(e)));{let e={};for(let t in this.metrics)e[t]=Sr(Ih(this.metrics[t]));return e}}getTrainingConfig(){return{loss:this.getLossIdentifiers(),metrics:this.getMetricIdentifiers(),optimizer_config:{class_name:this.optimizer.getClassName(),config:this.optimizer.getConfig()}}}loadTrainingConfig(e){if(e.weighted_metrics!=null)throw new Error("Loading weight_metrics is not supported yet.");if(e.loss_weights!=null)throw new Error("Loading loss_weights is not supported yet.");if(e.sample_weight_mode!=null)throw new Error("Loading sample_weight_mode is not supported yet.");let t=uc(e.optimizer_config),n=ja(t),a;if(typeof e.loss=="string")a=Ks(e.loss);else if(Array.isArray(e.loss))a=e.loss.map(s=>Ks(s));else if(e.loss!=null){a={};for(let s in e.loss)a[s]=Ks(e.loss[s])}let r;if(Array.isArray(e.metrics))r=e.metrics.map(s=>Ks(s));else if(e.metrics!=null){r={};for(let s in e.metrics)r[s]=Ks(e.metrics[s])}this.compile({loss:a,metrics:r,optimizer:n})}async save(e,t){if(typeof e=="string"){let i=en.getSaveHandlers(e);if(i.length===0)throw new H(`Cannot find any save handlers for URL '${e}'`);if(i.length>1)throw new H(`Found more than one (${i.length}) save handlers for URL '${e}'`);e=i[0]}if(e.save==null)throw new H("LayersModel.save() cannot proceed because the IOHandler provided does not have the `save` attribute defined.");let n=await en.encodeWeights(this.getNamedWeights(t)),a=!1,r=null,s={modelTopology:this.toJSON(r,a),format:TU,generatedBy:`TensorFlow.js tfjs-layers v${lw}`,convertedBy:null};if((t==null?!1:t.includeOptimizer)&&this.optimizer!=null){s.trainingConfig=this.getTrainingConfig();let i="optimizer",{data:o,specs:l}=await en.encodeWeights(await this.optimizer.getWeights(),i);n.specs.push(...l),n.data=en.concatenateArrayBuffers([n.data,o])}return this.userDefinedMetadata!=null&&(Ik(this.userDefinedMetadata,this.name,!0),s.userDefinedMetadata=this.userDefinedMetadata),s.weightData=n.data,s.weightSpecs=n.specs,e.save(s)}setUserDefinedMetadata(e){Ik(e,this.name),this.userDefinedMetadata=e}getUserDefinedMetadata(){return this.userDefinedMetadata}};Fr.className="Model";se.registerClass(Fr);var sN=class extends Fr{};sN.className="Functional";se.registerClass(sN);async function _U(e,t){"modelTopology"in e||(e={modelTopology:e}),e=e;let n=e.modelTopology;n.model_config!=null&&(n=n.model_config);let a=uc(n),r=ja(a,t);if(e.weightsManifest!=null){let s=await en.loadWeights(e.weightsManifest,e.pathPrefix,r.weights.map(o=>o.originalName)),i={};for(let o of r.weights)i[o.originalName]=s[o.originalName];r.loadWeights(i),Re(s)}return r}async function CU(e,t){if(t==null&&(t={}),typeof e=="string"){let n=en.getLoadHandlers(e,t);if(n.length===0)n.push(en.browserHTTPRequest(e,t));else if(n.length>1)throw new H(`Found more than one (${n.length}) load handlers for URL '${e}'`);e=n[0]}return EU(e,void 0,t)}async function EU(e,t,n){if(n==null&&(n={}),e.load==null)throw new H("Cannot proceed with model loading because the IOHandler provided does not have the `load` method implemented.");let a=await e.load(),r=a.modelTopology;r.model_config!=null&&(r=r.model_config);let s=n.strict==null?!0:n.strict,i=a.weightData!=null&&a.weightSpecs!=null&&s,o=ja(uc(r),t,i),l=a.trainingConfig;if(l!=null&&o.loadTrainingConfig(l),a.userDefinedMetadata!=null&&o.setUserDefinedMetadata(a.userDefinedMetadata),a.weightData!=null){if(a.weightSpecs==null)throw new H("LayersModel artifacts contains weight data, but not weight specs. Therefore loading of weights cannot proceed.");let{modelWeights:u,optimizerWeights:p}=FU(a.weightData,a.weightSpecs);o.loadWeights(u,s),o.optimizer!=null&&p.length>0&&await o.optimizer.setWeights(p),Re(u),Re(p.map(d=>d.tensor))}return o}function FU(e,t){let n=en.decodeWeights(e,t),a={},r=[];return t.forEach(s=>{s.group==="optimizer"?r.push({name:s.name,tensor:n[s.name]}):a[s.name]=n[s.name]}),{modelWeights:a,optimizerWeights:r}}var Fl=class extends Fr{constructor(e){if(super({inputs:[],outputs:[]}),e=e||{},this.trainable=!0,this.built=!1,this.name=e.name!=null?e.name:Of("sequential_"),e.layers!=null)for(let t of e.layers)this.add(t)}checkShape(e){if(e.inboundNodes[0].outputTensors[0].shape.some(t=>t<0))throw new H(`Negative dimension size caused by adding layer ${e.name} with input shape [${e.inboundNodes[0].inputTensors[0].shape}]`)}add(e){let t=e instanceof Fl||e instanceof Fr,n;if(t){if(n=e,n.outputs.length!==1)throw new H("All layers in a Sequential model should have a single output tensor. For multi-output layers, use the functional API.");if(n.inputs.length!==1)throw new H("All layers in a Sequential model should have a single input tensor. For multi-input layers, use the functional API.")}if(this.outputs.length===0){if(e.inboundNodes.length===0){if(e.batchInputShape==null)throw new H("The first layer in a Sequential model must get an `inputShape` or `batchInputShape` argument.");let a=z2({batchShape:e.batchInputShape,dtype:e.dtype,name:e.name+"_input"});e.apply(a)}if(t)this.outputs=n.outputs,this.inputs=n.inputs;else{if(e.inboundNodes.length!==1)throw new H(`A layer added to a Sequential model must not already be connected somewhere else. LayersModel received layer ${e.name} which has ${e.inboundNodes.length} pre-existing inbound connections.`);if(e.inboundNodes[0].outputTensors.length!==1)throw new H("All layers in a Sequential model should have a single output tensor. For multi-output layers, use the functional API.");this.checkShape(e),this.outputs=[e.inboundNodes[0].outputTensors[0]],this.inputs=L2(this.outputs[0])}this.inboundNodes=[],new qf({outboundLayer:this,inboundLayers:[],nodeIndices:[],tensorIndices:[],inputTensors:this.inputs,outputTensors:this.outputs,inputMasks:hi(null,this.inputs.length),outputMasks:[null],inputShapes:this.inputs.map(a=>a.shape),outputShapes:this.outputs[0].shape})}else{let a=e.apply(this.outputs[0]);if(Array.isArray(a))throw new TypeError("All layers in a Sequential model should have a single output tensor. For multi-output layers, use the functional API.");this.checkShape(e),this.outputs=[a],this.inboundNodes[0].outputTensors=this.outputs,this.inboundNodes[0].outputShapes=[this.outputs[0].shape]}this.layers.push(e),this.built=!1}pop(){if(this.layers.length===0)throw new TypeError("There are no layers in the model.");if(this.layers.pop(),this.layers.length===0)this.outputs=[],this.inboundNodes=[],this.outboundNodes=[];else{let e=this.layers.length-1;this.layers[e].outboundNodes=[],this.outputs=[this.layers[e].output],this.inboundNodes[0].outputTensors=this.outputs,this.inboundNodes[0].outputShapes=[this.outputs[0].shape]}}call(e,t){return this.model==null&&this.build(),this.model.call(e,t)}build(e){if(it(e),this.inputs.length===0||this.outputs.length===0)throw new TypeError("Sequential model cannot be built: model is empty. Add some layers first.");this.model=new Fr({inputs:this.inputs,outputs:this.outputs[0],name:this.name+"_model"}),this.model.trainable=this.trainable,this.supportsMasking=this.model.supportsMasking,this.inputLayers=this.model.inputLayers,this.inputLayersNodeIndices=this.model.inputLayersNodeIndices,this.inputLayersTensorIndices=this.model.inputLayersTensorIndices,this.outputLayers=this.model.outputLayers,this.outputLayersNodeIndices=this.model.outputLayersNodeIndices,this.outputLayersTensorIndices=this.model.outputLayersTensorIndices,this.nodesByDepth=this.model.nodesByDepth,this.containerNodes=this.model.containerNodes,this.outputNames=this.model.outputNames,this.inputNames=this.model.inputNames,this.built=!0}countParams(){return this.built||this.build(),super.countParams()}summary(e,t,n=console.log){this.built||this.build(),super.summary(e,t,n)}setWeights(e){this.model==null&&this.build(),this.model.setWeights(e)}evaluate(e,t,n={}){if(!this.built)throw new Va("The model needs to be compiled before being used.");return this.model.evaluate(e,t,n)}async evaluateDataset(e,t){if(!this.built)throw new Va("The model needs to be compiled before being used.");return this.model.evaluateDataset(e,t)}predict(e,t={}){return this.model==null&&this.build(),this.model.predict(e,t)}predictOnBatch(e){return this.model==null&&this.build(),this.model.predictOnBatch(e)}compile(e){this.build(),this.model.compile(e),this.optimizer_=this.model.optimizer,this.isOptimizerOwned=this.model.isOptimizerOwned,this.loss=this.model.loss,this.metrics=this.model.metrics,this.metricsTensors=this.model.metricsTensors,this.metricsNames=this.model.metricsNames}get optimizer(){return this.model==null?void 0:this.model.optimizer}set optimizer(e){this.model.optimizer=e}async fit(e,t,n={}){if(!this.built)throw new Va("The model needs to be compiled before being used.");return this.model.fit(e,t,n)}async fitDataset(e,t){if(!this.built)throw new Va("The model needs to be compiled before being used.");return this.model.fitDataset(e,t)}async trainOnBatch(e,t){return this.model.trainOnBatch(e,t)}static fromConfig(e,t,n={},a=!1){let r,s={};if(t instanceof Array){if(t[0].className==null||t[0].className==="Merge")throw new H("Legacy serialization format not supported yet.");r=t}else w.assert(t.layers!=null,()=>"When the config data for a Sequential model is not an Array, it must be an Object that contains the 'layers' field."),r=t.layers,delete t.layers,s=t;let i=new e(s);if(!(i instanceof Fl))throw new Oe(`Sequential.fromConfig called on non-Sequential input: ${i}`);for(let o of r){let l=ja(o,void 0,a);a&&l.setFastWeightInitDuringBuild(!0),i.add(l)}return i}set stopTraining(e){if(this.model==null)throw new H("Cannot set the stopTraining property of a sequential model before it is compiled.");this.model.stopTraining=e}get stopTraining(){if(this.model==null)throw new H("Cannot get the stopTraining property of a sequential model before it is compiled.");return this.model.stopTraining}getConfig(){let e=[];for(let t of this.layers){let n={};n.className=t.getClassName(),n.config=t.getConfig(),e.push(n)}return{name:this.name,layers:e}}};Fl.className="Sequential";se.registerClass(Fl);function AU(e){return new Fr(e)}function $U(e){return new Fl(e)}function DU(e,t){return t==null&&(t={}),CU(e,t)}function iN(e){return z2(e)}function RU(e,t){Sa.registerCallbackConstructor(e,t)}var Gn=class extends se.Serializable{getConfig(){return{}}},oN=class extends Gn{apply(e,t=1){return tV(e,t)}};oN.className="elu";se.registerClass(oN);var lN=class extends Gn{apply(e){return vf(e)}};lN.className="selu";se.registerClass(lN);var uN=class extends Gn{apply(e){return Xe(e)}};uN.className="relu";se.registerClass(uN);var pN=class extends Gn{apply(e){return O(()=>Uu(6,Xe(e)))}};pN.className="relu6";se.registerClass(pN);var cN=class extends Gn{apply(e){return e}};cN.className="linear";se.registerClass(cN);var dN=class extends Gn{apply(e){return ma(e)}};dN.className="sigmoid";se.registerClass(dN);var hN=class extends Gn{apply(e){return aV(e)}};hN.className="hardSigmoid";se.registerClass(hN);var mN=class extends Gn{apply(e){return xo(e)}};mN.className="softplus";se.registerClass(mN);var fN=class extends Gn{apply(e){return nV(e)}};fN.className="softsign";se.registerClass(fN);var gN=class extends Gn{apply(e){return ui(e)}};gN.className="tanh";se.registerClass(gN);var pw=class extends Gn{apply(e,t=-1){return Qa(e,t)}};pw.className="softmax";se.registerClass(pw);var yN=class extends Gn{apply(e,t=-1){return hf(e,t)}};yN.className="logSoftmax";se.registerClass(yN);var bN=class extends Gn{apply(e,t=1){return O(()=>B(ma(B(e,t)),e))}};bN.className="swish";se.registerClass(bN);var xN=class extends Gn{apply(e){return O(()=>B(e,ui(xo(e))))}};xN.className="mish";se.registerClass(xN);function ds(e){return e.getClassName()}function Nb(e,t={}){return Jc(e,se.SerializationMap.getMap().classNameMap,t,"activation")}function hs(e){if(e==null){let t={};return t.className="linear",t.config={},Nb(t)}if(typeof e=="string"){let t={};return t.className=e,t.config={},Nb(t)}else return e instanceof Gn?e:Nb(e)}function cw(e){if(e!=null&&typeof e!="object")throw new Error(`Argument to L1L2 regularizer's constructor is expected to be an object, but received: ${e}`)}var vN=class extends se.Serializable{},ad=class extends vN{constructor(e){super(),cw(e),this.l1=e==null||e.l1==null?.01:e.l1,this.l2=e==null||e.l2==null?.01:e.l2,this.hasL1=this.l1!==0,this.hasL2=this.l2!==0}apply(e){return O(()=>{let t=St([1]);return this.hasL1&&(t=J(t,be(B(this.l1,zt(e))))),this.hasL2&&(t=J(t,be(B(this.l2,ed(e))))),W(t,[])})}getConfig(){return{l1:this.l1,l2:this.l2}}static fromConfig(e,t){return new e({l1:t.l1,l2:t.l2})}};ad.className="L1L2";se.registerClass(ad);function MU(e){return cw(e),new ad({l1:e!=null?e.l1:null,l2:0})}function PU(e){return cw(e),new ad({l2:e!=null?e.l2:null,l1:0})}var Ek={l1l2:"L1L2"};function dt(e){return Vv(e)}function Fk(e,t={}){return Jc(e,se.SerializationMap.getMap().classNameMap,t,"regularizer")}function Tt(e){if(e==null)return null;if(typeof e=="string"){let t={className:e in Ek?Ek[e]:e,config:{}};return Fk(t)}else return e instanceof vN?e:Fk(e)}var dw=class extends Ye{constructor(e){super(e==null?{}:e),this.supportsMasking=!0,e!=null&&(this.maxValue=e.maxValue)}call(e,t){e=ze(e);let n=Xe(e);return this.maxValue!=null&&(n=an(n,0,this.maxValue)),n}computeOutputShape(e){return e}getConfig(){let e={maxValue:this.maxValue},t=super.getConfig();return Object.assign(e,t),e}};dw.className="ReLU";se.registerClass(dw);var hw=class extends Ye{constructor(e){super(e==null?{}:e),this.DEFAULT_ALPHA=.3,e==null&&(e={}),this.alpha=e.alpha==null?this.DEFAULT_ALPHA:e.alpha}call(e,t){let n=ze(e);return Gc(n,this.alpha)}computeOutputShape(e){return e}getConfig(){let e={alpha:this.alpha},t=super.getConfig();return Object.assign(e,t),e}};hw.className="LeakyReLU";se.registerClass(hw);var mw=class extends Ye{constructor(e){if(super(e==null?{}:e),this.DEFAULT_ALPHA_INITIALIZER="zeros",e==null&&(e={}),this.supportsMasking=!0,this.alphaInitializer=Nt(e.alphaInitializer||this.DEFAULT_ALPHA_INITIALIZER),this.alphaRegularizer=Tt(e.alphaRegularizer),this.alphaConstraint=Yt(e.alphaConstraint),e.sharedAxes==null)this.sharedAxes=null;else if(Array.isArray(e.sharedAxes))this.sharedAxes=e.sharedAxes;else if(typeof e.sharedAxes=="number")this.sharedAxes=[e.sharedAxes];else throw new H(`Expected sharedAxes to be a number or an array of numbers, but got ${e.sharedAxes}`)}build(e){e=it(e);let t=e.slice(1);if(this.sharedAxes!=null)for(let a of this.sharedAxes)t[a-1]=1;this.alpha=this.addWeight("alpha",t,"float32",this.alphaInitializer,this.alphaRegularizer,!0,this.alphaConstraint);let n={};if(this.sharedAxes!=null)for(let a=1;a<e.length;++a)n[a]=e[a];this.inputSpec=[new Bt({ndim:e.length,axes:n})],this.built=!0}call(e,t){return e=ze(e),Kc(e,this.alpha.read())}getConfig(){let e={alphaInitializer:Ft(this.alphaInitializer),alphaRegularizer:dt(this.alphaRegularizer),alphaConstraint:Xt(this.alphaConstraint),sharedAxes:this.sharedAxes},t=super.getConfig();return Object.assign(e,t),e}};mw.className="PReLU";se.registerClass(mw);var fw=class extends Ye{constructor(e){if(super(e==null?{}:e),this.DEFAULT_ALPHA=1,e==null&&(e={}),e.alpha!=null&&e.alpha!==this.DEFAULT_ALPHA)throw new Oe(`Non-default alpha value (${e.alpha}) is not supported by the ELU layer yet.`);this.alpha=e.alpha==null?this.DEFAULT_ALPHA:e.alpha}call(e,t){let n=ze(e);return Wu(n)}computeOutputShape(e){return e}getConfig(){let e={alpha:this.alpha},t=super.getConfig();return Object.assign(e,t),e}};fw.className="ELU";se.registerClass(fw);var gw=class extends Ye{constructor(e){super(e==null?{}:e),this.DEFAULT_THETA=1,e==null&&(e={}),this.theta=e.theta==null?this.DEFAULT_THETA:e.theta}call(e,t){let n=ze(e);return B(n,oe(Un(n,this.theta),"float32"))}computeOutputShape(e){return e}getConfig(){let e={theta:this.theta},t=super.getConfig();return Object.assign(e,t),e}};gw.className="ThresholdedReLU";se.registerClass(gw);var yw=class extends Ye{constructor(e){super(e==null?{}:e),this.DEFAULT_AXIS=1,e==null&&(e={}),this.softmax=new pw().apply,this.axis=e.axis==null?this.DEFAULT_AXIS:e.axis}call(e,t){let n=ze(e);return this.softmax(n,this.axis)}computeOutputShape(e){return e}getConfig(){let e={axis:this.axis},t=super.getConfig();return Object.assign(e,t),e}};yw.className="Softmax";se.registerClass(yw);function vl(e,t,n){if(typeof e=="number")return hi(e,t);if(e.length!==t)throw new H(`The ${n} argument must be an integer or tuple of ${t} integers. Received: ${e.length} elements.`);for(let a=0;a<t;++a){let r=e[a];if(!Q4(r))throw new H(`The ${n} argument must be an integer or tuple of ${t} integers. Received: ${JSON.stringify(e)} including a non-integer number ${r}`)}return e}function qa(e,t,n,a,r=1){if(e==null)return e;let s=t+(t-1)*(r-1),i;return n==="same"?i=e:i=e-s+1,Math.floor((i+a-1)/a)}function or(e,t,n,a){if(e==null)return null;if(a==="valid")e=e*t+cs([n-t,0]);else if(a==="same")e=e*t;else throw new H(`Unsupport padding mode: ${a}.`);return e}function bw(e,t){return O(()=>(Ot(t),t==="channelsFirst"?Ae(e,[0,2,3,1]):e))}function wN(e,t){return O(()=>(Ot(t),t==="channelsFirst"?Ae(e,[0,2,3,4,1]):e))}function OU(e,t,n,a=1,r="valid",s,i=1){return O(()=>{if(s==null&&(s=Xa()),Ot(s),e.shape.length!==3)throw new H(`The input of a conv1dWithBias operation should be 3, but is ${e.shape.length} instead.`);if(t.shape.length!==3)throw new H(`The kernel for a conv1dWithBias operation should be 3, but is ${t.shape.length} instead`);if(n!=null&&n.shape.length!==1)throw new H(`The bias for a conv1dWithBias operation should be 1, but is ${t.shape.length} instead`);if(s==="channelsFirst"&&(e=Ae(e,[0,2,1])),r==="causal")throw new Oe("The support for CAUSAL padding mode in conv1dWithBias is not implemented yet.");let o=of(e,t,a,r==="same"?"same":"valid","NWC",i);return n!=null&&(o=Za(o,n)),o})}function Ak(e,t,n,a=[1,1],r="valid",s,i,o=null){return O(()=>{if(s==null&&(s=Xa()),Ot(s),e.rank!==3&&e.rank!==4)throw new H(`conv2dWithBiasActivation expects input to be of rank 3 or 4, but received ${e.rank}.`);if(t.rank!==3&&t.rank!==4)throw new H(`conv2dWithBiasActivation expects kernel to be of rank 3 or 4, but received ${e.rank}.`);let l=bw(e,s);if(r==="causal")throw new Oe("The support for CAUSAL padding mode in conv1dWithBias is not implemented yet.");return l=ps.conv2d({x:l,filter:t,strides:a,pad:r==="same"?"same":"valid",dilations:i,dataFormat:"NHWC",bias:n,activation:o}),s==="channelsFirst"&&(l=Ae(l,[0,3,1,2])),l})}function LU(e,t,n,a=[1,1,1],r="valid",s,i){return O(()=>{if(s==null&&(s=Xa()),Ot(s),e.rank!==4&&e.rank!==5)throw new H(`conv3dWithBias expects input to be of rank 4 or 5, but received ${e.rank}.`);if(t.rank!==4&&t.rank!==5)throw new H(`conv3dWithBias expects kernel to be of rank 4 or 5, but received ${e.rank}.`);let o=wN(e,s);if(r==="causal")throw new Oe("The support for CAUSAL padding mode in conv3dWithBias is not implemented yet.");return o=uv(o,t,a,r==="same"?"same":"valid","NDHWC",i),n!=null&&(o=Za(o,n)),s==="channelsFirst"&&(o=Ae(o,[0,4,1,2,3])),o})}var xw=class extends Ye{constructor(e,t){if(super(t),this.bias=null,this.DEFAULT_KERNEL_INITIALIZER="glorotNormal",this.DEFAULT_BIAS_INITIALIZER="zeros",xw.verifyArgs(t),this.rank=e,nn(this.rank,"rank"),this.rank!==1&&this.rank!==2&&this.rank!==3)throw new Oe(`Convolution layer for rank other than 1, 2, or 3 (${this.rank}) is not implemented yet.`);if(this.kernelSize=vl(t.kernelSize,e,"kernelSize"),this.strides=vl(t.strides==null?1:t.strides,e,"strides"),this.padding=t.padding==null?"valid":t.padding,ba(this.padding),this.dataFormat=t.dataFormat==null?"channelsLast":t.dataFormat,Ot(this.dataFormat),this.activation=hs(t.activation),this.useBias=t.useBias==null?!0:t.useBias,this.biasInitializer=Nt(t.biasInitializer||this.DEFAULT_BIAS_INITIALIZER),this.biasConstraint=Yt(t.biasConstraint),this.biasRegularizer=Tt(t.biasRegularizer),this.activityRegularizer=Tt(t.activityRegularizer),this.dilationRate=vl(t.dilationRate==null?1:t.dilationRate,e,"dilationRate"),this.rank===1&&Array.isArray(this.dilationRate)&&this.dilationRate.length!==1)throw new H(`dilationRate must be a number or an array of a single number for 1D convolution, but received ${JSON.stringify(this.dilationRate)}`);if(this.rank===2){if(typeof this.dilationRate=="number")this.dilationRate=[this.dilationRate,this.dilationRate];else if(this.dilationRate.length!==2)throw new H(`dilationRate must be a number or array of two numbers for 2D convolution, but received ${JSON.stringify(this.dilationRate)}`)}else if(this.rank===3){if(typeof this.dilationRate=="number")this.dilationRate=[this.dilationRate,this.dilationRate,this.dilationRate];else if(this.dilationRate.length!==3)throw new H(`dilationRate must be a number or array of three numbers for 3D convolution, but received ${JSON.stringify(this.dilationRate)}`)}}static verifyArgs(e){if(ir("kernelSize"in e,"required key 'kernelSize' not in config"),typeof e.kernelSize!="number"&&!Uv(e.kernelSize,"number",1,3))throw new H(`BaseConv expects config.kernelSize to be number or number[] with length 1, 2, or 3, but received ${JSON.stringify(e.kernelSize)}.`)}getConfig(){let e={kernelSize:this.kernelSize,strides:this.strides,padding:this.padding,dataFormat:this.dataFormat,dilationRate:this.dilationRate,activation:ds(this.activation),useBias:this.useBias,biasInitializer:Ft(this.biasInitializer),biasRegularizer:dt(this.biasRegularizer),activityRegularizer:dt(this.activityRegularizer),biasConstraint:Xt(this.biasConstraint)},t=super.getConfig();return Object.assign(e,t),e}},rd=class extends xw{constructor(e,t){super(e,t),this.kernel=null,rd.verifyArgs(t),this.filters=t.filters,nn(this.filters,"filters"),this.kernelInitializer=Nt(t.kernelInitializer||this.DEFAULT_KERNEL_INITIALIZER),this.kernelConstraint=Yt(t.kernelConstraint),this.kernelRegularizer=Tt(t.kernelRegularizer)}build(e){e=it(e);let t=this.dataFormat==="channelsFirst"?1:e.length-1;if(e[t]==null)throw new H(`The channel dimension of the input should be defined. Found ${e[t]}`);let n=e[t],a=this.kernelSize.concat([n,this.filters]);this.kernel=this.addWeight("kernel",a,null,this.kernelInitializer,this.kernelRegularizer,!0,this.kernelConstraint),this.useBias&&(this.bias=this.addWeight("bias",[this.filters],null,this.biasInitializer,this.biasRegularizer,!0,this.biasConstraint)),this.inputSpec=[{ndim:this.rank+2,axes:{[t]:n}}],this.built=!0}call(e,t){return O(()=>{e=ze(e);let n,a=this.bias==null?null:this.bias.read(),r=E2(this.activation.getClassName());if(r!=null&&this.rank===2)n=Ak(e,this.kernel.read(),a,this.strides,this.padding,this.dataFormat,this.dilationRate,r);else{if(this.rank===1)n=OU(e,this.kernel.read(),a,this.strides[0],this.padding,this.dataFormat,this.dilationRate[0]);else if(this.rank===2)n=Ak(e,this.kernel.read(),a,this.strides,this.padding,this.dataFormat,this.dilationRate);else if(this.rank===3)n=LU(e,this.kernel.read(),a,this.strides,this.padding,this.dataFormat,this.dilationRate);else throw new Oe("convolutions greater than 3D are not implemented yet.");this.activation!=null&&(n=this.activation.apply(n))}return n})}computeOutputShape(e){e=it(e);let t=[],n=this.dataFormat==="channelsLast"?e.slice(1,e.length-1):e.slice(2);for(let r=0;r<n.length;++r){let s=qa(n[r],this.kernelSize[r],this.padding,this.strides[r],typeof this.dilationRate=="number"?this.dilationRate:this.dilationRate[r]);t.push(s)}let a=[e[0]];return this.dataFormat==="channelsLast"?(a=a.concat(t),a.push(this.filters)):(a.push(this.filters),a=a.concat(t)),a}getConfig(){let e={filters:this.filters,kernelInitializer:Ft(this.kernelInitializer),kernelRegularizer:dt(this.kernelRegularizer),kernelConstraint:Xt(this.kernelConstraint)},t=super.getConfig();return Object.assign(e,t),e}static verifyArgs(e){if(!("filters"in e)||typeof e.filters!="number"||e.filters<1)throw new H(`Convolution layer expected config.filters to be a 'number' > 0 but got ${JSON.stringify(e.filters)}`)}},sd=class extends rd{constructor(e){super(2,e),sd.verifyArgs(e)}getConfig(){let e=super.getConfig();return delete e.rank,e}static verifyArgs(e){if(typeof e.kernelSize!="number"&&!Uv(e.kernelSize,"number",1,2))throw new H(`Conv2D expects config.kernelSize to be number or number[] with length 1 or 2, but received ${JSON.stringify(e.kernelSize)}.`)}};sd.className="Conv2D";se.registerClass(sd);var id=class extends rd{constructor(e){super(3,e),id.verifyArgs(e)}getConfig(){let e=super.getConfig();return delete e.rank,e}static verifyArgs(e){if(typeof e.kernelSize!="number"&&!(Array.isArray(e.kernelSize)&&(e.kernelSize.length===1||e.kernelSize.length===3)))throw new H(`Conv3D expects config.kernelSize to be number or [number, number, number], but received ${JSON.stringify(e.kernelSize)}.`)}};id.className="Conv3D";se.registerClass(id);var vw=class extends sd{constructor(e){if(super(e),this.inputSpec=[new Bt({ndim:4})],this.padding!=="same"&&this.padding!=="valid")throw new H(`Conv2DTranspose currently supports only padding modes 'same' and 'valid', but received padding mode ${this.padding}`)}build(e){if(e=it(e),e.length!==4)throw new H("Input should have rank 4; Received input shape: "+JSON.stringify(e));let t=this.dataFormat==="channelsFirst"?1:e.length-1;if(e[t]==null)throw new H("The channel dimension of the inputs should be defined. Found `None`.");let n=e[t],a=this.kernelSize.concat([this.filters,n]);this.kernel=this.addWeight("kernel",a,"float32",this.kernelInitializer,this.kernelRegularizer,!0,this.kernelConstraint),this.useBias&&(this.bias=this.addWeight("bias",[this.filters],"float32",this.biasInitializer,this.biasRegularizer,!0,this.biasConstraint)),this.inputSpec=[new Bt({ndim:4,axes:{[t]:n}})],this.built=!0}call(e,t){return O(()=>{let n=ze(e);if(n.shape.length!==4)throw new H(`Conv2DTranspose.call() expects input tensor to be rank-4, but received a tensor of rank-${n.shape.length}`);let a=n.shape,r=a[0],s,i;this.dataFormat==="channelsFirst"?(s=2,i=3):(s=1,i=2);let o=a[s],l=a[i],u=this.kernelSize[0],p=this.kernelSize[1],d=this.strides[0],c=this.strides[1],h=or(o,d,u,this.padding),m=or(l,c,p,this.padding),f=[r,h,m,this.filters];this.dataFormat!=="channelsLast"&&(n=Ae(n,[0,2,3,1]));let g=lf(n,this.kernel.read(),f,this.strides,this.padding);return this.dataFormat!=="channelsLast"&&(g=Ae(g,[0,3,1,2])),this.bias!=null&&(g=Za(g,this.bias.read(),this.dataFormat)),this.activation!=null&&(g=this.activation.apply(g)),g})}computeOutputShape(e){e=it(e);let t=e.slice(),n,a,r;this.dataFormat==="channelsFirst"?(n=1,a=2,r=3):(n=3,a=1,r=2);let s=this.kernelSize[0],i=this.kernelSize[1],o=this.strides[0],l=this.strides[1];return t[n]=this.filters,t[a]=or(t[a],o,s,this.padding),t[r]=or(t[r],l,i,this.padding),t}getConfig(){let e=super.getConfig();return delete e.dilationRate,e}};vw.className="Conv2DTranspose";se.registerClass(vw);var ww=class extends id{constructor(e){if(super(e),this.inputSpec=[new Bt({ndim:5})],this.padding!=="same"&&this.padding!=="valid")throw new H(`Conv3DTranspose currently supports only padding modes 'same' and 'valid', but received padding mode ${this.padding}`)}build(e){if(e=it(e),e.length!==5)throw new H("Input should have rank 5; Received input shape: "+JSON.stringify(e));let t=this.dataFormat==="channelsFirst"?1:e.length-1;if(e[t]==null)throw new H("The channel dimension of the inputs should be defined. Found `None`.");let n=e[t],a=this.kernelSize.concat([this.filters,n]);this.kernel=this.addWeight("kernel",a,"float32",this.kernelInitializer,this.kernelRegularizer,!0,this.kernelConstraint),this.useBias&&(this.bias=this.addWeight("bias",[this.filters],"float32",this.biasInitializer,this.biasRegularizer,!0,this.biasConstraint)),this.inputSpec=[new Bt({ndim:5,axes:{[t]:n}})],this.built=!0}call(e,t){return O(()=>{let n=ze(e);if(n.shape.length!==5)throw new H(`Conv3DTranspose.call() expects input tensor to be rank-4, but received a tensor of rank-${n.shape.length}`);let a=n.shape,r=a[0],s,i,o;this.dataFormat==="channelsFirst"?(o=2,s=3,i=4):(o=1,s=2,i=3);let l=a[o],u=a[s],p=a[i],d=this.kernelSize[0],c=this.kernelSize[1],h=this.kernelSize[2],m=this.strides[0],f=this.strides[1],g=this.strides[2],y=or(l,m,d,this.padding),b=or(u,f,c,this.padding),x=or(p,g,h,this.padding),v=[r,y,b,x,this.filters];this.dataFormat!=="channelsLast"&&(n=Ae(n,[0,2,3,4,1]));let k=LS(n,this.kernel.read(),v,this.strides,this.padding);return this.dataFormat!=="channelsLast"&&(k=Ae(k,[0,4,1,2,3])),this.bias!==null&&(k=Za(k,this.bias.read(),this.dataFormat)),this.activation!==null&&(k=this.activation.apply(k)),k})}computeOutputShape(e){e=it(e);let t=e.slice(),n,a,r,s;this.dataFormat==="channelsFirst"?(n=1,a=2,r=3,s=4):(n=4,a=1,r=2,s=3);let i=this.kernelSize[0],o=this.kernelSize[1],l=this.kernelSize[2],u=this.strides[0],p=this.strides[1],d=this.strides[2];return t[n]=this.filters,t[a]=or(t[a],u,i,this.padding),t[r]=or(t[r],p,o,this.padding),t[s]=or(t[s],d,l,this.padding),t}getConfig(){let e=super.getConfig();return delete e.dilationRate,e}};ww.className="Conv3DTranspose";se.registerClass(ww);var kN=class extends rd{constructor(e,t){if(super(e,t),this.DEFAULT_DEPTHWISE_INITIALIZER="glorotUniform",this.DEFAULT_POINTWISE_INITIALIZER="glorotUniform",this.depthwiseKernel=null,this.pointwiseKernel=null,t.filters==null)throw new H("The `filters` configuration field is required by SeparableConv, but is unspecified.");if(t.kernelInitializer!=null||t.kernelRegularizer!=null||t.kernelConstraint!=null)throw new H("Fields kernelInitializer, kernelRegularizer and kernelConstraint are invalid for SeparableConv2D. Use depthwiseInitializer, depthwiseRegularizer, depthwiseConstraint, pointwiseInitializer, pointwiseRegularizer and pointwiseConstraint instead.");if(t.padding!=null&&t.padding!=="same"&&t.padding!=="valid")throw new H(`SeparableConv${this.rank}D supports only padding modes: 'same' and 'valid', but received ${JSON.stringify(t.padding)}`);this.depthMultiplier=t.depthMultiplier==null?1:t.depthMultiplier,this.depthwiseInitializer=Nt(t.depthwiseInitializer||this.DEFAULT_DEPTHWISE_INITIALIZER),this.depthwiseRegularizer=Tt(t.depthwiseRegularizer),this.depthwiseConstraint=Yt(t.depthwiseConstraint),this.pointwiseInitializer=Nt(t.depthwiseInitializer||this.DEFAULT_POINTWISE_INITIALIZER),this.pointwiseRegularizer=Tt(t.pointwiseRegularizer),this.pointwiseConstraint=Yt(t.pointwiseConstraint)}build(e){if(e=it(e),e.length<this.rank+2)throw new H(`Inputs to SeparableConv${this.rank}D should have rank ${this.rank+2}, but received input shape: ${JSON.stringify(e)}`);let t=this.dataFormat==="channelsFirst"?1:e.length-1;if(e[t]==null||e[t]<0)throw new H(`The channel dimension of the inputs should be defined, but found ${JSON.stringify(e[t])}`);let n=e[t],a=this.kernelSize.concat([n,this.depthMultiplier]),r=[];for(let i=0;i<this.rank;++i)r.push(1);r.push(n*this.depthMultiplier,this.filters);let s=!0;this.depthwiseKernel=this.addWeight("depthwise_kernel",a,"float32",this.depthwiseInitializer,this.depthwiseRegularizer,s,this.depthwiseConstraint),this.pointwiseKernel=this.addWeight("pointwise_kernel",r,"float32",this.pointwiseInitializer,this.pointwiseRegularizer,s,this.pointwiseConstraint),this.useBias?this.bias=this.addWeight("bias",[this.filters],"float32",this.biasInitializer,this.biasRegularizer,s,this.biasConstraint):this.bias=null,this.inputSpec=[new Bt({ndim:this.rank+2,axes:{[t]:n}})],this.built=!0}call(e,t){return O(()=>{e=ze(e);let n;if(this.rank===1)throw new Oe("1D separable convolution is not implemented yet.");return this.rank===2&&(this.dataFormat==="channelsFirst"&&(e=Ae(e,[0,2,3,1])),n=vo(e,this.depthwiseKernel.read(),this.pointwiseKernel.read(),this.strides,this.padding,this.dilationRate,"NHWC")),this.useBias&&(n=Za(n,this.bias.read(),this.dataFormat)),this.activation!=null&&(n=this.activation.apply(n)),this.dataFormat==="channelsFirst"&&(n=Ae(n,[0,3,1,2])),n})}getConfig(){let e=super.getConfig();return delete e.rank,delete e.kernelInitializer,delete e.kernelRegularizer,delete e.kernelConstraint,e.depthwiseInitializer=Ft(this.depthwiseInitializer),e.pointwiseInitializer=Ft(this.pointwiseInitializer),e.depthwiseRegularizer=dt(this.depthwiseRegularizer),e.pointwiseRegularizer=dt(this.pointwiseRegularizer),e.depthwiseConstraint=Xt(this.depthwiseConstraint),e.pointwiseConstraint=Xt(this.pointwiseConstraint),e}};kN.className="SeparableConv";var kw=class extends kN{constructor(e){super(2,e)}};kw.className="SeparableConv2D";se.registerClass(kw);var Yf=class extends rd{constructor(e){super(1,e),Yf.verifyArgs(e),this.inputSpec=[{ndim:3}]}getConfig(){let e=super.getConfig();return delete e.rank,delete e.dataFormat,e}static verifyArgs(e){if(typeof e.kernelSize!="number"&&!Uv(e.kernelSize,"number",1,1))throw new H(`Conv1D expects config.kernelSize to be number or number[] with length 1, but received ${JSON.stringify(e.kernelSize)}.`)}};Yf.className="Conv1D";se.registerClass(Yf);var Iw=class extends Ye{constructor(e){super(e),typeof e.cropping=="number"?this.cropping=[[e.cropping,e.cropping],[e.cropping,e.cropping]]:typeof e.cropping[0]=="number"?this.cropping=[[e.cropping[0],e.cropping[0]],[e.cropping[1],e.cropping[1]]]:this.cropping=e.cropping,this.dataFormat=e.dataFormat===void 0?"channelsLast":e.dataFormat,this.inputSpec=[{ndim:4}]}computeOutputShape(e){return this.dataFormat==="channelsFirst"?[e[0],e[1],e[2]-this.cropping[0][0]-this.cropping[0][1],e[3]-this.cropping[1][0]-this.cropping[1][1]]:[e[0],e[1]-this.cropping[0][0]-this.cropping[0][1],e[2]-this.cropping[1][0]-this.cropping[1][1],e[3]]}call(e,t){return O(()=>{if(e=ze(e),this.dataFormat==="channelsLast"){let n=kh(e,this.cropping[0][0],e.shape[1]-this.cropping[0][0]-this.cropping[0][1],2);return kh(n,this.cropping[1][0],e.shape[2]-this.cropping[1][1]-this.cropping[1][0],3)}else{let n=kh(e,this.cropping[0][0],e.shape[2]-this.cropping[0][0]-this.cropping[0][1],3);return kh(n,this.cropping[1][0],e.shape[3]-this.cropping[1][1]-this.cropping[1][0],4)}})}getConfig(){let e={cropping:this.cropping,dataFormat:this.dataFormat},t=super.getConfig();return Object.assign(e,t),e}};Iw.className="Cropping2D";se.registerClass(Iw);var Sw=class extends Ye{constructor(e){super(e),this.DEFAULT_SIZE=[2,2],this.inputSpec=[{ndim:4}],this.size=e.size==null?this.DEFAULT_SIZE:e.size,this.dataFormat=e.dataFormat==null?"channelsLast":e.dataFormat,Ot(this.dataFormat),this.interpolation=e.interpolation==null?"nearest":e.interpolation,K4(this.interpolation)}computeOutputShape(e){if(this.dataFormat==="channelsFirst"){let t=e[2]==null?null:this.size[0]*e[2],n=e[3]==null?null:this.size[1]*e[3];return[e[0],e[1],t,n]}else{let t=e[1]==null?null:this.size[0]*e[1],n=e[2]==null?null:this.size[1]*e[2];return[e[0],t,n,e[3]]}}call(e,t){return O(()=>{let n=ze(e),a=n.shape;if(this.dataFormat==="channelsFirst"){n=Ae(n,[0,2,3,1]);let r=this.size[0]*a[2],s=this.size[1]*a[3],i=this.interpolation==="nearest"?Ln.resizeNearestNeighbor(n,[r,s]):Ln.resizeBilinear(n,[r,s]);return Ae(i,[0,3,1,2])}else{let r=this.size[0]*a[1],s=this.size[1]*a[2];return this.interpolation==="nearest"?Ln.resizeNearestNeighbor(n,[r,s]):Ln.resizeBilinear(n,[r,s])}})}getConfig(){let e={size:this.size,dataFormat:this.dataFormat,interpolation:this.interpolation},t=super.getConfig();return Object.assign(e,t),e}};Sw.className="UpSampling2D";se.registerClass(Sw);function zU(e,t,n=[1,1],a="valid",r,s){return O(()=>{r==null&&(r=Xa()),Ot(r);let i=bw(e,r);if(e.rank!==4)throw new H(`Input for depthwiseConv2d is required to be 4-D, but is instead ${e.rank}-D`);if(t.rank!==4)throw new H(`depthwiseKernel is required to be 4-D, but is instead ${t.rank}-D`);return i=Ss(i,t,n,a==="same"?"same":"valid","NHWC",s),r==="channelsFirst"&&(i=Ae(i,[0,3,1,2])),i})}var Nw=class extends xw{constructor(e){super(2,e),this.depthwiseKernel=null,this.depthMultiplier=e.depthMultiplier==null?1:e.depthMultiplier,this.depthwiseInitializer=Nt(e.depthwiseInitializer||this.DEFAULT_KERNEL_INITIALIZER),this.depthwiseConstraint=Yt(e.depthwiseConstraint),this.depthwiseRegularizer=Tt(e.depthwiseRegularizer)}build(e){if(e=it(e),e.length<4)throw new H(`Inputs to DepthwiseConv2D should have rank 4. Received input shape: ${JSON.stringify(e)}.`);let t=this.dataFormat==="channelsFirst"?1:3;if(e[t]==null||e[t]<0)throw new H(`The channel dimension of the inputs to DepthwiseConv2D should be defined, but is not (${e[t]}).`);let n=e[t],a=[this.kernelSize[0],this.kernelSize[1],n,this.depthMultiplier];this.depthwiseKernel=this.addWeight("depthwise_kernel",a,null,this.depthwiseInitializer,this.depthwiseRegularizer,!0,this.depthwiseConstraint),this.useBias?this.bias=this.addWeight("bias",[n*this.depthMultiplier],null,this.biasInitializer,this.biasRegularizer,!0,this.biasConstraint):this.bias=null,this.built=!0}call(e,t){return O(()=>{e=ze(e);let n=zU(e,this.depthwiseKernel.read(),this.strides,this.padding,this.dataFormat,null);return this.useBias&&(n=Za(n,this.bias.read(),this.dataFormat)),this.activation!=null&&(n=this.activation.apply(n)),n})}computeOutputShape(e){e=it(e);let t=this.dataFormat==="channelsFirst"?e[2]:e[1],n=this.dataFormat==="channelsFirst"?e[3]:e[2],a=this.dataFormat==="channelsFirst"?e[1]*this.depthMultiplier:e[3]*this.depthMultiplier,r=qa(t,this.kernelSize[0],this.padding,this.strides[0]),s=qa(n,this.kernelSize[1],this.padding,this.strides[1]);return this.dataFormat==="channelsFirst"?[e[0],a,r,s]:[e[0],r,s,a]}getConfig(){let e=super.getConfig();return e.depthMultiplier=this.depthMultiplier,e.depthwiseInitializer=Ft(this.depthwiseInitializer),e.depthwiseRegularizer=dt(this.depthwiseRegularizer),e.depthwiseConstraint=Xt(this.depthwiseRegularizer),e}};Nw.className="DepthwiseConv2D";se.registerClass(Nw);function IN(e,t,n,a){if(Array.isArray(e)){if(t!=null||n!=null)throw new H("When inputs is an array, neither initialState or constants should be provided");a!=null&&(n=e.slice(e.length-a,e.length),e=e.slice(0,e.length-a)),e.length>1&&(t=e.slice(1,e.length)),e=e[0]}function r(s){return s==null||Array.isArray(s)?s:[s]}return t=r(t),n=r(n),{inputs:e,initialState:t,constants:n}}function SN(e,t,n,a=!1,r,s,i=!1,o=!1){return O(()=>{let l=t.shape.length;if(l<3)throw new H(`Input should be at least 3D, but is ${l}D.`);let u=[1,0].concat(Ka(2,l));if(t=Ae(t,u),s!=null)throw new Oe("The rnn() functoin of the deeplearn.js backend does not support constants yet.");i&&console.warn("Backend rnn(): the unroll = true option is not applicable to the imperative deeplearn.js backend."),r!=null&&(r=oe(oe(r,"bool"),"float32"),r.rank===l-1&&(r=mn(r,-1)),r=Ae(r,u)),a&&(t=na(t,0),r!=null&&(r=na(r,0)));let p=[],d,c=n,h=t.shape[0],m=ht(t),f;r!=null&&(f=ht(r));for(let y=0;y<h;++y){let b=m[y],x=O(()=>e(b,c));if(r==null)d=x[0],c=x[1];else{let v=O(()=>{let k=f[y],T=ce(ta(k),k),_=J(B(x[0],k),B(c[0],T)),E=c.map((A,M)=>J(B(x[1][M],k),B(A,T)));return{output:_,newStates:E}});d=v.output,c=v.newStates}o&&p.push(d)}let g;return o&&(g=Mt(p,1)),[d,g,c]})}var yr=class extends Ye{constructor(e){super(e);let t;if(e.cell==null)throw new H("cell property is missing for the constructor of RNN.");if(Array.isArray(e.cell)?t=new Zf({cells:e.cell}):t=e.cell,t.stateSize==null)throw new H("The RNN cell should have an attribute `stateSize` (tuple of integers, one integer per RNN state).");this.cell=t,this.returnSequences=e.returnSequences==null?!1:e.returnSequences,this.returnState=e.returnState==null?!1:e.returnState,this.goBackwards=e.goBackwards==null?!1:e.goBackwards,this._stateful=e.stateful==null?!1:e.stateful,this.unroll=e.unroll==null?!1:e.unroll,this.supportsMasking=!0,this.inputSpec=[new Bt({ndim:3})],this.stateSpec=null,this.states_=null,this.numConstants=null,this.keptStates=[]}getStates(){if(this.states_==null){let e=Array.isArray(this.cell.stateSize)?this.cell.stateSize.length:1;return Ka(0,e).map(t=>null)}else return this.states_}setStates(e){this.states_=e}computeOutputShape(e){Yb(e)&&(e=e[0]),e=e;let t=this.cell.stateSize;Array.isArray(t)||(t=[t]);let n=t[0],a;if(this.returnSequences?a=[e[0],e[1],n]:a=[e[0],n],this.returnState){let r=[];for(let s of t)r.push([e[0],s]);return[a].concat(r)}else return a}computeMask(e,t){return O(()=>{Array.isArray(t)&&(t=t[0]);let n=this.returnSequences?t:null;if(this.returnState){let a=this.states.map(r=>null);return[n].concat(a)}else return n})}get states(){if(this.states_==null){let e=Array.isArray(this.cell.stateSize)?this.cell.stateSize.length:1,t=[];for(let n=0;n<e;++n)t.push(null);return t}else return this.states_}set states(e){this.states_=e}build(e){if(this.numConstants!=null)throw new Oe("Constants support is not implemented in RNN yet.");Yb(e)&&(e=e[0]),e=e;let t=this.stateful?e[0]:null,n=e.slice(2);this.inputSpec[0]=new Bt({shape:[t,null,...n]});let a=[e[0]].concat(e.slice(2));this.cell.build(a);let r;if(Array.isArray(this.cell.stateSize)?r=this.cell.stateSize:r=[this.cell.stateSize],this.stateSpec!=null){if(!w.arraysEqual(this.stateSpec.map(s=>s.shape[s.shape.length-1]),r))throw new H(`An initialState was passed that is not compatible with cell.stateSize. Received stateSpec=${this.stateSpec}; However cell.stateSize is ${this.cell.stateSize}`)}else this.stateSpec=r.map(s=>new Bt({shape:[null,s]}));this.stateful&&this.resetStates()}resetStates(e,t=!1){O(()=>{if(!this.stateful)throw new Ir("Cannot call resetStates() on an RNN Layer that is not stateful.");let n=this.inputSpec[0].shape[0];if(n==null)throw new H("If an RNN is stateful, it needs to know its batch size. Specify the batch size of your input tensors: \n- If using a Sequential model, specify the batch size by passing a `batchInputShape` option to your first layer.\n- If using the functional API, specify the batch size by passing a `batchShape` option to your Input layer.");if(this.states_==null)Array.isArray(this.cell.stateSize)?this.states_=this.cell.stateSize.map(a=>St([n,a])):this.states_=[St([n,this.cell.stateSize])];else if(e==null)Re(this.states_),this.keptStates!=null&&(Re(this.keptStates),this.keptStates=[]),Array.isArray(this.cell.stateSize)?this.states_=this.cell.stateSize.map(a=>St([n,a])):this.states_[0]=St([n,this.cell.stateSize]);else{if(Array.isArray(e)||(e=[e]),e.length!==this.states_.length)throw new H(`Layer ${this.name} expects ${this.states_.length} state(s), but it received ${e.length} state value(s). Input received: ${e}`);t===!0?this.keptStates.push(this.states_.slice()):Re(this.states_);for(let a=0;a<this.states_.length;++a){let r=e[a],s=Array.isArray(this.cell.stateSize)?this.cell.stateSize[a]:this.cell.stateSize,i=[n,s];if(!w.arraysEqual(r.shape,i))throw new H(`State ${a} is incompatible with layer ${this.name}: expected shape=${i}, received shape=${r.shape}`);this.states_[a]=r}}this.states_=this.states_.map(a=>tn(a.clone()))})}apply(e,t){let n=t==null?null:t.initialState,a=t==null?null:t.constants;t==null&&(t={});let r=IN(e,n,a,this.numConstants);e=r.inputs,n=r.initialState,a=r.constants;let s=[],i=[];if(n!=null){t.initialState=n,s=s.concat(n),this.stateSpec=[];for(let o of n)this.stateSpec.push(new Bt({shape:o.shape}));i=i.concat(this.stateSpec)}if(a!=null&&(t.constants=a,s=s.concat(a),this.numConstants=a.length),s[0]instanceof Ua){let o=[e].concat(s),l=this.inputSpec.concat(i),u=this.inputSpec;this.inputSpec=l;let p=super.apply(o,t);return this.inputSpec=u,p}else return super.apply(e,t)}call(e,t){return O(()=>{let n=t==null?null:t.mask,a=t==null?null:t.training,r=t==null?null:t.initialState;e=ze(e),r==null&&(this.stateful?r=this.states_:r=this.getInitialState(e));let s=Array.isArray(this.cell.stateSize)?this.cell.stateSize.length:1;if(r.length!==s)throw new H(`RNN Layer has ${s} state(s) but was passed ${r.length} initial state(s).`);this.unroll&&console.warn("Ignoring unroll = true for RNN layer, due to imperative backend.");let i={training:a},o=SN((c,h)=>{let m=this.cell.call([c].concat(h),i);return[m[0],m.slice(1)]},e,r,this.goBackwards,n,null,this.unroll,this.returnSequences),l=o[0],u=o[1],p=o[2];this.stateful&&this.resetStates(p,a);let d=this.returnSequences?u:l;return this.returnState?[d].concat(p):d})}getInitialState(e){return O(()=>{let t=St(e.shape);return t=be(t,[1,2]),t=Zc(t),Array.isArray(this.cell.stateSize)?this.cell.stateSize.map(n=>n>1?Kb(t,[1,n]):t):this.cell.stateSize>1?[Kb(t,[1,this.cell.stateSize])]:[t]})}get trainableWeights(){return this.trainable?this.cell.trainableWeights:[]}get nonTrainableWeights(){return this.trainable?this.cell.nonTrainableWeights:this.cell.weights}setFastWeightInitDuringBuild(e){super.setFastWeightInitDuringBuild(e),this.cell!=null&&this.cell.setFastWeightInitDuringBuild(e)}getConfig(){let e=super.getConfig(),t={returnSequences:this.returnSequences,returnState:this.returnState,goBackwards:this.goBackwards,stateful:this.stateful,unroll:this.unroll};this.numConstants!=null&&(t.numConstants=this.numConstants);let n=this.cell.getConfig();return this.getClassName()===yr.className&&(t.cell={className:this.cell.getClassName(),config:n}),Object.assign({},n,e,t)}static fromConfig(e,t,n={}){let a=t.cell,r=ja(a,n);return new e(Object.assign(t,{cell:r}))}};yr.className="RNN";se.registerClass(yr);var od=class extends Ye{},Qf=class extends od{constructor(e){super(e),this.DEFAULT_ACTIVATION="tanh",this.DEFAULT_KERNEL_INITIALIZER="glorotNormal",this.DEFAULT_RECURRENT_INITIALIZER="orthogonal",this.DEFAULT_BIAS_INITIALIZER="zeros",this.units=e.units,nn(this.units,"units"),this.activation=hs(e.activation==null?this.DEFAULT_ACTIVATION:e.activation),this.useBias=e.useBias==null?!0:e.useBias,this.kernelInitializer=Nt(e.kernelInitializer||this.DEFAULT_KERNEL_INITIALIZER),this.recurrentInitializer=Nt(e.recurrentInitializer||this.DEFAULT_RECURRENT_INITIALIZER),this.biasInitializer=Nt(e.biasInitializer||this.DEFAULT_BIAS_INITIALIZER),this.kernelRegularizer=Tt(e.kernelRegularizer),this.recurrentRegularizer=Tt(e.recurrentRegularizer),this.biasRegularizer=Tt(e.biasRegularizer),this.kernelConstraint=Yt(e.kernelConstraint),this.recurrentConstraint=Yt(e.recurrentConstraint),this.biasConstraint=Yt(e.biasConstraint),this.dropout=Cl([1,cs([0,e.dropout==null?0:e.dropout])]),this.recurrentDropout=Cl([1,cs([0,e.recurrentDropout==null?0:e.recurrentDropout])]),this.dropoutFunc=e.dropoutFunc,this.stateSize=this.units,this.dropoutMask=null,this.recurrentDropoutMask=null}build(e){e=it(e),this.kernel=this.addWeight("kernel",[e[e.length-1],this.units],null,this.kernelInitializer,this.kernelRegularizer,!0,this.kernelConstraint),this.recurrentKernel=this.addWeight("recurrent_kernel",[this.units,this.units],null,this.recurrentInitializer,this.recurrentRegularizer,!0,this.recurrentConstraint),this.useBias?this.bias=this.addWeight("bias",[this.units],null,this.biasInitializer,this.biasRegularizer,!0,this.biasConstraint):this.bias=null,this.built=!0}call(e,t){return O(()=>{if(e=e,e.length!==2)throw new H(`SimpleRNNCell expects 2 input Tensors, got ${e.length}.`);let n=e[1];e=e[0];let a=t.training==null?!1:t.training;0<this.dropout&&this.dropout<1&&this.dropoutMask==null&&(this.dropoutMask=ms({ones:()=>ta(e),rate:this.dropout,training:a,dropoutFunc:this.dropoutFunc})),0<this.recurrentDropout&&this.recurrentDropout<1&&this.recurrentDropoutMask==null&&(this.recurrentDropoutMask=ms({ones:()=>ta(n),rate:this.recurrentDropout,training:a,dropoutFunc:this.dropoutFunc}));let r,s=this.dropoutMask,i=this.recurrentDropoutMask;s!=null?r=ur(B(e,s),this.kernel.read()):r=ur(e,this.kernel.read()),this.bias!=null&&(r=Za(r,this.bias.read())),i!=null&&(n=B(n,i));let o=J(r,ur(n,this.recurrentKernel.read()));return this.activation!=null&&(o=this.activation.apply(o)),[o,o]})}getConfig(){let e=super.getConfig(),t={units:this.units,activation:ds(this.activation),useBias:this.useBias,kernelInitializer:Ft(this.kernelInitializer),recurrentInitializer:Ft(this.recurrentInitializer),biasInitializer:Ft(this.biasInitializer),kernelRegularizer:dt(this.kernelRegularizer),recurrentRegularizer:dt(this.recurrentRegularizer),biasRegularizer:dt(this.biasRegularizer),activityRegularizer:dt(this.activityRegularizer),kernelConstraint:Xt(this.kernelConstraint),recurrentConstraint:Xt(this.recurrentConstraint),biasConstraint:Xt(this.biasConstraint),dropout:this.dropout,recurrentDropout:this.recurrentDropout};return Object.assign({},e,t)}};Qf.className="SimpleRNNCell";se.registerClass(Qf);var Tw=class extends yr{constructor(e){e.cell=new Qf(e),super(e)}call(e,t){return O(()=>{this.cell.dropoutMask!=null&&(Re(this.cell.dropoutMask),this.cell.dropoutMask=null),this.cell.recurrentDropoutMask!=null&&(Re(this.cell.recurrentDropoutMask),this.cell.recurrentDropoutMask=null);let n=t==null?null:t.mask,a=t==null?null:t.training,r=t==null?null:t.initialState;return super.call(e,{mask:n,training:a,initialState:r})})}static fromConfig(e,t){return new e(t)}};Tw.className="SimpleRNN";se.registerClass(Tw);var Jf=class extends od{constructor(e){if(super(e),this.DEFAULT_ACTIVATION="tanh",this.DEFAULT_RECURRENT_ACTIVATION="hardSigmoid",this.DEFAULT_KERNEL_INITIALIZER="glorotNormal",this.DEFAULT_RECURRENT_INITIALIZER="orthogonal",this.DEFAULT_BIAS_INITIALIZER="zeros",e.resetAfter)throw new H("GRUCell does not support reset_after parameter set to true.");this.units=e.units,nn(this.units,"units"),this.activation=hs(e.activation===void 0?this.DEFAULT_ACTIVATION:e.activation),this.recurrentActivation=hs(e.recurrentActivation===void 0?this.DEFAULT_RECURRENT_ACTIVATION:e.recurrentActivation),this.useBias=e.useBias==null?!0:e.useBias,this.kernelInitializer=Nt(e.kernelInitializer||this.DEFAULT_KERNEL_INITIALIZER),this.recurrentInitializer=Nt(e.recurrentInitializer||this.DEFAULT_RECURRENT_INITIALIZER),this.biasInitializer=Nt(e.biasInitializer||this.DEFAULT_BIAS_INITIALIZER),this.kernelRegularizer=Tt(e.kernelRegularizer),this.recurrentRegularizer=Tt(e.recurrentRegularizer),this.biasRegularizer=Tt(e.biasRegularizer),this.kernelConstraint=Yt(e.kernelConstraint),this.recurrentConstraint=Yt(e.recurrentConstraint),this.biasConstraint=Yt(e.biasConstraint),this.dropout=Cl([1,cs([0,e.dropout==null?0:e.dropout])]),this.recurrentDropout=Cl([1,cs([0,e.recurrentDropout==null?0:e.recurrentDropout])]),this.dropoutFunc=e.dropoutFunc,this.implementation=e.implementation,this.stateSize=this.units,this.dropoutMask=null,this.recurrentDropoutMask=null}build(e){e=it(e);let t=e[e.length-1];this.kernel=this.addWeight("kernel",[t,this.units*3],null,this.kernelInitializer,this.kernelRegularizer,!0,this.kernelConstraint),this.recurrentKernel=this.addWeight("recurrent_kernel",[this.units,this.units*3],null,this.recurrentInitializer,this.recurrentRegularizer,!0,this.recurrentConstraint),this.useBias?this.bias=this.addWeight("bias",[this.units*3],null,this.biasInitializer,this.biasRegularizer,!0,this.biasConstraint):this.bias=null,this.built=!0}call(e,t){return O(()=>{if(e=e,e.length!==2)throw new H(`GRUCell expects 2 input Tensors (inputs, h, c), got ${e.length}.`);let n=t.training==null?!1:t.training,a=e[1];e=e[0],0<this.dropout&&this.dropout<1&&this.dropoutMask==null&&(this.dropoutMask=ms({ones:()=>ta(e),rate:this.dropout,training:n,count:3,dropoutFunc:this.dropoutFunc})),0<this.recurrentDropout&&this.recurrentDropout<1&&this.recurrentDropoutMask==null&&(this.recurrentDropoutMask=ms({ones:()=>ta(a),rate:this.recurrentDropout,training:n,count:3,dropoutFunc:this.dropoutFunc}));let r=this.dropoutMask,s=this.recurrentDropoutMask,i,o,l;0<this.dropout&&this.dropout<1&&(e=B(e,r[0]));let u=ur(e,this.kernel.read());this.useBias&&(u=Za(u,this.bias.read())),0<this.recurrentDropout&&this.recurrentDropout<1&&(a=B(a,s[0]));let p=this.recurrentKernel.read(),[d,c]=zn(p,[2*this.units,this.units],p.rank-1),h=ur(a,d),[m,f,g]=zn(u,3,u.rank-1),[y,b]=zn(h,2,h.rank-1);i=this.recurrentActivation.apply(J(m,y)),o=this.recurrentActivation.apply(J(f,b));let x=ur(B(o,a),c);l=this.activation.apply(J(g,x));let v=J(B(i,a),B(J(1,It(i)),l));return[v,v]})}getConfig(){let e=super.getConfig(),t={units:this.units,activation:ds(this.activation),recurrentActivation:ds(this.recurrentActivation),useBias:this.useBias,kernelInitializer:Ft(this.kernelInitializer),recurrentInitializer:Ft(this.recurrentInitializer),biasInitializer:Ft(this.biasInitializer),kernelRegularizer:dt(this.kernelRegularizer),recurrentRegularizer:dt(this.recurrentRegularizer),biasRegularizer:dt(this.biasRegularizer),activityRegularizer:dt(this.activityRegularizer),kernelConstraint:Xt(this.kernelConstraint),recurrentConstraint:Xt(this.recurrentConstraint),biasConstraint:Xt(this.biasConstraint),dropout:this.dropout,recurrentDropout:this.recurrentDropout,implementation:this.implementation,resetAfter:!1};return Object.assign({},e,t)}};Jf.className="GRUCell";se.registerClass(Jf);var _w=class extends yr{constructor(e){e.implementation===0&&console.warn("`implementation=0` has been deprecated, and now defaults to `implementation=1`. Please update your layer call."),e.cell=new Jf(e),super(e)}call(e,t){return O(()=>{this.cell.dropoutMask!=null&&(Re(this.cell.dropoutMask),this.cell.dropoutMask=null),this.cell.recurrentDropoutMask!=null&&(Re(this.cell.recurrentDropoutMask),this.cell.recurrentDropoutMask=null);let n=t==null?null:t.mask,a=t==null?null:t.training,r=t==null?null:t.initialState;return super.call(e,{mask:n,training:a,initialState:r})})}static fromConfig(e,t){return t.implmentation===0&&(t.implementation=1),new e(t)}};_w.className="GRU";se.registerClass(_w);var ld=class extends od{constructor(e){super(e),this.DEFAULT_ACTIVATION="tanh",this.DEFAULT_RECURRENT_ACTIVATION="hardSigmoid",this.DEFAULT_KERNEL_INITIALIZER="glorotNormal",this.DEFAULT_RECURRENT_INITIALIZER="orthogonal",this.DEFAULT_BIAS_INITIALIZER="zeros",this.units=e.units,nn(this.units,"units"),this.activation=hs(e.activation===void 0?this.DEFAULT_ACTIVATION:e.activation),this.recurrentActivation=hs(e.recurrentActivation===void 0?this.DEFAULT_RECURRENT_ACTIVATION:e.recurrentActivation),this.useBias=e.useBias==null?!0:e.useBias,this.kernelInitializer=Nt(e.kernelInitializer||this.DEFAULT_KERNEL_INITIALIZER),this.recurrentInitializer=Nt(e.recurrentInitializer||this.DEFAULT_RECURRENT_INITIALIZER),this.biasInitializer=Nt(e.biasInitializer||this.DEFAULT_BIAS_INITIALIZER),this.unitForgetBias=e.unitForgetBias,this.kernelRegularizer=Tt(e.kernelRegularizer),this.recurrentRegularizer=Tt(e.recurrentRegularizer),this.biasRegularizer=Tt(e.biasRegularizer),this.kernelConstraint=Yt(e.kernelConstraint),this.recurrentConstraint=Yt(e.recurrentConstraint),this.biasConstraint=Yt(e.biasConstraint),this.dropout=Cl([1,cs([0,e.dropout==null?0:e.dropout])]),this.recurrentDropout=Cl([1,cs([0,e.recurrentDropout==null?0:e.recurrentDropout])]),this.dropoutFunc=e.dropoutFunc,this.implementation=e.implementation,this.stateSize=[this.units,this.units],this.dropoutMask=null,this.recurrentDropoutMask=null}build(e){var t;e=it(e);let n=e[e.length-1];this.kernel=this.addWeight("kernel",[n,this.units*4],null,this.kernelInitializer,this.kernelRegularizer,!0,this.kernelConstraint),this.recurrentKernel=this.addWeight("recurrent_kernel",[this.units,this.units*4],null,this.recurrentInitializer,this.recurrentRegularizer,!0,this.recurrentConstraint);let a;if(this.useBias){if(this.unitForgetBias){let r=this.biasInitializer,s=this.units;a=new(t=class extends Fa{apply(i,o){let l=r.apply([s]),u=new Bf().apply([s]),p=r.apply([s*2]);return fk(fk(l,u),p)}},t.className="CustomInit",t)}else a=this.biasInitializer;this.bias=this.addWeight("bias",[this.units*4],null,a,this.biasRegularizer,!0,this.biasConstraint)}else this.bias=null;this.built=!0}call(e,t){return O(()=>{let n=t.training==null?!1:t.training;if(e=e,e.length!==3)throw new H(`LSTMCell expects 3 input Tensors (inputs, h, c), got ${e.length}.`);let a=e[1],r=e[2];e=e[0],0<this.dropout&&this.dropout<1&&this.dropoutMask==null&&(this.dropoutMask=ms({ones:()=>ta(e),rate:this.dropout,training:n,count:4,dropoutFunc:this.dropoutFunc})),0<this.recurrentDropout&&this.recurrentDropout<1&&this.recurrentDropoutMask==null&&(this.recurrentDropoutMask=ms({ones:()=>ta(a),rate:this.recurrentDropout,training:n,count:4,dropoutFunc:this.dropoutFunc}));let s=this.dropoutMask,i=this.recurrentDropoutMask,o,l,u,p;0<this.dropout&&this.dropout<1&&(e=B(e,s[0]));let d=ur(e,this.kernel.read());0<this.recurrentDropout&&this.recurrentDropout<1&&(a=B(a,i[0])),d=J(d,ur(a,this.recurrentKernel.read())),this.useBias&&(d=Za(d,this.bias.read()));let[c,h,m,f]=zn(d,4,d.rank-1);o=this.recurrentActivation.apply(c),l=this.recurrentActivation.apply(h),u=J(B(l,r),B(o,this.activation.apply(m))),p=this.recurrentActivation.apply(f);let g=B(p,this.activation.apply(u));return[g,g,u]})}getConfig(){let e=super.getConfig(),t={units:this.units,activation:ds(this.activation),recurrentActivation:ds(this.recurrentActivation),useBias:this.useBias,kernelInitializer:Ft(this.kernelInitializer),recurrentInitializer:Ft(this.recurrentInitializer),biasInitializer:Ft(this.biasInitializer),unitForgetBias:this.unitForgetBias,kernelRegularizer:dt(this.kernelRegularizer),recurrentRegularizer:dt(this.recurrentRegularizer),biasRegularizer:dt(this.biasRegularizer),activityRegularizer:dt(this.activityRegularizer),kernelConstraint:Xt(this.kernelConstraint),recurrentConstraint:Xt(this.recurrentConstraint),biasConstraint:Xt(this.biasConstraint),dropout:this.dropout,recurrentDropout:this.recurrentDropout,implementation:this.implementation};return Object.assign({},e,t)}};ld.className="LSTMCell";se.registerClass(ld);var Cw=class extends yr{constructor(e){e.implementation===0&&console.warn("`implementation=0` has been deprecated, and now defaults to `implementation=1`. Please update your layer call."),e.cell=new ld(e),super(e)}call(e,t){return O(()=>{this.cell.dropoutMask!=null&&(Re(this.cell.dropoutMask),this.cell.dropoutMask=null),this.cell.recurrentDropoutMask!=null&&(Re(this.cell.recurrentDropoutMask),this.cell.recurrentDropoutMask=null);let n=t==null?null:t.mask,a=t==null?null:t.training,r=t==null?null:t.initialState;return super.call(e,{mask:n,training:a,initialState:r})})}static fromConfig(e,t){return t.implmentation===0&&(t.implementation=1),new e(t)}};Cw.className="LSTM";se.registerClass(Cw);var Zf=class extends od{constructor(e){super(e),this.cells=e.cells}get stateSize(){let e=[];for(let t of this.cells.slice().reverse())Array.isArray(t.stateSize)?e.push(...t.stateSize):e.push(t.stateSize);return e}call(e,t){return O(()=>{e=e;let n=e.slice(1),a=[];for(let i of this.cells.slice().reverse())Array.isArray(i.stateSize)?a.push(n.splice(0,i.stateSize.length)):a.push(n.splice(0,1));a.reverse();let r=[],s;for(let i=0;i<this.cells.length;++i){let o=this.cells[i];n=a[i],i===0?s=[e[0]].concat(n):s=[s[0]].concat(n),s=o.call(s,t),r.push(s.slice(1))}n=[];for(let i of r.slice().reverse())n.push(...i);return[s[0]].concat(n)})}build(e){Yb(e)&&(e=e[0]),e=e;let t;this.cells.forEach((n,a)=>{ei(`RNNCell_${a}`,()=>{n.build(e),Array.isArray(n.stateSize)?t=n.stateSize[0]:t=n.stateSize,e=[e[0],t]})}),this.built=!0}getConfig(){let e=super.getConfig(),t=a=>({className:a.getClassName(),config:a.getConfig()}),n={cells:this.cells.map(t)};return Object.assign({},e,n)}static fromConfig(e,t,n={}){let a=[];for(let r of t.cells)a.push(ja(r,n));return new e({cells:a})}get trainableWeights(){if(!this.trainable)return[];let e=[];for(let t of this.cells)e.push(...t.trainableWeights);return e}get nonTrainableWeights(){let e=[];for(let t of this.cells)e.push(...t.nonTrainableWeights);if(!this.trainable){let t=[];for(let n of this.cells)t.push(...n.trainableWeights);return t.concat(e)}return e}getWeights(){let e=[];for(let t of this.cells)e.push(...t.weights);return Qb(e)}setWeights(e){let t=[];for(let n of this.cells){let a=n.weights.length,r=e.splice(a);for(let s=0;s<n.weights.length;++s)t.push([n.weights[s],r[s]])}Jv(t)}};Zf.className="StackedRNNCells";se.registerClass(Zf);function ms(e){let{ones:t,rate:n,training:a=!1,count:r=1,dropoutFunc:s}=e,i=()=>s!=null?s(t(),n):P2(t(),n),o=()=>td(i,t,a);return!r||r<=1?tn(o().clone()):Array(r).fill(void 0).map(o).map(l=>tn(l.clone()))}var BU=function(e,t){var n={};for(var a in e)Object.prototype.hasOwnProperty.call(e,a)&&t.indexOf(a)<0&&(n[a]=e[a]);if(e!=null&&typeof Object.getOwnPropertySymbols=="function")for(var r=0,a=Object.getOwnPropertySymbols(e);r<a.length;r++)t.indexOf(a[r])<0&&Object.prototype.propertyIsEnumerable.call(e,a[r])&&(n[a[r]]=e[a[r]]);return n},NN=class extends yr{constructor(e){if(e.unroll)throw new Oe("Unrolling is not possible with convolutional RNNs.");if(Array.isArray(e.cell))throw new Oe("It is not possible at the moment to stack convolutional cells.");super(e),this.inputSpec=[new Bt({ndim:5})]}call(e,t){return O(()=>{if(this.cell.dropoutMask!=null&&(Re(this.cell.dropoutMask),this.cell.dropoutMask=null),this.cell.recurrentDropoutMask!=null&&(Re(this.cell.recurrentDropoutMask),this.cell.recurrentDropoutMask=null),t&&t.constants)throw new H("ConvRNN2D cell does not support constants");let n=t==null?null:t.mask,a=t==null?null:t.training,r=t==null?null:t.initialState;return super.call(e,{mask:n,training:a,initialState:r})})}computeOutputShape(e){let t=this.computeSingleOutputShape(e);return this.returnSequences||(t=[t[0],...t.slice(2)]),this.returnState&&(t=[t,...Array(2).fill([e[0],...t.slice(-3)])]),t}getInitialState(e){return O(()=>{let{stateSize:t}=this.cell,n=e.shape,a=this.computeSingleOutputShape(n),r=[a[0],...a.slice(2)],s=St(r);return Array.isArray(t)?Array(t.length).fill(s):[s]})}resetStates(e,t=!1){O(()=>{if(!this.stateful)throw new Ir("Cannot call resetStates() on an RNN Layer that is not stateful.");let n=this.inputSpec[0].shape,a=this.computeSingleOutputShape(n),r=[a[0],...a.slice(2)];if(n[0]==null)throw new H("If an RNN is stateful, it needs to know its batch size. Specify the batch size of your input tensors: \n- If using a Sequential model, specify the batch size by passing a `batchInputShape` option to your first layer.\n- If using the functional API, specify the batch size by passing a `batchShape` option to your Input layer.");if(this.getStates()==null)Array.isArray(this.cell.stateSize)?this.states_=this.cell.stateSize.map(()=>St(r)):this.states_=[St(r)];else if(e==null)Re(this.states_),this.keptStates!=null&&(Re(this.keptStates),this.keptStates=[]),Array.isArray(this.cell.stateSize)?this.states_=this.cell.stateSize.map(()=>St(r)):this.states_[0]=St(r);else{if(Array.isArray(e)||(e=[e]),e.length!==this.states_.length)throw new H(`Layer ${this.name} expects ${this.states_.length} state(s), but it received ${e.length} state value(s). Input received: ${e}`);t?this.keptStates.push(this.states_.slice()):Re(this.states_);for(let s=0;s<this.states_.length;++s){let i=e[s],o=r;if(!w.arraysEqual(i.shape,o))throw new H(`State ${s} is incompatible with layer ${this.name}: expected shape=${o}, received shape=${i.shape}`);this.states_[s]=i}}this.states_=this.states_.map(s=>tn(s.clone()))})}computeSingleOutputShape(e){let{dataFormat:t,filters:n,kernelSize:a,padding:r,strides:s,dilationRate:i}=this.cell,o=t==="channelsFirst",l=e[o?3:2],u=e[o?4:3],p=qa(l,a[0],r,s[0],i[0]),d=qa(u,a[1],r,s[1],i[1]);return[...e.slice(0,2),...o?[n,p,d]:[p,d,n]]}};NN.className="ConvRNN2D";var eg=class extends ld{constructor(e){let{filters:t,kernelSize:n,strides:a,padding:r,dataFormat:s,dilationRate:i}=e;super(Object.assign({},e,{units:t})),this.filters=t,nn(this.filters,"filters"),this.kernelSize=vl(n,2,"kernelSize"),this.kernelSize.forEach(o=>nn(o,"kernelSize")),this.strides=vl(a||1,2,"strides"),this.strides.forEach(o=>nn(o,"strides")),this.padding=r||"valid",ba(this.padding),this.dataFormat=s||"channelsLast",Ot(this.dataFormat),this.dilationRate=vl(i||1,2,"dilationRate"),this.dilationRate.forEach(o=>nn(o,"dilationRate"))}build(e){var t;e=it(e);let n=this.dataFormat==="channelsFirst"?1:e.length-1;if(e[n]==null)throw new H(`The channel dimension of the input should be defined. Found ${e[n]}`);let a=e[n],r=4,s=this.kernelSize.concat([a,this.filters*r]);this.kernel=this.addWeight("kernel",s,null,this.kernelInitializer,this.kernelRegularizer,!0,this.kernelConstraint);let i=this.kernelSize.concat([this.filters,this.filters*r]);if(this.recurrentKernel=this.addWeight("recurrent_kernel",i,null,this.recurrentInitializer,this.recurrentRegularizer,!0,this.recurrentConstraint),this.useBias){let o;if(this.unitForgetBias){let l=this.biasInitializer,u=this.filters;o=new(t=class extends Fa{apply(p,d){let c=l.apply([u]),h=Qn([u]),m=l.apply([u*2]);return Gv([c,h,m])}},t.className="CustomInit",t)}else o=this.biasInitializer;this.bias=this.addWeight("bias",[this.filters*r],null,o,this.biasRegularizer,!0,this.biasConstraint)}this.built=!0}call(e,t){return O(()=>{if(e.length!==3)throw new H(`ConvLSTM2DCell expects 3 input Tensors (inputs, h, c), got ${e.length}.`);let n=t.training||!1,a=e[0],r=e[1],s=e[2],i=4;0<this.dropout&&this.dropout<1&&this.dropoutMask==null&&(this.dropoutMask=ms({ones:()=>ta(a),rate:this.dropout,training:n,count:i,dropoutFunc:this.dropoutFunc}));let o=this.dropoutMask,l=(ee,re,Y)=>!re||!re[Y]?ee:B(re[Y],ee),u=l(a,o,0),p=l(a,o,1),d=l(a,o,2),c=l(a,o,3);0<this.recurrentDropout&&this.recurrentDropout<1&&this.recurrentDropoutMask==null&&(this.recurrentDropoutMask=ms({ones:()=>ta(r),rate:this.recurrentDropout,training:n,count:i,dropoutFunc:this.dropoutFunc}));let h=this.recurrentDropoutMask,m=l(r,h,0),f=l(r,h,1),g=l(r,h,2),y=l(r,h,3),b=3,[x,v,k,T]=zn(this.kernel.read(),i,b),[_,E,A,M]=this.useBias?zn(this.bias.read(),i):[null,null,null,null];u=this.inputConv(u,x,_,this.padding),p=this.inputConv(p,v,E,this.padding),d=this.inputConv(d,k,A,this.padding),c=this.inputConv(c,T,M,this.padding);let[$,S,P,V]=zn(this.recurrentKernel.read(),i,b);m=this.recurrentConv(m,$),f=this.recurrentConv(f,S),g=this.recurrentConv(g,P),y=this.recurrentConv(y,V);let j=this.recurrentActivation.apply(J(u,m)),q=this.recurrentActivation.apply(J(p,f)),K=J(B(q,s),B(j,this.activation.apply(J(d,g)))),Z=B(this.recurrentActivation.apply(J(c,y)),this.activation.apply(K));return[Z,Z,K]})}getConfig(){let e=super.getConfig(),{units:t}=e,n=BU(e,["units"]),a={filters:this.filters,kernelSize:this.kernelSize,padding:this.padding,dataFormat:this.dataFormat,dilationRate:this.dilationRate,strides:this.strides};return Object.assign({},n,a)}inputConv(e,t,n,a){let r=Rt(e,t,this.strides,a||"valid",this.dataFormat==="channelsFirst"?"NCHW":"NHWC",this.dilationRate);return n?Za(r,n,this.dataFormat):r}recurrentConv(e,t){return Rt(e,t,1,"same",this.dataFormat==="channelsFirst"?"NCHW":"NHWC")}};eg.className="ConvLSTM2DCell";se.registerClass(eg);var Ew=class extends NN{constructor(e){let t=new eg(e);super(Object.assign({},e,{cell:t}))}static fromConfig(e,t){return new e(t)}};Ew.className="ConvLSTM2D";se.registerClass(Ew);var tg=class extends Ye{constructor(e){super(e),this.rate=Math.max(Math.min(e.rate,1),0),this.noiseShape=e.noiseShape,this.seed=e.seed,this.supportsMasking=!0}getNoiseShape(e){if(this.noiseShape==null)return this.noiseShape;let t=e.shape,n=[];for(let a=0;a<this.noiseShape.length;++a)n.push(this.noiseShape[a]==null?t[a]:this.noiseShape[a]);return n}call(e,t){return O(()=>{this.invokeCallHook(e,t);let n=ze(e);if(0<this.rate&&this.rate<1){let a=t.training==null?!1:t.training,r=this.getNoiseShape(n);return td(()=>P2(n,this.rate,r,this.seed),()=>n,a)}return e})}getConfig(){let e={rate:this.rate,noiseShape:this.noiseShape,seed:this.seed},t=super.getConfig();return Object.assign(e,t),e}dispose(){return super.dispose()}};tg.className="Dropout";se.registerClass(tg);var Fw=class extends tg{constructor(e){super(e),this.inputSpec=[{ndim:3}]}getNoiseShape(e){let t=e.shape;return[t[0],1,t[2]]}};Fw.className="SpatialDropout1D";se.registerClass(Fw);var Aw=class extends Ye{constructor(e){if(super(e),this.activation=null,this.useBias=!0,this.kernel=null,this.bias=null,this.DEFAULT_KERNEL_INITIALIZER="glorotNormal",this.DEFAULT_BIAS_INITIALIZER="zeros",e.batchInputShape==null&&e.inputShape==null&&e.inputDim!=null){let t=null;e.batchSize!=null&&(t=e.batchSize),this.batchInputShape=[t,e.inputDim]}this.units=e.units,nn(this.units,"units"),this.activation=hs(e.activation),e.useBias!=null&&(this.useBias=e.useBias),this.kernelInitializer=Nt(e.kernelInitializer||this.DEFAULT_KERNEL_INITIALIZER),this.biasInitializer=Nt(e.biasInitializer||this.DEFAULT_BIAS_INITIALIZER),this.kernelConstraint=Yt(e.kernelConstraint),this.biasConstraint=Yt(e.biasConstraint),this.kernelRegularizer=Tt(e.kernelRegularizer),this.biasRegularizer=Tt(e.biasRegularizer),this.activityRegularizer=Tt(e.activityRegularizer),this.supportsMasking=!0,this.inputSpec=[{minNDim:2}]}build(e){e=it(e);let t=e[e.length-1];this.kernel==null&&(this.kernel=this.addWeight("kernel",[t,this.units],null,this.kernelInitializer,this.kernelRegularizer,!0,this.kernelConstraint),this.useBias&&(this.bias=this.addWeight("bias",[this.units],null,this.biasInitializer,this.biasRegularizer,!0,this.biasConstraint))),this.inputSpec=[{minNDim:2,axes:{[-1]:t}}],this.built=!0}computeOutputShape(e){e=it(e);let t=e.slice();return t[t.length-1]=this.units,t}call(e,t){return O(()=>{this.invokeCallHook(e,t);let n=ze(e),a=E2(this.activation.getClassName()),r;return a!=null?r=ur(n,this.kernel.read(),a,this.bias?this.bias.read():null):(r=ur(n,this.kernel.read()),this.bias!=null&&(r=Za(r,this.bias.read())),this.activation!=null&&(r=this.activation.apply(r))),r})}getConfig(){let e={units:this.units,activation:ds(this.activation),useBias:this.useBias,kernelInitializer:Ft(this.kernelInitializer),biasInitializer:Ft(this.biasInitializer),kernelRegularizer:dt(this.kernelRegularizer),biasRegularizer:dt(this.biasRegularizer),activityRegularizer:dt(this.activityRegularizer),kernelConstraint:Xt(this.kernelConstraint),biasConstraint:Xt(this.biasConstraint)},t=super.getConfig();return Object.assign(e,t),e}};Aw.className="Dense";se.registerClass(Aw);var $w=class extends Ye{constructor(e){e=e||{},super(e),this.inputSpec=[{minNDim:3}],this.dataFormat=e.dataFormat}computeOutputShape(e){e=it(e);for(let t of e.slice(1))if(t==null)throw new H(`The shape of the input to "Flatten" is not fully defined (got ${e.slice(1)}). Make sure to pass a complete "input_shape" or "batch_input_shape" argument to the first layer in your model.`);return[e[0],is(e,1)]}call(e,t){return O(()=>{this.invokeCallHook(e,t);let n=ze(e);if(this.dataFormat==="channelsFirst"&&n.rank>1){let a=[0];for(let r=2;r<n.rank;++r)a.push(r);a.push(1),n=Ae(n,a)}return eV(n)})}getConfig(){let e={};this.dataFormat!=null&&(e.dataFormat=this.dataFormat);let t=super.getConfig();return Object.assign(e,t),e}};$w.className="Flatten";se.registerClass($w);var Dw=class extends Ye{constructor(e){super(e),this.supportsMasking=!0,this.activation=hs(e.activation)}call(e,t){return O(()=>{this.invokeCallHook(e,t);let n=ze(e);return this.activation.apply(n)})}getConfig(){let e={activation:ds(this.activation)},t=super.getConfig();return Object.assign(e,t),e}};Dw.className="Activation";se.registerClass(Dw);var Rw=class extends Ye{constructor(e){super(e),this.n=e.n,this.inputSpec=[{ndim:2}]}computeOutputShape(e){return[e[0],this.n,e[1]]}call(e,t){return O(()=>(e=ze(e),J4(e,this.n)))}getConfig(){let e={n:this.n},t=super.getConfig();return Object.assign(e,t),e}};Rw.className="RepeatVector";se.registerClass(Rw);var Mw=class extends Ye{constructor(e){super(e),this.targetShape=e.targetShape;for(let t=0;t<this.targetShape.length;++t)this.isUnknown(this.targetShape[t])&&(this.targetShape[t]=null)}isUnknown(e){return e<0||e==null}fixUnknownDimension(e,t){let n="Total size of new array must be unchanged.",a=t.slice(),r=1,s=null;for(let o=0;o<a.length;++o){let l=a[o];if(this.isUnknown(l))if(s===null)s=o;else throw new H("Can only specifiy one unknown dimension.");else r*=l}let i=is(e);if(s!==null){if(r===0||i%r!==0)throw new H(n);a[s]=i/r}else if(i!==r)throw new H(n);return a}computeOutputShape(e){let t=!1;for(let n=0;n<e.length;++n)if(this.isUnknown(e[n])){t=!0;break}return t?e.slice(0,1).concat(this.targetShape):e.slice(0,1).concat(this.fixUnknownDimension(e.slice(1),this.targetShape))}call(e,t){return O(()=>{this.invokeCallHook(e,t);let n=ze(e),a=n.shape,r=a.slice(0,1).concat(this.fixUnknownDimension(a.slice(1),this.targetShape));return W(n,r)})}getConfig(){let e={targetShape:this.targetShape},t=super.getConfig();return Object.assign(e,t),e}};Mw.className="Reshape";se.registerClass(Mw);var Pw=class extends Ye{constructor(e){if(super(e),e.dims==null)throw new Error("Required configuration field `dims` is missing during Permute constructor call.");if(!Array.isArray(e.dims))throw new Error(`Permute constructor requires \`dims\` to be an Array, but received ${e.dims} instead.`);let t=Ka(1,e.dims.length+1);if(!w.arraysEqual(e.dims.slice().sort(),t))throw new Error("Invalid permutation `dims`: "+JSON.stringify(e.dims)+" `dims` must contain consecutive integers starting from 1.");this.dims=e.dims,this.dimsIncludingBatch=[0].concat(this.dims),this.inputSpec=[new Bt({ndim:this.dims.length+1})]}computeOutputShape(e){e=it(e);let t=e.slice();return this.dims.forEach((n,a)=>{t[a+1]=e[n]}),t}call(e,t){return Ae(ze(e),this.dimsIncludingBatch)}getConfig(){let e={dims:this.dims},t=super.getConfig();return Object.assign(e,t),e}};Pw.className="Permute";se.registerClass(Pw);var Ow=class extends Ye{constructor(e){super(e==null?{}:e),this.supportsMasking=!0,e!=null?this.maskValue=e.maskValue==null?0:e.maskValue:this.maskValue=0}computeOutputShape(e){return e}getConfig(){let e=super.getConfig(),t={maskValue:this.maskValue};return Object.assign(t,e),t}computeMask(e,t){let n=ze(e),a=-1;return rc(di(n,this.maskValue),a)}call(e,t){return O(()=>{this.invokeCallHook(e,t);let n=ze(e),a=-1,r=!0,s=rc(di(n,this.maskValue),a,r);return B(n,oe(s,n.dtype))})}};Ow.className="Masking";se.registerClass(Ow);var Lw=class extends Ye{constructor(e){if(super(e),this.embeddings=null,this.DEFAULT_EMBEDDINGS_INITIALIZER="randomUniform",e.batchInputShape==null&&e.inputShape==null){let t=null;e.batchSize!=null&&(t=e.batchSize),e.inputLength==null?this.batchInputShape=[t,null]:this.batchInputShape=[t].concat(kt(e.inputLength))}this.inputDim=e.inputDim,nn(this.inputDim,"inputDim"),this.outputDim=e.outputDim,nn(this.outputDim,"outputDim"),this.embeddingsInitializer=Nt(e.embeddingsInitializer||this.DEFAULT_EMBEDDINGS_INITIALIZER),this.embeddingsRegularizer=Tt(e.embeddingsRegularizer),this.activityRegularizer=Tt(e.activityRegularizer),this.embeddingsConstraint=Yt(e.embeddingsConstraint),this.maskZero=e.maskZero,this.supportsMasking=e.maskZero,this.inputLength=e.inputLength}build(e){this.embeddings=this.addWeight("embeddings",[this.inputDim,this.outputDim],this.dtype,this.embeddingsInitializer,this.embeddingsRegularizer,!0,this.embeddingsConstraint),this.built=!0}warnOnIncompatibleInputShape(e){}computeMask(e,t){return O(()=>this.maskZero?(e=ze(e),di(e,Ke(e))):null)}computeOutputShape(e){if(e=it(e),this.inputLength==null)return[...e,this.outputDim];let t=kt(this.inputLength);if(t.length!==e.length-1)throw new H(`"inputLength" is ${this.inputLength}, but received input shape has shape ${e}`);{let n=0;for(let a=0;a<t.length;++a){let r=t[a],s=e[a+1];if(r!=null&&s!=null&&r!==s)throw new H(`"inputLength" is ${this.inputLength}, but received input shape has shape ${e}`);r==null&&(t[n]=s),n++}}return[e[0],...t,this.outputDim]}call(e,t){return O(()=>{this.invokeCallHook(e,t);let n=ze(e);n.dtype!=="int32"&&(n=Lf(n,"int32"));let a=M2(this.embeddings.read(),W(n,[n.size]));return W(a,it(this.computeOutputShape(n.shape)))})}getConfig(){let e={inputDim:this.inputDim,outputDim:this.outputDim,embeddingsInitializer:Ft(this.embeddingsInitializer),embeddingsRegularizer:dt(this.embeddingsRegularizer),activityRegularizer:dt(this.activityRegularizer),embeddingsConstraint:Xt(this.embeddingsConstraint),maskZero:this.maskZero,inputLength:this.inputLength},t=super.getConfig();return Object.assign(e,t),e}};Lw.className="Embedding";se.registerClass(Lw);var Io=class extends Ye{constructor(e){super(e||{}),this.supportsMasking=!0}mergeFunction(e){throw new Oe}computeElementwiseOpOutputShape(e,t){if(e==null||t==null)return null;if(e.length<t.length)return this.computeElementwiseOpOutputShape(t,e);if(t.length===0)return e;let n=e.slice(0,e.length-t.length);for(let a=0;a<t.length;++a){let r=e[e.length-t.length+a],s=t[a];if(r==null||s==null||r<0||s<0)n.push(null);else if(r===1)n.push(s);else if(s===1)n.push(r);else{if(r!==s)throw new H("Operands could not be broadcast together with shapes "+JSON.stringify(e)+" "+JSON.stringify(t));n.push(r)}}return n}build(e){if(Array.isArray(e)&&!Array.isArray(e[0])&&(e=[it(e)]),e=e,e.length<2)throw new H(`A merge layer should be called on an Array of at least 2 inputs. Got ${e.length} input(s).`);let t=[];for(let r of e)r!=null&&r[0]!==null&&t.push(r[0]);if(t=ss(t),t.length>1)throw new H(`Can not merge tensors with different batch sizes. Got tensors with shapes: ${JSON.stringify(e)}.`);let n=e[0]==null?null:e[0].slice(1);for(let r=1;r<e.length;++r){let s=e[r]==null?null:e[r].slice(1);n=this.computeElementwiseOpOutputShape(n,s)}let a=e.map(r=>r.length);e.indexOf(null)===-1&&ss(a).length===1?this.reshapeRequired=!1:this.reshapeRequired=!0}call(e,t){return O(()=>{if(e=e,this.reshapeRequired){let n=[],a=e.map(r=>r.rank);if(a.indexOf(null)===-1){let r=cs(a);for(let s of e){let i=s.rank;for(let o=0;o<r-i;++o)s=Zc(s,1);n.push(s)}return this.mergeFunction(n)}else{let r=!1;for(let o of e){let l=o.rank;if(l==null){let u=o.shape,p=u[0],d=u.slice(1).concat([p]),c=W(o,[p].concat(is(u.slice(1))));c=Ae(c,[1,0]),c=W(c,d),n.push(c),r=!0}else if(l>1){let u=Ka(1,l).concat([0]);n.push(Ae(o,u)),r=!0}else n.push(o)}let s=this.mergeFunction(n),i=s.rank;if(r){if(i==null){let o=s.shape,l=o.length,u=o[l-1],p=[u].concat(o.slice(0,o.length-1));s=W(Ae(W(s,[-1,u]),[1,0]),p)}else if(i>1){let o=[i-1].concat(Ka(0,i-1));s=Ae(s,o)}}return s}}else return this.mergeFunction(e)})}computeOutputShape(e){e=e;let t;e[0]==null?t=null:t=e[0].slice(1);for(let a=1;a<e.length;++a){let r=e[a]==null?null:e[a].slice(1);t=this.computeElementwiseOpOutputShape(t,r)}let n=[];for(let a of e)a!=null&&a[0]!==null&&n.push(a[0]);return n=ss(n),n.length===1?t=n.concat(t):t=[null].concat(t),t}computeMask(e,t){return O(()=>{if(t==null)return null;if(!Array.isArray(t))throw new H("`mask` should be an Array");if(!Array.isArray(e))throw new H("`inputs` should be an Array");if(t.length!==e.length)throw new H(`The Array 'inputs' and 'mask' are expected to have the same length, but have different lengths (${e.length} vs ${t.length})`);if(t.every(a=>a==null))return null;t=t.map(a=>a==null?a:mn(a,0));let n=t[0];for(let a=1;a<t.length-1;++a)n=Ca(n,t[a]);return n})}},zw=class extends Io{constructor(e){super(e)}mergeFunction(e){return O(()=>{let t=e[0].clone();for(let n=1;n<e.length;++n)t=J(t,e[n]);return t})}};zw.className="Add";se.registerClass(zw);var Bw=class extends Io{constructor(e){super(e)}mergeFunction(e){return O(()=>{let t=e[0].clone();for(let n=1;n<e.length;++n)t=B(t,e[n]);return t})}};Bw.className="Multiply";se.registerClass(Bw);var Ww=class extends Io{constructor(e){super(e)}mergeFunction(e){return O(()=>{let t=e[0].clone();for(let n=1;n<e.length;++n)t=J(t,e[n]);return B(1/e.length,t)})}};Ww.className="Average";se.registerClass(Ww);var Vw=class extends Io{constructor(e){super(e)}mergeFunction(e){return O(()=>{let t=e[0];for(let n=1;n<e.length;++n)t=fr(t,e[n]);return t})}};Vw.className="Maximum";se.registerClass(Vw);var Uw=class extends Io{constructor(e){super(e)}mergeFunction(e){return O(()=>{let t=e[0];for(let n=1;n<e.length;++n)t=Uu(t,e[n]);return t})}};Uw.className="Minimum";se.registerClass(Uw);var Gw=class extends Io{constructor(e){super(e),this.DEFAULT_AXIS=-1,e==null&&(e={}),this.axis=e.axis==null?this.DEFAULT_AXIS:e.axis,this.supportsMasking=!0,this.reshapeRequired=!1}build(e){if(!(Array.isArray(e)&&Array.isArray(e[0]))||e.length===1)throw new H("A `Concatenate` layer should be called on a list of at least 2 inputs");e=e;let t=!0;for(let a of e)if(a!=null){t=!1;break}if(t)return;let n=[];for(let a=0;a<e.length;++a){let r=e[a].slice();r.splice(this.axis,1);let s=!1;for(let i of n)if(w.arraysEqual(i,r)){s=!0;break}s||n.push(r)}if(n.length>1)throw new H("A `Concatenate` layer requires inputs with matching shapes except for the concat axis. Got input shapes: "+JSON.stringify(e))}mergeFunction(e){return O(()=>Gv(e,this.axis))}computeOutputShape(e){if(!(Array.isArray(e)&&Array.isArray(e[0])))throw new H("A `Concatenate` layer should be called on a list of inputs.");let t=e,n=t[0].slice(),a=this.axis<0?n.length+this.axis:this.axis;for(let r of t.slice(1)){if(n[a]==null||r[a]==null){n[a]=null;break}n[a]+=r[a]}return n}computeMask(e,t){if(t==null)return null;if(!Array.isArray(t))throw new H("`mask` should be an array for Concatenate");if(!Array.isArray(e))throw new H("`inputs` should be an array for Concatenate");if(t.length!==e.length)throw new H(`Mismatch in the length of mask (${t.length}) and the legnth of inputs (${e.length})`);return O(()=>{let n=!0;if(t.forEach(s=>{if(s!=null){n=!1;return}}),n)return null;let a=[];for(let s=0;s<e.length;++s)t[s]==null?a.push(oe(ta(e[s]),"bool")):t[s].rank<e[s].rank?a.push(mn(t[s],-1)):a.push(t[s]);let r=Ze(a,this.axis);return sf(r,-1,!1)})}getConfig(){let e={axis:this.axis},t=super.getConfig();return Object.assign(e,t),e}};Gw.className="Concatenate";se.registerClass(Gw);function Pp(e,t){for(;e<0;)e+=t;return e}function WU(e,t,n){if(e.shape.length>3||t.shape.length>3)throw new Oe("batchDot is not implemented for tensors of 4D or higher rank yet");if(w.assert(e.shape.length>=2,()=>`batchDot requires the rank of x to be >= 2, but got ${e.shape.length}`),w.assert(e.shape.length>=2,()=>`batchDot requires the rank of y to be >= 2, but got ${t.shape.length}`),typeof n=="number"&&(n=[n,n]),e.dtype==="complex64"||t.dtype==="complex64")throw new Oe("batchDot is not implemented for complex64-type Tensors yet.");let a=e.shape.length,r=t.shape.length;n==null&&(n=[a-1,r-2]);let s=n;return O(()=>{let i;if(a>r){i=a-r;let l=[];for(let u=0;u<i;++u)l.push(1);t=W(t,t.shape.concat(l))}else if(r>a){i=r-a;let l=[];for(let u=0;u<i;++u)l.push(1);e=W(e,e.shape.concat(l))}else i=0;let o;if(e.shape.length===2&&t.shape.length===2)s[0]===s[1]?o=be(B(e,t),s[0]):o=be(B(Ae(e,[1,0]),t),s[1]);else{let l=s[0]!==e.shape.length-1,u=s[1]===t.shape.length-1;o=De(e,t,l,u)}if(i>0){let l;a>r?l=a+r-3:l=a-1;let u=[];for(let p=l;p<l+i;++p)u.push(p);o=dr(o,u)}return o.shape.length===1&&(o=mn(o,1)),o})}var Hw=class extends Io{constructor(e){super(e),this.axes=e.axes,this.normalize=e.normalize==null?!1:e.normalize,this.supportsMasking=!0,this.reshapeRequired=!1}build(e){w.assert(Array.isArray(e)&&e.length===2&&Array.isArray(e[0])&&Array.isArray(e[1]),()=>"A `Dot` layer should be called on a list of exactly 2 inputs.");let t=e[0],n=e[1];if(t.length>3||n.length>3)throw new Oe("Dot layer does not support tensors of 4D or higher rank yet.");let a=this.interpretAxes(t,n);if(t[a[0]]!==n[a[1]])throw new H(`Dimension incompatibility: ${t[a[0]]} !== ${n[a[1]]}`)}mergeFunction(e){if(e.length!==2)throw new H(`A \`Dot\` layer must be called on exactly 2 inputs, but received ${e.length} input(s).`);let t=e[0],n=e[1],a;return Array.isArray(this.axes)?a=this.axes.map((r,s)=>Pp(r,e[s].shape.length)):a=[Pp(this.axes,t.shape.length),Pp(this.axes,n.shape.length)],this.normalize&&(t=rm(t,a[0]),n=rm(n,a[1])),WU(t,n,a)}interpretAxes(e,t){let n;return Array.isArray(this.axes)?n=this.axes:n=[Pp(this.axes,e.length),Pp(this.axes,t.length)],n}computeOutputShape(e){w.assert(Array.isArray(e)&&e.length===2&&Array.isArray(e[0])&&Array.isArray(e[1]),()=>"A `Dot` layer should be called on a list of exactly 2 inputs.");let t=e[0].slice(),n=e[1].slice();if(t.length>3||n.length>3)throw new Oe("Dot layer does not support tensors of 4D or higher rank yet.");let a=this.interpretAxes(t,n);t.splice(a[0],1),n.splice(a[1],1),n.splice(0,1);let r=t.concat(n);return r.length===1&&r.push(1),r}computeMask(e,t){return null}getConfig(){let e={axes:this.axes,normalize:this.normalize},t=super.getConfig();return Object.assign(e,t),e}};Hw.className="Dot";se.registerClass(Hw);var jw=class extends Ye{constructor(e){super(e),this.supportsMasking=!0,this.stddev=e.stddev}computeOutputShape(e){return e}getConfig(){let e=super.getConfig(),t={stddev:this.stddev};return Object.assign(t,e),t}call(e,t){return O(()=>{this.invokeCallHook(e,t);let n=ze(e);return td(()=>J(zf(n.shape,0,this.stddev),n),()=>n,t.training||!1)})}};jw.className="GaussianNoise";se.registerClass(jw);var qw=class extends Ye{constructor(e){super(e),this.supportsMasking=!0,this.rate=e.rate}computeOutputShape(e){return e}getConfig(){let e=super.getConfig(),t={rate:this.rate};return Object.assign(t,e),t}call(e,t){return O(()=>{this.invokeCallHook(e,t);let n=ze(e);return this.rate>0&&this.rate<1?td(()=>{let a=Math.sqrt(this.rate/(1-this.rate));return B(n,zf(n.shape,1,a))},()=>n,t.training||!1):n})}};qw.className="GaussianDropout";se.registerClass(qw);var Kw=class extends Ye{constructor(e){super(e),this.supportsMasking=!0,this.rate=e.rate,this.noiseShape=e.noiseShape}_getNoiseShape(e){return this.noiseShape||ze(e).shape}computeOutputShape(e){return e}getConfig(){let e=super.getConfig(),t={rate:this.rate};return Object.assign(t,e),t}call(e,t){return O(()=>{if(this.rate<1&&this.rate>0){let n=this._getNoiseShape(e);return td(()=>{let a=ze(e),r=1.6732632423543772,s=1.0507009873554805,i=-r*s,o=Ns(Gu(n),this.rate);o=Lf(o,"float32");let l=((1-this.rate)*(1+this.rate*i**2))**-.5,u=-l*i*this.rate,p=J(B(a,o),B(J(o,-1),i));return J(B(p,l),u)},()=>ze(e),t.training||!1)}return e})}};Kw.className="AlphaDropout";se.registerClass(Kw);function pc(e,t,n,a,r,s=.001){let i;if(e.rank===2)i=ES(e,t,n,a,r,s);else if(e.rank===3)i=FS(e,t,n,a,r,s);else if(e.rank===4)i=AS(e,t,n,a,r,s);else throw new Oe(`batchNormalization is not implemented for array of rank ${e.rank} yet`);return i}function VU(e,t,n,a,r=.001){return O(()=>{let s=ff(e,a),i=s.mean,o=s.variance;return[pc(e,i,o,n,t,r),i,o]})}function UU(e,t,n,a,r=.001){return O(()=>{let s=ff(e,a),i=s.mean,o=s.variance,l=[];for(let h of Ka(0,e.rank))a.indexOf(h)!==-1?l.push(1):l.push(e.shape[h]);let u=W(i,l),p=W(o,l),d=t==null?null:W(t,l),c=n==null?null:W(n,l);return[pc(e,u,p,c,d,r),i,o]})}function GU(e,t,n,a,r=.001){return w.arraysEqual(a.slice().sort(),Ka(0,e.rank-1))?VU(e,t,n,a,r):UU(e,t,n,a,r)}var Xw=class extends Ye{constructor(e){e==null&&(e={}),super(e),this.supportsMasking=!0,this.axis=e.axis==null?-1:e.axis,this.momentum=e.momentum==null?.99:e.momentum,this.epsilon=e.epsilon==null?.001:e.epsilon,this.center=e.center==null?!0:e.center,this.scale=e.scale==null?!0:e.scale,this.betaInitializer=Nt(e.betaInitializer||"zeros"),this.gammaInitializer=Nt(e.gammaInitializer||"ones"),this.movingMeanInitializer=Nt(e.movingMeanInitializer||"zeros"),this.movingVarianceInitializer=Nt(e.movingVarianceInitializer||"ones"),this.betaConstraint=Yt(e.betaConstraint),this.gammaConstraint=Yt(e.gammaConstraint),this.betaRegularizer=Tt(e.betaRegularizer),this.gammaRegularizer=Tt(e.gammaRegularizer)}build(e){e=it(e);let t=this.axis>=0?this.axis:this.axis+e.length,n=e[t];if(n==null)throw new H(`Axis ${t} of input tensor should have a defined dimension but the layer received an input with shape ${JSON.stringify(e)}.`);this.inputSpec=[new Bt({ndim:e.length,axes:{[t]:n}})];let a=[n];this.scale&&(this.gamma=this.addWeight("gamma",a,null,this.gammaInitializer,this.gammaRegularizer,!0,this.gammaConstraint)),this.center&&(this.beta=this.addWeight("beta",a,null,this.betaInitializer,this.betaRegularizer,!0,this.betaConstraint)),this.movingMean=this.addWeight("moving_mean",a,null,this.movingMeanInitializer,null,!1),this.movingVariance=this.addWeight("moving_variance",a,null,this.movingVarianceInitializer,null,!1),this.built=!0}call(e,t){return O(()=>{let n=t.training==null?!1:t.training,a=ze(e),r=a.shape,s=r.length,i=Ka(0,s),o=this.axis>=0?this.axis:this.axis+s;i.splice(o,1);let l=hi(1,s);l[o]=r[o];let u=i.slice();u.sort();let p=!w.arraysEqual(u,Ka(0,s).slice(0,s-1)),d=()=>{if(p){let g=W(this.movingMean.read(),l),y=W(this.movingVariance.read(),l),b=this.center?W(this.beta.read(),l):null,x=this.scale?W(this.gamma.read(),l):null;return pc(a,g,y,b,x,this.epsilon)}else return pc(a,this.movingMean.read(),this.movingVariance.read(),this.beta==null?null:this.beta.read(),this.gamma==null?null:this.gamma.read(),this.epsilon)};if(!n)return d();let[c,h,m]=GU(a,this.gamma.read(),this.beta.read(),i,this.epsilon),f=(g,y,b)=>{O(()=>{let x=1-b,v=g.read(),k=B(ce(v,y),x);g.write(ce(v,k))})};return f(this.movingMean,h,this.momentum),f(this.movingVariance,m,this.momentum),c})}getConfig(){let e={axis:this.axis,momentum:this.momentum,epsilon:this.epsilon,center:this.center,scale:this.scale,betaInitializer:Ft(this.betaInitializer),gammaInitializer:Ft(this.gammaInitializer),movingMeanInitializer:Ft(this.movingMeanInitializer),movingVarianceInitializer:Ft(this.movingVarianceInitializer),betaRegularizer:dt(this.betaRegularizer),gammaRegularizer:dt(this.gammaRegularizer),betaConstraint:Xt(this.betaConstraint),gammaConstraint:Xt(this.gammaConstraint)},t=super.getConfig();return Object.assign(e,t),e}};Xw.className="BatchNormalization";se.registerClass(Xw);var Yw=class extends Ye{constructor(e){if(e==null&&(e={}),super(e),this.axis=e.axis==null?-1:e.axis,typeof this.axis=="number"){if(!Number.isInteger(this.axis))throw new Error(`Expected axis to be an integer, but received ${this.axis}`)}else if(Array.isArray(this.axis)){for(let t of this.axis)if(!Number.isInteger(t))throw new Error(`Expected axis to be an array of integers, but received ${JSON.stringify(this.axis)}`)}else throw new Error(`Expected axis to be an integer or an array of integers, but received ${JSON.stringify(this.axis)}`);this.epsilon=e.epsilon==null?.001:e.epsilon,this.center=e.center==null?!0:e.center,this.scale=e.scale==null?!0:e.scale,this.betaInitializer=Nt(e.betaInitializer||"zeros"),this.gammaInitializer=Nt(e.gammaInitializer||"ones"),this.betaRegularizer=Tt(e.betaRegularizer),this.gammaRegularizer=Tt(e.gammaRegularizer),this.supportsMasking=!0}build(e){e=it(e);let t=e.length;typeof this.axis=="number"&&(this.axis=[this.axis]);for(let r=0;r<this.axis.length;++r)this.axis[r]<0&&(this.axis[r]+=t);for(let r of this.axis)if(r<0||r>=t)throw new Error(`Invalid axis: ${r}`);if(this.axis.length!==ss(this.axis).length)throw new Error(`Found duplicate axes in: ${this.axis}`);let n=this.axis.map(r=>e[r]),a=!0;this.scale?this.gamma=this.addWeight("gamma",n,"float32",this.gammaInitializer,this.gammaRegularizer,a):this.gamma=null,this.center?this.beta=this.addWeight("beta",n,"float32",this.betaInitializer,this.betaRegularizer,a):this.beta=null,this.built=!0}call(e,t){let n=ze(e),a=n.shape,r=a.length;return O(()=>{let{mean:s,variance:i}=ff(n,this.axis,!0),o=hi(1,r);for(let h of this.axis)o[h]=a[h];let l=h=>h!=null&&h.shape.length!==r?W(h,o):h,u=this.scale?l(this.gamma.read()):null,p=this.center?l(this.beta.read()):null,d=[],c=[];for(let h=0;h<r;++h)this.axis.indexOf(h)!==-1?(d.push(a[h]),c.push(1)):(d.push(1),c.push(a[h]));return s=On(s,d),i=On(i,d),u!=null&&(u=On(u,c)),p!=null&&(p=On(p,c)),pc(n,s,i,p,u,this.epsilon)})}getConfig(){let e={axis:this.axis,epsilon:this.epsilon,center:this.center,scale:this.scale,betaInitializer:Ft(this.betaInitializer),gammaInitializer:Ft(this.gammaInitializer),betaRegularizer:dt(this.betaRegularizer),gammaRegularizer:dt(this.gammaRegularizer)},t=super.getConfig();return Object.assign(e,t),e}};Yw.className="LayerNormalization";se.registerClass(Yw);function HU(e,t,n){return O(()=>{if(e.rank!==4)throw new H(`temporalPadding expects input tensor to be 4-D, but received a ${e.rank}-D tensor.`);if(t==null&&(t=[[1,1],[1,1]]),t.length!==2||t[0].length!==2||t[1].length!==2)throw new H("spatial2dPadding expects `padding` to be an Array of two Arrays, each of which is an Array of two integers.");if(n==null&&(n=Xa()),n!=="channelsLast"&&n!=="channelsFirst")throw new H(`Unknown data format: ${n}. Supported data formats are 'channelsLast' and 'channelsFirst.`);let a;return n==="channelsFirst"?a=[[0,0],[0,0],t[0],t[1]]:a=[[0,0],t[0],t[1],[0,0]],ya(e,a)})}var Qw=class extends Ye{constructor(e){if(e==null&&(e={}),super(e),this.dataFormat=e.dataFormat==null?Xa():e.dataFormat,e.padding==null)this.padding=[[1,1],[1,1]];else if(typeof e.padding=="number")this.padding=[[e.padding,e.padding],[e.padding,e.padding]];else{if(e.padding=e.padding,e.padding.length!==2)throw new H(`ZeroPadding2D expects padding to be a length-2 array, but received a length-${e.padding.length} array.`);let t,n;if(typeof e.padding[0]=="number")t=[e.padding[0],e.padding[0]],n=[e.padding[1],e.padding[1]];else{if(e.padding=e.padding,e.padding[0].length!==2)throw new H(`ZeroPadding2D expects height padding to be a length-2 array, but received a length-${e.padding[0].length} array.`);if(t=e.padding[0],e.padding[1].length!==2)throw new H(`ZeroPadding2D expects width padding to be a length-2 array, but received a length-${e.padding[1].length} array.`);n=e.padding[1]}this.padding=[t,n]}this.inputSpec=[new Bt({ndim:4})]}computeOutputShape(e){e=it(e);let t,n;return this.dataFormat==="channelsFirst"?(e[2]!=null&&e[2]>=0?t=e[2]+this.padding[0][0]+this.padding[0][1]:t=null,e[3]!=null&&e[3]>=0?n=e[3]+this.padding[1][0]+this.padding[1][1]:n=null,[e[0],e[1],t,n]):(e[1]!=null&&e[1]>=0?t=e[1]+this.padding[0][0]+this.padding[0][1]:t=null,e[2]!=null&&e[2]>=0?n=e[2]+this.padding[1][0]+this.padding[1][1]:n=null,[e[0],t,n,e[3]])}call(e,t){return O(()=>HU(ze(e),this.padding,this.dataFormat))}getConfig(){let e={padding:this.padding,dataFormat:this.dataFormat},t=super.getConfig();return Object.assign(e,t),e}};Qw.className="ZeroPadding2D";se.registerClass(Qw);function ng(e,t,n,a,r,s){return O(()=>{Ot(r),A2(s),ba(a),n==null&&(n=[1,1]),a==null&&(a="valid"),r==null&&(r=Xa()),s==null&&(s="max"),e=bw(e,r);let i,o=a==="same"?"same":"valid";return s==="max"?i=Pt(e,t,n,o):i=ga(e,t,n,o),r==="channelsFirst"&&(i=Ae(i,[0,3,1,2])),i})}function TN(e,t,n,a,r,s){return O(()=>{Ot(r),A2(s),ba(a),n==null&&(n=[1,1,1]),a==null&&(a="valid"),r==null&&(r=Xa()),s==null&&(s="max"),e=wN(e,r);let i,o=a==="same"?"same":"valid";return s==="max"?i=Iv(e,t,n,o):i=sv(e,t,n,o),r==="channelsFirst"&&(i=Ae(i,[0,4,1,2,3])),i})}var _N=class extends Ye{constructor(e){if(e.poolSize==null&&(e.poolSize=2),super(e),typeof e.poolSize=="number")this.poolSize=[e.poolSize];else if(Array.isArray(e.poolSize)&&e.poolSize.length===1&&typeof e.poolSize[0]=="number")this.poolSize=e.poolSize;else throw new H(`poolSize for 1D convolutional layer must be a number or an Array of a single number, but received ${JSON.stringify(e.poolSize)}`);if(nn(this.poolSize,"poolSize"),e.strides==null)this.strides=this.poolSize;else if(typeof e.strides=="number")this.strides=[e.strides];else if(Array.isArray(e.strides)&&e.strides.length===1&&typeof e.strides[0]=="number")this.strides=e.strides;else throw new H(`strides for 1D convolutional layer must be a number or an Array of a single number, but received ${JSON.stringify(e.strides)}`);nn(this.strides,"strides"),this.padding=e.padding==null?"valid":e.padding,ba(this.padding),this.inputSpec=[new Bt({ndim:3})]}computeOutputShape(e){e=it(e);let t=qa(e[1],this.poolSize[0],this.padding,this.strides[0]);return[e[0],t,e[2]]}call(e,t){return O(()=>{this.invokeCallHook(e,t),e=Zc(ze(e),2);let n=this.poolingFunction(ze(e),[this.poolSize[0],1],[this.strides[0],1],this.padding,"channelsLast");return dr(n,[2])})}getConfig(){let e={poolSize:this.poolSize,padding:this.padding,strides:this.strides},t=super.getConfig();return Object.assign(e,t),e}},Jw=class extends _N{constructor(e){super(e)}poolingFunction(e,t,n,a,r){return Ot(r),ba(a),ng(e,t,n,a,r,"max")}};Jw.className="MaxPooling1D";se.registerClass(Jw);var Zw=class extends _N{constructor(e){super(e)}poolingFunction(e,t,n,a,r){return Ot(r),ba(a),ng(e,t,n,a,r,"avg")}};Zw.className="AveragePooling1D";se.registerClass(Zw);var CN=class extends Ye{constructor(e){if(e.poolSize==null&&(e.poolSize=[2,2]),super(e),this.poolSize=Array.isArray(e.poolSize)?e.poolSize:[e.poolSize,e.poolSize],e.strides==null)this.strides=this.poolSize;else if(Array.isArray(e.strides)){if(e.strides.length!==2)throw new H(`If the strides property of a 2D pooling layer is an Array, it is expected to have a length of 2, but received length ${e.strides.length}.`);this.strides=e.strides}else this.strides=[e.strides,e.strides];nn(this.poolSize,"poolSize"),nn(this.strides,"strides"),this.padding=e.padding==null?"valid":e.padding,this.dataFormat=e.dataFormat==null?"channelsLast":e.dataFormat,Ot(this.dataFormat),ba(this.padding),this.inputSpec=[new Bt({ndim:4})]}computeOutputShape(e){e=it(e);let t=this.dataFormat==="channelsFirst"?e[2]:e[1],n=this.dataFormat==="channelsFirst"?e[3]:e[2];return t=qa(t,this.poolSize[0],this.padding,this.strides[0]),n=qa(n,this.poolSize[1],this.padding,this.strides[1]),this.dataFormat==="channelsFirst"?[e[0],e[1],t,n]:[e[0],t,n,e[3]]}call(e,t){return O(()=>(this.invokeCallHook(e,t),this.poolingFunction(ze(e),this.poolSize,this.strides,this.padding,this.dataFormat)))}getConfig(){let e={poolSize:this.poolSize,padding:this.padding,strides:this.strides,dataFormat:this.dataFormat},t=super.getConfig();return Object.assign(e,t),e}},e0=class extends CN{constructor(e){super(e)}poolingFunction(e,t,n,a,r){return Ot(r),ba(a),ng(e,t,n,a,r,"max")}};e0.className="MaxPooling2D";se.registerClass(e0);var t0=class extends CN{constructor(e){super(e)}poolingFunction(e,t,n,a,r){return Ot(r),ba(a),ng(e,t,n,a,r,"avg")}};t0.className="AveragePooling2D";se.registerClass(t0);var EN=class extends Ye{constructor(e){if(e.poolSize==null&&(e.poolSize=[2,2,2]),super(e),this.poolSize=Array.isArray(e.poolSize)?e.poolSize:[e.poolSize,e.poolSize,e.poolSize],e.strides==null)this.strides=this.poolSize;else if(Array.isArray(e.strides)){if(e.strides.length!==3)throw new H(`If the strides property of a 3D pooling layer is an Array, it is expected to have a length of 3, but received length ${e.strides.length}.`);this.strides=e.strides}else this.strides=[e.strides,e.strides,e.strides];nn(this.poolSize,"poolSize"),nn(this.strides,"strides"),this.padding=e.padding==null?"valid":e.padding,this.dataFormat=e.dataFormat==null?"channelsLast":e.dataFormat,Ot(this.dataFormat),ba(this.padding),this.inputSpec=[new Bt({ndim:5})]}computeOutputShape(e){e=it(e);let t=this.dataFormat==="channelsFirst"?e[2]:e[1],n=this.dataFormat==="channelsFirst"?e[3]:e[2],a=this.dataFormat==="channelsFirst"?e[4]:e[3];return t=qa(t,this.poolSize[0],this.padding,this.strides[0]),n=qa(n,this.poolSize[1],this.padding,this.strides[1]),a=qa(a,this.poolSize[2],this.padding,this.strides[2]),this.dataFormat==="channelsFirst"?[e[0],e[1],t,n,a]:[e[0],t,n,a,e[4]]}call(e,t){return O(()=>(this.invokeCallHook(e,t),this.poolingFunction(ze(e),this.poolSize,this.strides,this.padding,this.dataFormat)))}getConfig(){let e={poolSize:this.poolSize,padding:this.padding,strides:this.strides,dataFormat:this.dataFormat},t=super.getConfig();return Object.assign(e,t),e}},n0=class extends EN{constructor(e){super(e)}poolingFunction(e,t,n,a,r){return Ot(r),ba(a),TN(e,t,n,a,r,"max")}};n0.className="MaxPooling3D";se.registerClass(n0);var a0=class extends EN{constructor(e){super(e)}poolingFunction(e,t,n,a,r){return Ot(r),ba(a),TN(e,t,n,a,r,"avg")}};a0.className="AveragePooling3D";se.registerClass(a0);var FN=class extends Ye{constructor(e){super(e),this.inputSpec=[new Bt({ndim:3})]}computeOutputShape(e){return[e[0],e[2]]}call(e,t){throw new Oe}},r0=class extends FN{constructor(e){super(e||{})}call(e,t){return O(()=>{let n=ze(e);return Et(n,1)})}};r0.className="GlobalAveragePooling1D";se.registerClass(r0);var s0=class extends FN{constructor(e){super(e||{})}call(e,t){return O(()=>{let n=ze(e);return Ta(n,1)})}};s0.className="GlobalMaxPooling1D";se.registerClass(s0);var AN=class extends Ye{constructor(e){super(e),this.dataFormat=e.dataFormat==null?"channelsLast":e.dataFormat,Ot(this.dataFormat),this.inputSpec=[new Bt({ndim:4})]}computeOutputShape(e){return e=e,this.dataFormat==="channelsLast"?[e[0],e[3]]:[e[0],e[1]]}call(e,t){throw new Oe}getConfig(){let e={dataFormat:this.dataFormat},t=super.getConfig();return Object.assign(e,t),e}},i0=class extends AN{call(e,t){return O(()=>{let n=ze(e);return this.dataFormat==="channelsLast"?Et(n,[1,2]):Et(n,[2,3])})}};i0.className="GlobalAveragePooling2D";se.registerClass(i0);var o0=class extends AN{call(e,t){return O(()=>{let n=ze(e);return this.dataFormat==="channelsLast"?Ta(n,[1,2]):Ta(n,[2,3])})}};o0.className="GlobalMaxPooling2D";se.registerClass(o0);var $N=class extends Ye{constructor(e){super(e),this.layer=e.layer}build(e){this.built=!0}get trainable(){return this.layer!=null?this.layer.trainable:!1}set trainable(e){this.layer!=null&&(this.layer.trainable=e)}get trainableWeights(){return this.layer.trainableWeights}get nonTrainableWeights(){return this.layer.nonTrainableWeights}get updates(){return this.layer._updates}get losses(){return this.layer.losses}getWeights(){return this.layer.getWeights()}setWeights(e){this.layer.setWeights(e)}getConfig(){let e={layer:{className:this.layer.getClassName(),config:this.layer.getConfig()}},t=super.getConfig();return Object.assign(e,t),e}setFastWeightInitDuringBuild(e){super.setFastWeightInitDuringBuild(e),this.layer!=null&&this.layer.setFastWeightInitDuringBuild(e)}static fromConfig(e,t,n={}){let a=t.layer,r=ja(a,n);delete t.layer;let s={layer:r};return Object.assign(s,t),new e(s)}},l0=class extends $N{constructor(e){super(e),this.supportsMasking=!0}build(e){if(e=it(e),e.length<3)throw new H(`TimeDistributed layer expects an input shape >= 3D, but received input shape ${JSON.stringify(e)}`);this.inputSpec=[{shape:e}];let t=[e[0]].concat(e.slice(2));this.layer.built||(this.layer.build(t),this.layer.built=!0),super.build(e)}computeOutputShape(e){e=it(e);let t=[e[0]].concat(e.slice(2)),n=this.layer.computeOutputShape(t),a=e[1];return[n[0],a].concat(n.slice(1))}call(e,t){return O(()=>(e=ze(e),SN((n,a)=>[ze(this.layer.call(n,t)),[]],e,[],!1,null,null,!1,!0)[1]))}};l0.className="TimeDistributed";se.registerClass(l0);function jU(e){wo(q4,"BidirectionalMergeMode",e)}var qU="concat",u0=class extends $N{constructor(e){super(e);let t=e.layer.getConfig(),n={};n.className=e.layer.getClassName(),n.config=t,this.forwardLayer=ja(n),t.goBackwards=t.goBackwards!==!0;let a={};if(a.className=e.layer.getClassName(),a.config=t,this.backwardLayer=ja(a),this.forwardLayer.name="forward_"+this.forwardLayer.name,this.backwardLayer.name="backward_"+this.backwardLayer.name,this.mergeMode=e.mergeMode===void 0?qU:e.mergeMode,jU(this.mergeMode),e.weights)throw new Oe("weights support is not implemented for Bidirectional layer yet.");this._stateful=e.layer.stateful,this.returnSequences=e.layer.returnSequences,this.returnState=e.layer.returnState,this.supportsMasking=!0,this._trainable=!0,this.inputSpec=e.layer.inputSpec,this.numConstants=null}get trainable(){return this._trainable}set trainable(e){this._trainable=e,this.forwardLayer!=null&&(this.forwardLayer.trainable=e),this.backwardLayer!=null&&(this.backwardLayer.trainable=e)}getWeights(){return this.forwardLayer.getWeights().concat(this.backwardLayer.getWeights())}setWeights(e){let t=e.length,n=Math.floor(t/2);this.forwardLayer.setWeights(e.slice(0,n)),this.backwardLayer.setWeights(e.slice(n))}computeOutputShape(e){let t=this.forwardLayer.computeOutputShape(e);Array.isArray(t)&&Array.isArray(t[0])||(t=[t]),t=t;let n,a,r;return this.returnState&&(r=t.slice(1)),n=t[0],n=n,this.mergeMode==="concat"?(n[n.length-1]*=2,a=[n]):this.mergeMode==null?a=[n,n.slice()]:a=[n],this.returnState?this.mergeMode==null?a.concat(r).concat(r.slice()):[n].concat(r).concat(r.slice()):Pn(a)}apply(e,t){let n=t==null?null:t.initialState,a=t==null?null:t.constants;t==null&&(t={});let r=IN(e,n,a,this.numConstants);if(e=r.inputs,n=r.initialState,a=r.constants,Array.isArray(e)&&(n=e.slice(1),e=e[0]),(n==null||n.length===0)&&a==null)return super.apply(e,t);let s=[],i=[];if(n!=null){let l=n.length;if(l%2>0)throw new H("When passing `initialState` to a Bidrectional RNN, the state should be an Array containing the states of the underlying RNNs.");t.initialState=n,s.push(...n);let u=n.map(p=>new Bt({shape:p.shape}));this.forwardLayer.stateSpec=u.slice(0,l/2),this.backwardLayer.stateSpec=u.slice(l/2),i.push(...u)}if(a!=null)throw new Oe("Support for constants in Bidirectional layers is not implemented yet.");let o=s[0]instanceof Ua;for(let l of s)if(l instanceof Ua!==o)throw new H("The initial state of a Bidirectional layer cannot be specified as a mix of symbolic and non-symbolic tensors");if(o){let l=[e].concat(s),u=this.inputSpec.concat(i),p=this.inputSpec;this.inputSpec=u;let d=super.apply(l,t);return this.inputSpec=p,d}else return super.apply(e,t)}call(e,t){return O(()=>{let n=t.initialState,a,r;if(n==null)a=this.forwardLayer.call(e,t),r=this.backwardLayer.call(e,t);else{let o=n.slice(0,n.length/2),l=n.slice(n.length/2);a=this.forwardLayer.call(e,Object.assign(t,{initialState:o})),r=this.backwardLayer.call(e,Object.assign(t,{initialState:l}))}let s;this.returnState&&(Array.isArray(a)&&(s=a.slice(1).concat(r.slice(1))),a=a[0],r=r[0]),this.returnSequences&&(r=na(r,1));let i;return this.mergeMode==="concat"?i=Gv([a,r]):this.mergeMode==="sum"?i=J(a,r):this.mergeMode==="ave"?i=B(.5,J(a,r)):this.mergeMode==="mul"?i=B(a,r):this.mergeMode==null&&(i=[a,r]),this.returnState?this.mergeMode==null?i.concat(s):[i].concat(s):i})}resetStates(e){this.forwardLayer.resetStates(),this.backwardLayer.resetStates()}build(e){ei(this.forwardLayer.name,()=>{this.forwardLayer.build(e)}),ei(this.backwardLayer.name,()=>{this.backwardLayer.build(e)}),this.built=!0}computeMask(e,t){Array.isArray(t)&&(t=t[0]);let n;if(this.returnSequences?this.mergeMode==null?n=[t,t]:n=t:this.mergeMode==null?n=[null,null]:n=null,this.returnState){let a=this.forwardLayer.states.map(r=>null);return Array.isArray(n)?n.concat(a).concat(a):[n].concat(a).concat(a)}else return n}get trainableWeights(){return this.forwardLayer.trainableWeights.concat(this.backwardLayer.trainableWeights)}get nonTrainableWeights(){return this.forwardLayer.nonTrainableWeights.concat(this.backwardLayer.nonTrainableWeights)}setFastWeightInitDuringBuild(e){super.setFastWeightInitDuringBuild(e),this.forwardLayer!=null&&this.forwardLayer.setFastWeightInitDuringBuild(e),this.backwardLayer!=null&&this.backwardLayer.setFastWeightInitDuringBuild(e)}getConfig(){let e={mergeMode:this.mergeMode},t=super.getConfig();return Object.assign(e,t),e}static fromConfig(e,t){let n=ja(t.layer);if(delete t.layer,t.numConstants!=null)throw new Oe("Deserialization of a Bidirectional layer with numConstants present is not supported yet.");let a=t;return a.layer=n,new e(a)}};u0.className="Bidirectional";se.registerClass(u0);function KU(e){return new Ku(e)}function XU(e){return new fw(e)}function YU(e){return new dw(e)}function QU(e){return new hw(e)}function JU(e){return new mw(e)}function ZU(e){return new yw(e)}function eG(e){return new gw(e)}function tG(e){return new Yf(e)}function nG(e){return new sd(e)}function aG(e){return new vw(e)}function rG(e){return new id(e)}function sG(e){return new ww(e)}function iG(e){return new kw(e)}function oG(e){return new Iw(e)}function lG(e){return new Sw(e)}function uG(e){return new Nw(e)}function pG(e){return new Dw(e)}function cG(e){return new Aw(e)}function dG(e){return new tg(e)}function hG(e){return new Fw(e)}function mG(e){return new $w(e)}function fG(e){return new Rw(e)}function gG(e){return new Mw(e)}function yG(e){return new Pw(e)}function bG(e){return new Lw(e)}function xG(e){return new zw(e)}function vG(e){return new Ww(e)}function wG(e){return new Gw(e)}function kG(e){return new Vw(e)}function IG(e){return new Uw(e)}function SG(e){return new Bw(e)}function NG(e){return new Hw(e)}function TG(e){return new Xw(e)}function _G(e){return new Yw(e)}function CG(e){return new Qw(e)}function p0(e){return new Zw(e)}function EG(e){return p0(e)}function FG(e){return p0(e)}function c0(e){return new t0(e)}function AG(e){return c0(e)}function $G(e){return c0(e)}function d0(e){return new a0(e)}function DG(e){return d0(e)}function RG(e){return d0(e)}function MG(e){return new r0(e)}function PG(e){return new i0(e)}function DN(e){return new s0(e)}function RN(e){return new o0(e)}function MN(e){return new Jw(e)}function PN(e){return new e0(e)}function OG(e){return new n0(e)}function LG(e){return new _w(e)}function zG(e){return new Jf(e)}function BG(e){return new Cw(e)}function WG(e){return new ld(e)}function VG(e){return new Tw(e)}function UG(e){return new Qf(e)}function GG(e){return new Ew(e)}function HG(e){return new eg(e)}function jG(e){return new yr(e)}function qG(e){return new Zf(e)}function KG(e){return new u0(e)}function XG(e){return new l0(e)}var YG=DN,QG=RN,JG=MN,ZG=PN;function e6(e){return new jw(e)}function t6(e){return new qw(e)}function n6(e){return new Kw(e)}function a6(e){return new Ow(e)}var ON={};Me(ON,{MAPE:()=>m6,MSE:()=>y6,binaryAccuracy:()=>r6,binaryCrossentropy:()=>s6,categoricalAccuracy:()=>o6,categoricalCrossentropy:()=>l6,cosineProximity:()=>c6,mape:()=>f6,meanAbsoluteError:()=>d6,meanAbsolutePercentageError:()=>h6,meanSquaredError:()=>g6,mse:()=>b6,precision:()=>u6,recall:()=>p6,sparseCategoricalAccuracy:()=>i6});function r6(e,t){return sw(e,t)}function s6(e,t){return Q2(e,t)}function i6(e,t){return J2(e,t)}function o6(e,t){return iw(e,t)}function l6(e,t){return ow(e,t)}function u6(e,t){return Y2(e,t)}function p6(e,t){return QV(e,t)}function c6(e,t){return rw(e,t)}function d6(e,t){return Kf(e,t)}function h6(e,t){return Xu(e,t)}function m6(e,t){return Xu(e,t)}function f6(e,t){return Xu(e,t)}function g6(e,t){return ko(e,t)}function y6(e,t){return ko(e,t)}function b6(e,t){return ko(e,t)}var LN={};Me(LN,{modelFromJSON:()=>_U});var zN={};Me(zN,{l1:()=>v6,l1l2:()=>x6,l2:()=>w6});function x6(e){return new ad(e)}function v6(e){return MU(e)}function w6(e){return PU(e)}var BN=class extends El{constructor(){super(...arguments),this.model=null}setModel(e){if(!(e instanceof Fr))throw new Error("model must be a LayersModel, not some other Container");this.model=e}};function Sh(e,t){return e<t}function $k(e,t){return e>t}var WN=class extends BN{constructor(e){if(super(),e==null&&(e={}),e.restoreBestWeights)throw new Oe("restoreBestWeights = True is not implemented in EarlyStopping yet.");this.monitor=e.monitor||"val_loss",this.minDelta=Math.abs(e.minDelta||0),this.patience=e.patience||0,this.verbose=e.verbose||0,this.mode=e.mode||"auto",this.baseline=e.baseline,["auto","min","max"].indexOf(this.mode)===-1&&(console.warn(`EarlyStopping mode '${this.mode}' is invalid. Falling back to mode 'auto'.`),this.mode="auto"),this.mode==="min"?this.monitorFunc=Sh:this.mode==="max"?this.monitorFunc=$k:this.monitor.indexOf("acc")!==-1?this.monitorFunc=$k:this.monitorFunc=Sh,this.monitorFunc===Sh&&(this.minDelta*=-1)}async onTrainBegin(e){this.wait=0,this.stoppedEpoch=0,this.baseline!=null?this.best=this.baseline:this.best=this.monitorFunc===Sh?1/0:-1/0}async onEpochEnd(e,t){await Zr(t);let n=this.getMonitorValue(t);n!=null&&(this.monitorFunc(n-this.minDelta,this.best)?(this.best=n,this.wait=0):(this.wait++,this.wait>=this.patience&&(this.stoppedEpoch=e,this.model.stopTraining=!0)))}async onTrainEnd(e){this.stoppedEpoch>0&&this.verbose&&console.log(`Epoch ${this.stoppedEpoch}: early stopping.`)}getMonitorValue(e){e==null&&(e={});let t=e[this.monitor];return t==null&&console.warn(`Metric for EarlyStopping ${this.monitor} is not available. Available metrics are: ${Object.keys(e)}`),t}};function k6(e){return new WN(e)}var I6={earlyStopping:k6},S6=X();S6.registerFlag("KEEP_INTERMEDIATE_TENSORS",()=>!1,e=>{e&&console.warn("Keep intermediate tensors is ON. This will print the values of all intermediate tensors during model inference. Not all models support this mode. For details, check e2e/benchmarks/ model_config.js. This significantly impacts performance.")});var Ia;(function(e){e[e.DT_INVALID=0]="DT_INVALID",e[e.DT_FLOAT=1]="DT_FLOAT",e[e.DT_DOUBLE=2]="DT_DOUBLE",e[e.DT_INT32=3]="DT_INT32",e[e.DT_UINT8=4]="DT_UINT8",e[e.DT_INT16=5]="DT_INT16",e[e.DT_INT8=6]="DT_INT8",e[e.DT_STRING=7]="DT_STRING",e[e.DT_COMPLEX64=8]="DT_COMPLEX64",e[e.DT_INT64=9]="DT_INT64",e[e.DT_BOOL=10]="DT_BOOL",e[e.DT_QINT8=11]="DT_QINT8",e[e.DT_QUINT8=12]="DT_QUINT8",e[e.DT_QINT32=13]="DT_QINT32",e[e.DT_BFLOAT16=14]="DT_BFLOAT16",e[e.DT_QINT16=15]="DT_QINT16",e[e.DT_QUINT16=16]="DT_QUINT16",e[e.DT_UINT16=17]="DT_UINT16",e[e.DT_COMPLEX128=18]="DT_COMPLEX128",e[e.DT_HALF=19]="DT_HALF",e[e.DT_RESOURCE=20]="DT_RESOURCE",e[e.DT_VARIANT=21]="DT_VARIANT",e[e.DT_UINT32=22]="DT_UINT32",e[e.DT_UINT64=23]="DT_UINT64",e[e.DT_FLOAT_REF=101]="DT_FLOAT_REF",e[e.DT_DOUBLE_REF=102]="DT_DOUBLE_REF",e[e.DT_INT32_REF=103]="DT_INT32_REF",e[e.DT_UINT8_REF=104]="DT_UINT8_REF",e[e.DT_INT16_REF=105]="DT_INT16_REF",e[e.DT_INT8_REF=106]="DT_INT8_REF",e[e.DT_STRING_REF=107]="DT_STRING_REF",e[e.DT_COMPLEX64_REF=108]="DT_COMPLEX64_REF",e[e.DT_INT64_REF=109]="DT_INT64_REF",e[e.DT_BOOL_REF=110]="DT_BOOL_REF",e[e.DT_QINT8_REF=111]="DT_QINT8_REF",e[e.DT_QUINT8_REF=112]="DT_QUINT8_REF",e[e.DT_QINT32_REF=113]="DT_QINT32_REF",e[e.DT_BFLOAT16_REF=114]="DT_BFLOAT16_REF",e[e.DT_QINT16_REF=115]="DT_QINT16_REF",e[e.DT_QUINT16_REF=116]="DT_QUINT16_REF",e[e.DT_UINT16_REF=117]="DT_UINT16_REF",e[e.DT_COMPLEX128_REF=118]="DT_COMPLEX128_REF",e[e.DT_HALF_REF=119]="DT_HALF_REF",e[e.DT_RESOURCE_REF=120]="DT_RESOURCE_REF",e[e.DT_VARIANT_REF=121]="DT_VARIANT_REF",e[e.DT_UINT32_REF=122]="DT_UINT32_REF",e[e.DT_UINT64_REF=123]="DT_UINT64_REF"})(Ia||(Ia={}));var Dk;(function(e){let t;(function(n){n[n.LEGACY=0]="LEGACY",n[n.V1=1]="V1",n[n.V2=2]="V2"})(t=e.CheckpointFormatVersion||(e.CheckpointFormatVersion={}))})(Dk||(Dk={}));var h0={};function N6(e,t){let n={tfOpName:e,category:"custom",inputs:[],attrs:[],customExecutor:t};h0[e]=n}function VN(e){return h0[e]}function T6(e){delete h0[e]}function I(e,t,n,a,r){let s=t.inputParams[e];if(s&&s.inputIndexStart!==void 0){let o=s.inputIndexStart,l=s.inputIndexEnd===0?void 0:s.inputIndexEnd===void 0?o+1:s.inputIndexEnd;if(s.type==="tensor")return Sn(t.inputNames[s.inputIndexStart],n,a,r);if(s.type==="tensors")return t.inputNames.slice(o,l).map(d=>Sn(d,n,a,r));let u=Sn(t.inputNames.slice(o)[0],n,a,r),p=u.dataSync();return s.type==="number"?p[0]:w.toNestedArray(u.shape,p)}let i=t.attrParams[e];return i&&i.value}function Sn(e,t,n,a){let[r,s]=Xn(e);if(a!=null){let o=a.getHashTableHandleByName(r);if(o!=null)return o}let i=n.currentContextIds.find(o=>!!t[um(r,o)]);return i!==void 0?t[um(r,i)][s]:void 0}function _6(e,t,n){return t[um(e,n.currentContextId)]}function lr(e,t){let[n,a,r]=Xn(e);return[um(n,t&&t.currentContextId),a,r]}function um(e,t){return t?`${e}-${t}`:e}function Xn(e){let t=e.split(":");if(t.length===1)return[e,0,void 0];let n=t[0],a=t.length===3?t[1]:void 0,r=Number(t[t.length-1]);return[n,r,a]}function Dh(e,t,n){let a=I("pad",e,t,n);if(a==="explicit"){a=I("explicitPaddings",e,t,n);let r=[[0,0],[0,0],[0,0],[0,0]];for(let s=0;s<4;s++)r[s][0]=a[s*2],r[s][1]=a[s*2+1];return r}return a}function Nr(e){return e.kept?e:Er(e)}var UN={};Me(UN,{json:()=>C6});var C6=[{tfOpName:"Add",category:"arithmetic",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"AddV2",category:"arithmetic",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"AddN",category:"arithmetic",inputs:[{start:0,end:0,name:"tensors",type:"tensors"}]},{tfOpName:"BiasAdd",category:"arithmetic",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0},{tfName:"data_format",name:"dataFormat",type:"string",notSupported:!0}]},{tfOpName:"Sub",category:"arithmetic",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"RealDiv",category:"arithmetic",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Div",category:"arithmetic",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"DivNoNan",category:"arithmetic",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"FloorDiv",category:"arithmetic",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Mul",category:"arithmetic",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Maximum",category:"arithmetic",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Minimum",category:"arithmetic",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Pow",category:"arithmetic",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"SquaredDifference",category:"arithmetic",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Mod",category:"arithmetic",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"FloorMod",category:"arithmetic",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]}],GN={};Me(GN,{json:()=>E6});var E6=[{tfOpName:"Abs",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Acos",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Asin",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Atan",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Atan2",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"y",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Ceil",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"ClipByValue",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"clipValueMin",type:"number"},{start:2,name:"clipValueMax",type:"number"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Complex",category:"basic_math",inputs:[{start:0,name:"real",type:"tensor"},{start:1,name:"imag",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"ComplexAbs",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Cos",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Cosh",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Elu",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Exp",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Floor",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Log",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Imag",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0},{tfName:"Tout",name:"outputType",type:"dtype",notSupported:!0}]},{tfOpName:"Neg",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Real",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0},{tfName:"Tout",name:"outputType",type:"dtype",notSupported:!0}]},{tfOpName:"Prelu",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"alpha",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Relu",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Relu6",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Selu",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Sigmoid",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Sin",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Sinh",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Sqrt",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Rsqrt",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Square",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Tan",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Tanh",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Sign",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Round",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Expm1",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Log1p",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Reciprocal",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Softplus",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Asinh",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Acosh",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Atanh",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Erf",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Prod",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"axes",type:"number[]"}],attrs:[{tfName:"keep_dims",name:"keepDims",type:"bool",notSupported:!0},{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"LeakyRelu",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"alpha",name:"alpha",type:"number",defaultValue:.2},{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"IsNan",category:"basic_math",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]}],HN={};Me(HN,{json:()=>F6});var F6=[{tfOpName:"EmptyTensorList",category:"control",inputs:[{start:0,name:"elementShape",type:"shape"},{start:1,name:"maxNumElements",type:"number"}],attrs:[{tfName:"element_dtype",name:"elementDType",type:"dtype"}]},{tfOpName:"LoopCond",category:"control",inputs:[{start:0,name:"pred",type:"tensor"}]},{tfOpName:"Switch",category:"control",inputs:[{start:0,name:"data",type:"tensor"},{start:1,name:"pred",type:"tensor"}]},{tfOpName:"Merge",category:"control",inputs:[{start:0,end:0,name:"tensors",type:"tensors"}]},{tfOpName:"Enter",category:"control",inputs:[{start:0,name:"tensor",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0},{tfName:"frame_name",name:"frameName",type:"string"},{tfName:"is_constant",name:"isConstant",type:"bool"}]},{tfOpName:"Exit",category:"control",inputs:[{start:0,name:"tensor",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"NextIteration",category:"control",inputs:[{start:0,name:"tensor",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"TensorArrayV3",category:"control",inputs:[{start:0,name:"size",type:"number"}],attrs:[{tfName:"dtype",name:"dtype",type:"dtype"},{tfName:"element_shape",name:"elementShape",type:"shape"},{tfName:"dynamic_size",name:"dynamicSize",type:"bool"},{tfName:"clear_after_read",name:"clearAfterRead",type:"bool"},{tfName:"identical_element_shapes",name:"identicalElementShapes",type:"bool"},{tfName:"tensor_array_name",name:"name",type:"string"}]},{tfOpName:"TensorArrayWriteV3",category:"control",inputs:[{start:0,name:"tensorArrayId",type:"tensor"},{start:1,name:"index",type:"number"},{start:2,name:"tensor",type:"tensor"},{start:3,name:"flowIn",type:"number"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"TensorArrayReadV3",category:"control",inputs:[{start:0,name:"tensorArrayId",type:"tensor"},{start:1,name:"index",type:"number"},{start:2,name:"flowIn",type:"number"}],attrs:[{tfName:"dtype",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"TensorArrayGatherV3",category:"control",inputs:[{start:0,name:"tensorArrayId",type:"tensor"},{start:1,name:"indices",type:"number[]"},{start:2,name:"flowIn",type:"number"}],attrs:[{tfName:"dtype",name:"dtype",type:"dtype"},{tfName:"element_shape",name:"elementShape",type:"shape"}]},{tfOpName:"TensorArrayScatterV3",category:"control",inputs:[{start:0,name:"tensorArrayId",type:"tensor"},{start:1,name:"indices",type:"number[]"},{start:2,name:"tensor",type:"tensor"},{start:3,name:"flowIn",type:"number"}],attrs:[{tfName:"T",name:"dtype",type:"dtype"}]},{tfOpName:"TensorArrayConcatV3",category:"control",inputs:[{start:0,name:"tensorArrayId",type:"tensor"},{start:1,name:"flowIn",type:"number"}],attrs:[{tfName:"dtype",name:"dtype",type:"dtype"},{tfName:"element_shape_except0",name:"elementShapeExcept0",type:"shape",notSupported:!0}]},{tfOpName:"TensorArraySplitV3",category:"control",inputs:[{start:0,name:"tensorArrayId",type:"tensor"},{start:1,name:"tensor",type:"tensor"},{start:2,name:"lengths",type:"number[]"},{start:3,name:"flowIn",type:"number"}],attrs:[{tfName:"T",name:"dtype",type:"dtype"}]},{tfOpName:"TensorArraySizeV3",category:"control",inputs:[{start:0,name:"tensorArrayId",type:"tensor"},{start:1,name:"flowIn",type:"number"}]},{tfOpName:"TensorArrayCloseV3",category:"control",inputs:[{start:0,name:"tensorArrayId",type:"tensor"}]},{tfOpName:"StatelessIf",category:"control",inputs:[{start:0,name:"cond",type:"tensor"},{start:1,end:0,name:"args",type:"tensors"}],attrs:[{tfName:"then_branch",name:"thenBranch",type:"func"},{tfName:"else_branch",name:"elseBranch",type:"func"}]},{tfOpName:"If",category:"control",inputs:[{start:0,name:"cond",type:"tensor"},{start:1,end:0,name:"args",type:"tensors"}],attrs:[{tfName:"then_branch",name:"thenBranch",type:"func"},{tfName:"else_branch",name:"elseBranch",type:"func"}]},{tfOpName:"StatelessWhile",category:"control",inputs:[{start:0,end:0,name:"args",type:"tensors"}],attrs:[{tfName:"cond",name:"cond",type:"func"},{tfName:"body",name:"body",type:"func"}]},{tfOpName:"While",category:"control",inputs:[{start:0,end:0,name:"args",type:"tensors"}],attrs:[{tfName:"cond",name:"cond",type:"func"},{tfName:"body",name:"body",type:"func"}]},{tfOpName:"TensorListScatter",category:"control",inputs:[{start:0,name:"tensor",type:"tensor"},{start:1,name:"indices",type:"number[]"},{start:2,name:"elementShape",type:"shape"}],attrs:[{tfName:"element_dtype",name:"elementDType",type:"dtype"}]},{tfOpName:"TensorListScatterV2",category:"control",inputs:[{start:0,name:"tensor",type:"tensor"},{start:1,name:"indices",type:"number[]"},{start:2,name:"elementShape",type:"shape"},{start:3,name:"numElements",type:"number"}],attrs:[{tfName:"element_dtype",name:"elementDType",type:"dtype"}]},{tfOpName:"TensorListGather",category:"control",inputs:[{start:0,name:"tensorListId",type:"tensor"},{start:1,name:"indices",type:"number[]"},{start:2,name:"elementShape",type:"shape"}],attrs:[{tfName:"element_dtype",name:"elementDType",type:"dtype"}]},{tfOpName:"TensorListGetItem",category:"control",inputs:[{start:0,name:"tensorListId",type:"tensor"},{start:1,name:"index",type:"number"},{start:2,name:"elementShape",type:"shape"}],attrs:[{tfName:"element_dtype",name:"elementDType",type:"dtype"}]},{tfOpName:"TensorListSetItem",category:"control",inputs:[{start:0,name:"tensorListId",type:"tensor"},{start:1,name:"index",type:"number"},{start:2,name:"tensor",type:"tensor"}],attrs:[{tfName:"element_dtype",name:"elementDType",type:"dtype"}]},{tfOpName:"TensorListReserve",category:"control",inputs:[{start:0,name:"elementShape",type:"shape"},{start:1,name:"numElements",type:"number"}],attrs:[{tfName:"element_dtype",name:"elementDType",type:"dtype"}]},{tfOpName:"TensorListFromTensor",category:"control",inputs:[{start:0,name:"tensor",type:"tensor"},{start:1,name:"elementShape",type:"shape"}],attrs:[{tfName:"element_dtype",name:"elementDType",type:"dtype"}]},{tfOpName:"TensorListStack",category:"control",inputs:[{start:0,name:"tensorListId",type:"tensor"},{start:1,name:"elementShape",type:"shape"}],attrs:[{tfName:"element_dtype",name:"elementDType",type:"dtype"},{tfName:"num_elements",name:"numElements",type:"dtype"}]},{tfOpName:"TensorListSplit",category:"control",inputs:[{start:0,name:"tensor",type:"tensor"},{start:1,name:"elementShape",type:"shape"},{start:2,name:"lengths",type:"number[]"}],attrs:[{tfName:"element_dtype",name:"elementDType",type:"dtype"}]},{tfOpName:"TensorListConcat",category:"control",inputs:[{start:0,name:"tensorListId",type:"tensor"}],attrs:[{tfName:"element_shape",name:"elementShape",type:"shape"},{tfName:"element_dtype",name:"elementDType",type:"dtype"}]},{tfOpName:"TensorListConcatV2",category:"control",inputs:[{start:0,name:"tensorListId",type:"tensor"}],attrs:[{tfName:"element_shape",name:"elementShape",type:"shape"},{tfName:"element_dtype",name:"elementDType",type:"dtype"}]},{tfOpName:"TensorListPopBack",category:"control",inputs:[{start:0,name:"tensorListId",type:"tensor"},{start:1,name:"elementShape",type:"shape"}],attrs:[{tfName:"element_dtype",name:"elementDType",type:"dtype"}]},{tfOpName:"TensorListPushBack",category:"control",inputs:[{start:0,name:"tensorListId",type:"tensor"},{start:1,name:"tensor",type:"tensor"}],attrs:[{tfName:"element_dtype",name:"elementDType",type:"dtype"}]},{tfOpName:"TensorListLength",category:"control",inputs:[{start:0,name:"tensorListId",type:"tensor"}]},{tfOpName:"TensorListResize",category:"control",inputs:[{start:0,name:"tensorListId",type:"tensor"},{start:1,name:"size",type:"number"}]}],jN={};Me(jN,{json:()=>A6});var A6=[{tfOpName:"AvgPool",category:"convolution",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"strides",name:"strides",type:"number[]"},{tfName:"padding",name:"pad",type:"string"},{tfName:"data_format",name:"dataFormat",type:"string",notSupported:!0},{tfName:"ksize",name:"kernelSize",type:"number[]"},{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"MaxPool",category:"convolution",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"strides",name:"strides",type:"number[]"},{tfName:"padding",name:"pad",type:"string"},{tfName:"data_format",name:"dataFormat",type:"string",notSupported:!0},{tfName:"ksize",name:"kernelSize",type:"number[]"},{tfName:"explicit_paddings",name:"explicitPaddings",type:"number[]",defaultValue:[],notSupported:!0},{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"MaxPoolWithArgmax",category:"convolution",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"strides",name:"strides",type:"number[]"},{tfName:"padding",name:"pad",type:"string"},{tfName:"ksize",name:"kernelSize",type:"number[]"},{tfName:"include_batch_in_index",name:"includeBatchInIndex",type:"bool"},{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"AvgPool3D",category:"convolution",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"strides",name:"strides",type:"number[]"},{tfName:"padding",name:"pad",type:"string"},{tfName:"data_format",name:"dataFormat",type:"string",notSupported:!0},{tfName:"ksize",name:"kernelSize",type:"number[]"},{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"MaxPool3D",category:"convolution",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"strides",name:"strides",type:"number[]"},{tfName:"padding",name:"pad",type:"string"},{tfName:"data_format",name:"dataFormat",type:"string",notSupported:!0},{tfName:"ksize",name:"kernelSize",type:"number[]"},{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Conv1D",category:"convolution",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"filter",type:"tensor"}],attrs:[{tfName:"stride",name:"stride",type:"number"},{tfName:"padding",name:"pad",type:"string"},{tfName:"data_format",name:"dataFormat",type:"string",defaultValue:"NWC"},{tfName:"T",name:"dtype",type:"dtype",notSupported:!0},{tfName:"dilation",name:"dilation",type:"number",defaultValue:1}]},{tfOpName:"Conv2D",category:"convolution",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"filter",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0},{tfName:"strides",name:"strides",type:"number[]"},{tfName:"padding",name:"pad",type:"string"},{tfName:"useCudnnOnGpu",name:"useCudnnOnGpu",type:"bool"},{tfName:"data_format",name:"dataFormat",type:"string",defaultValue:"NHWC"},{tfName:"explicit_paddings",name:"explicitPaddings",type:"number[]",defaultValue:[]},{tfName:"dilations",name:"dilations",type:"number[]"}]},{tfOpName:"_FusedConv2D",category:"convolution",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"filter",type:"tensor"},{start:2,end:0,name:"args",type:"tensors"}],attrs:[{tfName:"num_args",name:"numArgs",type:"number"},{tfName:"T",name:"dtype",type:"dtype",notSupported:!0},{tfName:"strides",name:"strides",type:"number[]"},{tfName:"padding",name:"pad",type:"string"},{tfName:"explicit_paddings",name:"explicitPaddings",type:"number[]",defaultValue:[]},{tfName:"use_cudnn_on_gpu",name:"useCudnnOnGpu",type:"bool",defaultValue:!0},{tfName:"data_format",name:"dataFormat",type:"string",defaultValue:"NHWC"},{tfName:"dilations",name:"dilations",type:"number[]",defaultValue:[1,1,1,1]},{tfName:"fused_ops",name:"fusedOps",type:"string[]",defaultValue:[]},{tfName:"epsilon",name:"epsilon",type:"number",defaultValue:1e-4},{tfName:"leakyrelu_alpha",name:"leakyreluAlpha",type:"number"}]},{tfOpName:"Conv2DBackpropInput",category:"convolution",inputs:[{start:2,name:"x",type:"tensor"},{start:1,name:"filter",type:"tensor"},{start:0,name:"outputShape",type:"number[]"}],attrs:[{tfName:"strides",name:"strides",type:"number[]"},{tfName:"padding",name:"pad",type:"string"},{tfName:"data_format",name:"dataFormat",type:"string",notSupported:!0},{tfName:"explicit_paddings",name:"explicitPaddings",type:"number[]",defaultValue:[]},{tfName:"dilations",name:"dilations",type:"number[]",notSupported:!0}]},{tfOpName:"DepthwiseConv2d",category:"convolution",inputs:[{start:0,name:"input",type:"tensor"},{start:1,name:"filter",type:"tensor"}],attrs:[{tfName:"strides",name:"strides",type:"number[]"},{tfName:"padding",name:"pad",type:"string"},{tfName:"data_format",name:"dataFormat",type:"string",defaultValue:"NHWC"},{tfName:"explicit_paddings",name:"explicitPaddings",type:"number[]",defaultValue:[]},{tfName:"dilations",name:"dilations",type:"number[]"}]},{tfOpName:"DepthwiseConv2dNative",category:"convolution",inputs:[{start:0,name:"input",type:"tensor"},{start:1,name:"filter",type:"tensor"}],attrs:[{tfName:"strides",name:"strides",type:"number[]"},{tfName:"padding",name:"pad",type:"string"},{tfName:"data_format",name:"dataFormat",type:"string",defaultValue:"NHWC"},{tfName:"explicit_paddings",name:"explicitPaddings",type:"number[]",defaultValue:[]},{tfName:"dilations",name:"dilations",type:"number[]"}]},{tfOpName:"FusedDepthwiseConv2dNative",category:"convolution",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"filter",type:"tensor"},{start:2,end:0,name:"args",type:"tensors"}],attrs:[{tfName:"num_args",name:"numArgs",type:"number"},{tfName:"T",name:"dtype",type:"dtype",notSupported:!0},{tfName:"strides",name:"strides",type:"number[]"},{tfName:"padding",name:"pad",type:"string"},{tfName:"data_format",name:"dataFormat",type:"string",defaultValue:"NHWC"},{tfName:"dilations",name:"dilations",type:"number[]",defaultValue:[1,1,1,1]},{tfName:"fused_ops",name:"fusedOps",type:"string[]",defaultValue:[]},{tfName:"explicit_paddings",name:"explicitPaddings",type:"number[]",defaultValue:[]}]},{tfOpName:"Conv3D",category:"convolution",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"filter",type:"tensor"}],attrs:[{tfName:"strides",name:"strides",type:"number[]"},{tfName:"padding",name:"pad",type:"string"},{tfName:"data_format",name:"dataFormat",type:"string",defaultValue:"NHWC"},{tfName:"dilations",name:"dilations",type:"number[]"}]},{tfOpName:"Dilation2D",category:"convolution",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"filter",type:"tensor"}],attrs:[{tfName:"strides",name:"strides",type:"number[]"},{tfName:"rates",name:"dilations",type:"number[]"},{tfName:"padding",name:"pad",type:"string"}]}],qN={};Me(qN,{json:()=>$6});var $6=[{tfOpName:"Fill",category:"creation",inputs:[{start:0,name:"shape",type:"number[]"},{start:1,name:"value",type:"number"}],attrs:[{tfName:"T",name:"dtype",type:"dtype"}]},{tfOpName:"LinSpace",category:"creation",inputs:[{start:0,name:"start",type:"number"},{start:1,name:"stop",type:"number"},{start:2,name:"num",type:"number"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"OneHot",category:"creation",inputs:[{start:0,name:"indices",type:"tensor"},{start:1,name:"depth",type:"number"},{start:2,name:"onValue",type:"number",defaultValue:1},{start:3,name:"offValue",type:"number",defaultValue:0}],attrs:[{tfName:"axis",name:"axis",type:"number",notSupported:!0},{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Ones",category:"creation",inputs:[{start:0,name:"shape",type:"number[]"}],attrs:[{tfName:"T",name:"dtype",type:"dtype"}]},{tfOpName:"OnesLike",category:"creation",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"dtype",name:"dtype",type:"dtype"}]},{tfOpName:"RandomUniform",category:"creation",inputs:[{start:0,name:"shape",type:"number[]"}],attrs:[{tfName:"minval",name:"minval",type:"number",defaultValue:0},{tfName:"maxval",name:"maxval",type:"number",defaultValue:1},{tfName:"dtype",name:"dtype",type:"dtype"},{tfName:"seed",name:"seed",type:"number",defaultValue:0},{tfName:"seed2",name:"seed2",type:"number",defaultValue:0,notSupported:!0},{tfName:"T",name:"T",type:"number",notSupported:!0}]},{tfOpName:"Range",category:"creation",inputs:[{start:0,name:"start",type:"number"},{start:1,name:"stop",type:"number"},{start:2,name:"step",type:"number",defaultValue:0}],attrs:[{tfName:"Tidx",name:"dtype",type:"dtype"}]},{tfOpName:"TruncatedNormal",category:"creation",inputs:[{start:0,name:"shape",type:"number[]"}],attrs:[{tfName:"means",name:"mean",type:"number",defaultValue:0},{tfName:"stddev",name:"stdDev",type:"number",defaultValue:1},{tfName:"seed",name:"seed",type:"number"},{tfName:"seed2",name:"seed2",type:"number",defaultValue:0,notSupported:!0},{tfName:"dtype",name:"dtype",type:"dtype"},{tfName:"T",name:"T",type:"number",notSupported:!0}]},{tfOpName:"Zeros",category:"creation",inputs:[{start:0,name:"shape",type:"number[]"}],attrs:[{tfName:"T",name:"dtype",type:"dtype"}]},{tfOpName:"ZerosLike",category:"creation",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype"}]},{tfOpName:"Multinomial",category:"creation",inputs:[{start:0,name:"logits",type:"tensor"},{start:1,name:"numSamples",type:"number"}],attrs:[{tfName:"seed",name:"seed",type:"number"},{tfName:"seed2",name:"seed2",type:"number"},{tfName:"T",name:"dtype",type:"dtype"},{tfName:"output_dtype",name:"output_dtype",type:"dtype"}]}],KN={};Me(KN,{json:()=>D6});var D6=[{tfOpName:"NonMaxSuppressionV2",category:"dynamic",inputs:[{start:0,name:"boxes",type:"tensor"},{start:1,name:"scores",type:"tensor"},{start:2,name:"maxOutputSize",type:"number"},{start:3,name:"iouThreshold",type:"number"}]},{tfOpName:"NonMaxSuppressionV3",category:"dynamic",inputs:[{start:0,name:"boxes",type:"tensor"},{start:1,name:"scores",type:"tensor"},{start:2,name:"maxOutputSize",type:"number"},{start:3,name:"iouThreshold",type:"number"},{start:4,name:"scoreThreshold",type:"number"}]},{tfOpName:"NonMaxSuppressionV4",category:"dynamic",inputs:[{start:0,name:"boxes",type:"tensor"},{start:1,name:"scores",type:"tensor"},{start:2,name:"maxOutputSize",type:"number"},{start:3,name:"iouThreshold",type:"number"},{start:4,name:"scoreThreshold",type:"number"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0},{tfName:"T_threshold",name:"threshold",type:"dtype",notSupported:!0},{tfName:"pad_to_max_output_size",name:"padToMaxOutputSize",type:"bool"}]},{tfOpName:"NonMaxSuppressionV5",category:"dynamic",inputs:[{start:0,name:"boxes",type:"tensor"},{start:1,name:"scores",type:"tensor"},{start:2,name:"maxOutputSize",type:"number"},{start:3,name:"iouThreshold",type:"number"},{start:4,name:"scoreThreshold",type:"number"},{start:5,name:"softNmsSigma",type:"number"}]},{tfOpName:"Where",category:"dynamic",inputs:[{start:0,name:"condition",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"ListDiff",category:"dynamic",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"y",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]}],XN={};Me(XN,{json:()=>R6});var R6=[{tfOpName:"LowerBound",category:"evaluation",inputs:[{start:0,name:"sortedSequence",type:"tensor"},{start:1,name:"values",type:"tensor"}]},{tfOpName:"TopKV2",category:"evaluation",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"k",type:"number"}],attrs:[{tfName:"sorted",name:"sorted",type:"bool"}]},{tfOpName:"UpperBound",category:"evaluation",inputs:[{start:0,name:"sortedSequence",type:"tensor"},{start:1,name:"values",type:"tensor"}]},{tfOpName:"Unique",category:"evaluation",inputs:[{start:0,name:"x",type:"tensor"}]},{tfOpName:"UniqueV2",category:"evaluation",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"axis",type:"number"}]}],YN={};Me(YN,{json:()=>M6});var M6=[{tfOpName:"PlaceholderWithDefault",category:"graph",inputs:[{start:0,name:"default",type:"tensor"}],attrs:[{tfName:"shape",name:"shape",type:"shape"},{tfName:"dtype",name:"dtype",type:"dtype"}]},{tfOpName:"Placeholder",category:"graph",attrs:[{tfName:"shape",name:"shape",type:"shape"},{tfName:"dtype",name:"dtype",type:"dtype"}]},{tfOpName:"Const",category:"graph"},{tfOpName:"Identity",category:"graph",inputs:[{start:0,name:"x",type:"tensor"}]},{tfOpName:"IdentityN",category:"graph",inputs:[{start:0,end:0,name:"x",type:"tensors"}]},{tfOpName:"Snapshot",category:"graph",inputs:[{start:0,name:"x",type:"tensor"}]},{tfOpName:"Rank",category:"graph",inputs:[{start:0,name:"x",type:"tensor"}]},{tfOpName:"Size",category:"graph",inputs:[{start:0,name:"x",type:"tensor"}]},{tfOpName:"Shape",category:"graph",inputs:[{start:0,name:"x",type:"tensor"}]},{tfOpName:"ShapeN",category:"graph",inputs:[{start:0,end:0,name:"x",type:"tensors"}]},{tfOpName:"Print",category:"graph",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"data",type:"tensors"}],attrs:[{tfName:"message",name:"message",type:"string"},{tfName:"first_n",name:"firstN",type:"number",notSupported:!0},{tfName:"summarize",name:"summarize",type:"number",defaultValue:3}]},{tfOpName:"NoOp",category:"graph",inputs:[]},{tfOpName:"StopGradient",category:"graph",inputs:[{start:0,name:"x",type:"tensor"}]},{tfOpName:"FakeQuantWithMinMaxVars",category:"graph",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"min",name:"min",type:"number"},{tfName:"max",name:"max",type:"number"}]}],QN={};Me(QN,{json:()=>P6});var P6=[{tfOpName:"HashTable",category:"hash_table",inputs:[],attrs:[{tfName:"shared_name",name:"sharedName",type:"string"},{tfName:"use_node_name_sharing",name:"useNodeNameSharing",type:"bool"},{tfName:"key_dtype",name:"keyDType",type:"dtype"},{tfName:"value_dtype",name:"valueDType",type:"dtype"}]},{tfOpName:"HashTableV2",category:"hash_table",inputs:[],attrs:[{tfName:"shared_name",name:"sharedName",type:"string"},{tfName:"use_node_name_sharing",name:"useNodeNameSharing",type:"bool"},{tfName:"key_dtype",name:"keyDType",type:"dtype"},{tfName:"value_dtype",name:"valueDType",type:"dtype"}]},{tfOpName:"LookupTableImport",category:"hash_table",inputs:[{start:0,name:"tableHandle",type:"tensor"},{start:1,name:"keys",type:"tensor"},{start:2,name:"values",type:"tensor"}],attrs:[{tfName:"Tin",name:"tIn",type:"dtype",notSupported:!0},{tfName:"Tout",name:"tOut",type:"dtype",notSupported:!0}]},{tfOpName:"LookupTableImportV2",category:"hash_table",inputs:[{start:0,name:"tableHandle",type:"tensor"},{start:1,name:"keys",type:"tensor"},{start:2,name:"values",type:"tensor"}],attrs:[{tfName:"Tin",name:"tIn",type:"dtype",notSupported:!0},{tfName:"Tout",name:"tOut",type:"dtype",notSupported:!0}]},{tfOpName:"LookupTableFind",category:"hash_table",inputs:[{start:0,name:"tableHandle",type:"tensor"},{start:1,name:"keys",type:"tensor"},{start:2,name:"defaultValue",type:"tensor"}],attrs:[{tfName:"Tin",name:"tIn",type:"dtype",notSupported:!0},{tfName:"Tout",name:"tOut",type:"dtype",notSupported:!0}]},{tfOpName:"LookupTableFindV2",category:"hash_table",inputs:[{start:0,name:"tableHandle",type:"tensor"},{start:1,name:"keys",type:"tensor"},{start:2,name:"defaultValue",type:"tensor"}],attrs:[{tfName:"Tin",name:"tIn",type:"dtype",notSupported:!0},{tfName:"Tout",name:"tOut",type:"dtype",notSupported:!0}]},{tfOpName:"LookupTableSize",category:"hash_table",inputs:[{start:0,name:"tableHandle",type:"tensor"}]},{tfOpName:"LookupTableSizeV2",category:"hash_table",inputs:[{start:0,name:"tableHandle",type:"tensor"}]}],JN={};Me(JN,{json:()=>O6});var O6=[{tfOpName:"ResizeBilinear",category:"image",inputs:[{start:0,name:"images",type:"tensor"},{start:1,name:"size",type:"number[]"}],attrs:[{tfName:"align_corners",name:"alignCorners",type:"bool"},{tfName:"half_pixel_centers",name:"halfPixelCenters",type:"bool"},{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"ResizeNearestNeighbor",category:"image",inputs:[{start:0,name:"images",type:"tensor"},{start:1,name:"size",type:"number[]"}],attrs:[{tfName:"align_corners",name:"alignCorners",type:"bool"},{tfName:"half_pixel_centers",name:"halfPixelCenters",type:"bool"},{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"CropAndResize",category:"image",inputs:[{start:0,name:"image",type:"tensor"},{start:1,name:"boxes",type:"tensor"},{start:2,name:"boxInd",type:"tensor"},{start:3,name:"cropSize",type:"number[]"}],attrs:[{tfName:"method",name:"method",type:"string"},{tfName:"extrapolation_value",name:"extrapolationValue",type:"number"}]},{tfOpName:"ImageProjectiveTransformV3",category:"image",inputs:[{start:0,name:"images",type:"tensor"},{start:1,name:"transforms",type:"tensor"},{start:2,name:"outputShape",type:"number[]"},{start:3,name:"fillValue",type:"number"}],attrs:[{tfName:"interpolation",name:"interpolation",type:"string"},{tfName:"fill_mode",name:"fillMode",type:"string"}]}],ZN={};Me(ZN,{json:()=>L6});var L6=[{tfOpName:"Equal",category:"logical",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"NotEqual",category:"logical",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Greater",category:"logical",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"GreaterEqual",category:"logical",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Less",category:"logical",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"LessEqual",category:"logical",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"LogicalAnd",category:"logical",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"LogicalNot",category:"logical",inputs:[{start:0,name:"a",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"LogicalOr",category:"logical",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Select",category:"logical",inputs:[{start:0,name:"condition",type:"tensor"},{start:1,name:"a",type:"tensor"},{start:2,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"SelectV2",category:"logical",inputs:[{start:0,name:"condition",type:"tensor"},{start:1,name:"a",type:"tensor"},{start:2,name:"b",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]}],eT={};Me(eT,{json:()=>z6});var z6=[{tfOpName:"_FusedMatMul",category:"matrices",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"},{start:2,end:0,name:"args",type:"tensors"}],attrs:[{tfName:"num_args",name:"numArgs",type:"number"},{tfName:"fused_ops",name:"fusedOps",type:"string[]",defaultValue:[]},{tfName:"epsilon",name:"epsilon",type:"number",defaultValue:1e-4},{tfName:"transpose_a",name:"transposeA",type:"bool",defaultValue:!1},{tfName:"transpose_b",name:"transposeB",type:"bool",defaultValue:!1},{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"MatMul",category:"matrices",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"transpose_a",name:"transposeA",type:"bool",defaultValue:!1},{tfName:"transpose_b",name:"transposeB",type:"bool",defaultValue:!1},{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"BatchMatMul",category:"matrices",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"adj_x",name:"transposeA",type:"bool",defaultValue:!1},{tfName:"adj_y",name:"transposeB",type:"bool",defaultValue:!1},{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"BatchMatMulV2",category:"matrices",inputs:[{start:0,name:"a",type:"tensor"},{start:1,name:"b",type:"tensor"}],attrs:[{tfName:"adj_x",name:"transposeA",type:"bool",defaultValue:!1},{tfName:"adj_y",name:"transposeB",type:"bool",defaultValue:!1},{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Transpose",category:"matrices",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"perm",type:"number[]"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"Einsum",category:"matrices",inputs:[{start:0,end:0,name:"tensors",type:"tensors"}],attrs:[{tfName:"equation",name:"equation",type:"string"},{tfName:"N",name:"n",type:"number",defaultValue:2},{tfName:"T",name:"dtype",type:"dtype"}]}],tT={};Me(tT,{json:()=>B6});var B6=[{tfOpName:"EuclideanNorm",category:"normalization",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"axis",type:"number[]"}],attrs:[{tfName:"keep_dims",name:"keepDims",type:"bool",defaultValue:!1}]},{tfOpName:"FusedBatchNorm",category:"normalization",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"scale",type:"tensor"},{start:2,name:"offset",type:"tensor"},{start:3,name:"mean",type:"tensor"},{start:4,name:"variance",type:"tensor"}],attrs:[{tfName:"epsilon",name:"epsilon",type:"number",defaultValue:.001},{tfName:"data_format",name:"dataFormat",type:"string",notSupported:!0}]},{tfOpName:"FusedBatchNormV2",category:"normalization",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"scale",type:"tensor"},{start:2,name:"offset",type:"tensor"},{start:3,name:"mean",type:"tensor"},{start:4,name:"variance",type:"tensor"}],attrs:[{tfName:"epsilon",name:"epsilon",type:"number",defaultValue:.001},{tfName:"data_format",name:"dataFormat",type:"string",notSupported:!0}]},{tfOpName:"FusedBatchNormV3",category:"normalization",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"scale",type:"tensor"},{start:2,name:"offset",type:"tensor"},{start:3,name:"mean",type:"tensor"},{start:4,name:"variance",type:"tensor"}],attrs:[{tfName:"epsilon",name:"epsilon",type:"number",defaultValue:.001},{tfName:"data_format",name:"dataFormat",type:"string",notSupported:!0}]},{tfOpName:"LRN",category:"normalization",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"depth_radius",name:"radius",type:"number",defaultValue:5},{tfName:"bias",name:"bias",type:"number",defaultValue:1},{tfName:"alpha",name:"alpha",type:"number",defaultValue:1},{tfName:"beta",name:"beta",type:"number",defaultValue:.5}]},{tfOpName:"Softmax",category:"normalization",inputs:[{start:0,name:"x",type:"tensor"}]},{tfOpName:"LogSoftmax",category:"normalization",inputs:[{start:0,name:"x",type:"tensor"}]},{tfOpName:"SparseToDense",category:"normalization",inputs:[{start:0,name:"sparseIndices",type:"tensor"},{start:1,name:"outputShape",type:"number[]"},{start:2,name:"sparseValues",type:"tensor"},{start:3,name:"defaultValue",type:"tensor"}],attrs:[{tfName:"validate_indices",name:"validateIndices",type:"bool",defaultValue:!0,notSupported:!0}]}],nT={};Me(nT,{json:()=>W6});var W6=[{tfOpName:"Bincount",category:"reduction",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"size",type:"number"},{start:2,name:"weights",type:"tensor"}]},{tfOpName:"DenseBincount",category:"reduction",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"size",type:"number"},{start:2,name:"weights",type:"tensor"}],attrs:[{tfName:"binary_output",name:"binaryOutput",type:"bool"}]},{tfOpName:"Max",category:"reduction",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"axis",type:"number[]"}],attrs:[{tfName:"keep_dims",name:"keepDims",type:"bool"}]},{tfOpName:"Mean",category:"reduction",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"axis",type:"number[]"}],attrs:[{tfName:"keep_dims",name:"keepDims",type:"bool"}]},{tfOpName:"Min",category:"reduction",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"axis",type:"number[]"}],attrs:[{tfName:"keep_dims",name:"keepDims",type:"bool"}]},{tfOpName:"Sum",category:"reduction",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"axis",type:"number[]"}],attrs:[{tfName:"keep_dims",name:"keepDims",type:"bool"}]},{tfOpName:"All",category:"reduction",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"axis",type:"number[]"}],attrs:[{tfName:"keep_dims",name:"keepDims",type:"bool"}]},{tfOpName:"Any",category:"reduction",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"axis",type:"number[]"}],attrs:[{tfName:"keep_dims",name:"keepDims",type:"bool"}]},{tfOpName:"ArgMax",category:"reduction",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"axis",type:"number"}]},{tfOpName:"ArgMin",category:"reduction",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"axis",type:"number"}]},{tfOpName:"Prod",category:"reduction",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"axis",type:"number[]"}],attrs:[{tfName:"keep_dims",name:"keepDims",type:"bool"}]},{tfOpName:"Cumprod",category:"reduction",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"axis",type:"number"}],attrs:[{tfName:"exclusive",name:"exclusive",type:"bool"},{tfName:"reverse",name:"reverse",type:"bool"}]},{tfOpName:"Cumsum",category:"reduction",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"axis",type:"number"}],attrs:[{tfName:"exclusive",name:"exclusive",type:"bool"},{tfName:"reverse",name:"reverse",type:"bool"}]}],aT={};Me(aT,{json:()=>V6});var V6=[{tfOpName:"ConcatV2",category:"slice_join",inputs:[{start:0,end:-1,name:"tensors",type:"tensors"},{start:-1,name:"axis",type:"number"}],attrs:[{tfName:"N",name:"n",type:"number",defaultValue:2}]},{tfOpName:"Concat",category:"slice_join",inputs:[{start:1,end:0,name:"tensors",type:"tensors"},{start:0,name:"axis",type:"number"}],attrs:[{tfName:"N",name:"n",type:"number",defaultValue:2}]},{tfOpName:"GatherV2",category:"slice_join",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"indices",type:"tensor"},{start:2,name:"axis",type:"number",defaultValue:0}],attrs:[{tfName:"batch_dims",name:"batchDims",type:"number",defaultValue:0}]},{tfOpName:"Gather",category:"slice_join",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"indices",type:"tensor"}],attrs:[{tfName:"validate_indices",name:"validateIndices",type:"bool",notSupported:!0}]},{tfOpName:"Reverse",category:"slice_join",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"dims",type:"bool[]"}]},{tfOpName:"ReverseV2",category:"slice_join",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"axis",type:"number[]"}]},{tfOpName:"Slice",category:"slice_join",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"begin",type:"number[]"},{start:2,name:"size",type:"number[]"}]},{tfOpName:"StridedSlice",category:"slice_join",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"begin",type:"number[]"},{start:2,name:"end",type:"number[]"},{start:3,name:"strides",type:"number[]"}],attrs:[{tfName:"begin_mask",name:"beginMask",type:"number",defaultValue:0},{tfName:"end_mask",name:"endMask",type:"number",defaultValue:0},{tfName:"new_axis_mask",name:"newAxisMask",type:"number",defaultValue:0},{tfName:"ellipsis_mask",name:"ellipsisMask",type:"number",defaultValue:0},{tfName:"shrink_axis_mask",name:"shrinkAxisMask",type:"number",defaultValue:0}]},{tfOpName:"Pack",category:"slice_join",inputs:[{start:0,end:0,name:"tensors",type:"tensors"}],attrs:[{tfName:"axis",name:"axis",type:"number",defaultValue:0}]},{tfOpName:"Unpack",category:"slice_join",inputs:[{start:0,name:"tensor",type:"tensor"}],attrs:[{tfName:"axis",name:"axis",type:"number",defaultValue:0},{tfName:"num",name:"num",type:"number",defaultValue:0,notSupported:!0}]},{tfOpName:"Tile",category:"slice_join",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"reps",type:"number[]"}]},{tfOpName:"Split",category:"slice_join",inputs:[{start:0,name:"axis",type:"number",defaultValue:0},{start:1,name:"x",type:"tensor"}],attrs:[{tfName:"num_split",name:"numOrSizeSplits",type:"number",defaultValue:1}]},{tfOpName:"SplitV",category:"slice_join",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"numOrSizeSplits",type:"number[]"},{start:2,name:"axis",type:"number",defaultValue:0}]},{tfOpName:"ScatterNd",category:"slice_join",inputs:[{start:0,name:"indices",type:"tensor"},{start:1,name:"values",type:"tensor"},{start:2,name:"shape",type:"number[]"}]},{tfOpName:"GatherNd",category:"slice_join",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"indices",type:"tensor"}]},{tfOpName:"SparseToDense",category:"slice_join",inputs:[{start:0,name:"sparseIndices",type:"tensor"},{start:1,name:"outputShape",type:"number[]"},{start:2,name:"sparseValues",type:"tensor"},{start:3,name:"defaultValue",type:"tensor"}],attrs:[{tfName:"validate_indices",name:"validateIndices",type:"bool",defaultValue:!1,notSupported:!0}]}],rT={};Me(rT,{json:()=>U6});var U6=[{tfOpName:"SparseFillEmptyRows",category:"sparse",inputs:[{start:0,name:"indices",type:"tensor"},{start:1,name:"values",type:"tensor"},{start:2,name:"denseShape",type:"tensor"},{start:3,name:"defaultValue",type:"tensor"}]},{tfOpName:"SparseReshape",category:"sparse",inputs:[{start:0,name:"inputIndices",type:"tensor"},{start:1,name:"inputShape",type:"tensor"},{start:2,name:"newShape",type:"tensor"}],attrs:[{tfName:"T",name:"dtype",type:"dtype",notSupported:!0}]},{tfOpName:"SparseSegmentMean",category:"sparse",inputs:[{start:0,name:"data",type:"tensor"},{start:1,name:"indices",type:"tensor"},{start:2,name:"segmentIds",type:"tensor"}]},{tfOpName:"SparseSegmentSum",category:"sparse",inputs:[{start:0,name:"data",type:"tensor"},{start:1,name:"indices",type:"tensor"},{start:2,name:"segmentIds",type:"tensor"}]}],sT={};Me(sT,{json:()=>G6});var G6=[{tfOpName:"FFT",category:"spectral",inputs:[{start:0,name:"x",type:"tensor"}]},{tfOpName:"IFFT",category:"spectral",inputs:[{start:0,name:"x",type:"tensor"}]},{tfOpName:"RFFT",category:"spectral",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"fft_length",type:"number",notSupported:!0}]},{tfOpName:"IRFFT",category:"spectral",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"fft_length",type:"number",notSupported:!0}]}],iT={};Me(iT,{json:()=>H6});var H6=[{tfOpName:"StringNGrams",category:"string",inputs:[{start:0,name:"data",type:"tensor"},{start:1,name:"dataSplits",type:"tensor"}],attrs:[{tfName:"separator",name:"separator",type:"string"},{tfName:"ngram_widths",name:"nGramWidths",type:"number[]"},{tfName:"left_pad",name:"leftPad",type:"string"},{tfName:"right_pad",name:"rightPad",type:"string"},{tfName:"pad_width",name:"padWidth",type:"number"},{tfName:"preserve_short_sequences",name:"preserveShortSequences",type:"bool"}],outputs:["ngrams","ngrams_splits"]},{tfOpName:"StringSplit",category:"string",inputs:[{start:0,name:"input",type:"tensor"},{start:1,name:"delimiter",type:"tensor"}],attrs:[{tfName:"skip_empty",name:"skipEmpty",type:"bool"}],outputs:["indices","values","shape"]},{tfOpName:"StringToHashBucketFast",category:"string",inputs:[{start:0,name:"input",type:"tensor"}],attrs:[{tfName:"num_buckets",name:"numBuckets",type:"number"}]}],oT={};Me(oT,{json:()=>j6});var j6=[{tfOpName:"Cast",category:"transformation",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"SrcT",name:"sdtype",type:"dtype",notSupported:!0},{tfName:"DstT",name:"dtype",type:"dtype"}]},{tfOpName:"ExpandDims",category:"transformation",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"axis",type:"number"}]},{tfOpName:"MirrorPad",category:"transformation",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"padding",type:"number[]"}],attrs:[{tfName:"mode",name:"mode",type:"string"}]},{tfOpName:"Pad",category:"transformation",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"padding",type:"number[]"}],attrs:[{tfName:"constant_value",name:"constantValue",type:"number",defaultValue:0}]},{tfOpName:"PadV2",category:"transformation",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"padding",type:"number[]"},{start:2,name:"constantValue",type:"number",defaultValue:0}]},{tfOpName:"Reshape",category:"transformation",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"shape",type:"number[]"}]},{tfOpName:"Squeeze",category:"transformation",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"axis",tfDeprecatedName:"squeeze_dims",name:"axis",type:"number[]"}]},{tfOpName:"SpaceToBatchND",category:"transformation",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"blockShape",type:"number[]"},{start:2,name:"paddings",type:"number[]"}]},{tfOpName:"BatchToSpaceND",category:"transformation",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"blockShape",type:"number[]"},{start:2,name:"crops",type:"number[]"}]},{tfOpName:"DepthToSpace",category:"transformation",inputs:[{start:0,name:"x",type:"tensor"}],attrs:[{tfName:"block_size",name:"blockSize",type:"number"},{tfName:"data_format",name:"dataFormat",type:"string"}]},{tfOpName:"BroadcastTo",category:"transformation",inputs:[{start:0,name:"x",type:"tensor"},{start:1,name:"shape",type:"number[]"}],attrs:[]},{tfOpName:"BroadcastArgs",category:"transformation",inputs:[{start:0,name:"s0",type:"tensor"},{start:1,name:"s1",type:"tensor"}],attrs:[]}],Rk=class{static get Instance(){return this._instance||(this._instance=new this)}constructor(){let e=[UN,GN,HN,jN,qN,KN,XN,YN,QN,JN,ZN,eT,tT,nT,aT,rT,sT,iT,oT],t=[].concat(...e.map(n=>n.json));this.opMappers=t.reduce((n,a)=>(n[a.tfOpName]=a,n),{})}transformGraph(e,t={}){let n=e.node,a=[],r=[],s=[],i=n.reduce((m,f)=>(m[f.name]=this.mapNode(f),f.op.startsWith("Placeholder")?a.push(m[f.name]):f.op==="Const"?r.push(m[f.name]):(f.input==null||f.input.length===0)&&s.push(m[f.name]),m),{}),o=[],l=[],u={},p={};t!=null&&(u=this.mapSignatureEntries(t.inputs),p=this.mapSignatureEntries(t.outputs));let d=Object.keys(i);d.forEach(m=>{let f=i[m];f.inputNames.forEach((g,y)=>{let[b,,x]=lr(g),v=i[b];if(v.outputs!=null){let k=v.outputs.indexOf(x);if(k!==-1){let T=`${b}:${k}`;f.inputNames[y]=T}}f.inputs.push(v),v.children.push(f)})}),Object.keys(p).length===0?d.forEach(m=>{let f=i[m];f.children.length===0&&l.push(f)}):Object.keys(p).forEach(m=>{let[f]=lr(m),g=i[f];g!=null&&(g.signatureKey=p[m],l.push(g))}),Object.keys(u).length>0?Object.keys(u).forEach(m=>{let[f]=lr(m),g=i[f];g&&(g.signatureKey=u[m],o.push(g))}):o=a;let c={};e.library!=null&&e.library.function!=null&&(c=e.library.function.reduce((m,f)=>(m[f.signature.name]=this.mapFunction(f),m),{}));let h={nodes:i,inputs:o,outputs:l,weights:r,placeholders:a,signature:t,functions:c};return s.length>0&&(h.initNodes=s),h}mapSignatureEntries(e){return Object.keys(e||{}).reduce((t,n)=>(t[e[n].name]=n,t),{})}mapNode(e){let t=VN(e.op)||this.opMappers[e.op]||{};e.attr==null&&(e.attr={});let n={name:e.name,op:e.op,category:t.category,inputNames:(e.input||[]).map(a=>a.startsWith("^")?a.slice(1):a),inputs:[],children:[],inputParams:{},attrParams:{},rawAttrs:e.attr,outputs:t.outputs};return t.inputs!=null&&(n.inputParams=t.inputs.reduce((a,r)=>(a[r.name]={type:r.type,inputIndexStart:r.start,inputIndexEnd:r.end},a),{})),t.attrs!=null&&(n.attrParams=t.attrs.reduce((a,r)=>{let s=r.type,i;switch(r.type){case"string":i=ax(e.attr,r.tfName,r.defaultValue),i===void 0&&!!r.tfDeprecatedName&&(i=ax(e.attr,r.tfDeprecatedName,r.defaultValue));break;case"string[]":i=px(e.attr,r.tfName,r.defaultValue),i===void 0&&!!r.tfDeprecatedName&&(i=px(e.attr,r.tfDeprecatedName,r.defaultValue));break;case"number":i=sx(e.attr,r.tfName,r.defaultValue||0),i===void 0&&!!r.tfDeprecatedName&&(i=sx(e.attr,r.tfDeprecatedName,r.defaultValue));break;case"number[]":i=ux(e.attr,r.tfName,r.defaultValue),i===void 0&&!!r.tfDeprecatedName&&(i=ux(e.attr,r.tfDeprecatedName,r.defaultValue));break;case"bool":i=rx(e.attr,r.tfName,r.defaultValue),i===void 0&&!!r.tfDeprecatedName&&(i=rx(e.attr,r.tfDeprecatedName,r.defaultValue));break;case"bool[]":i=dx(e.attr,r.tfName,r.defaultValue),i===void 0&&!!r.tfDeprecatedName&&(i=dx(e.attr,r.tfDeprecatedName,r.defaultValue));break;case"shape":i=lx(e.attr,r.tfName,r.defaultValue),i===void 0&&!!r.tfDeprecatedName&&(i=lx(e.attr,r.tfDeprecatedName,r.defaultValue));break;case"shape[]":i=cx(e.attr,r.tfName,r.defaultValue),i===void 0&&!!r.tfDeprecatedName&&(i=cx(e.attr,r.tfDeprecatedName,r.defaultValue));break;case"dtype":i=ix(e.attr,r.tfName,r.defaultValue),i===void 0&&!!r.tfDeprecatedName&&(i=ix(e.attr,r.tfDeprecatedName,r.defaultValue));break;case"dtype[]":i=ox(e.attr,r.tfName,r.defaultValue),i===void 0&&!!r.tfDeprecatedName&&(i=ox(e.attr,r.tfDeprecatedName,r.defaultValue));break;case"func":i=Mk(e.attr,r.tfName,r.defaultValue),i===void 0&&!!r.tfDeprecatedName&&(i=Mk(e.attr,r.tfDeprecatedName,r.defaultValue));break;case"tensor":case"tensors":break;default:throw new Error(`Unsupported param type: ${r.type} for op: ${e.op}`)}return a[r.name]={value:i,type:s},a},{})),n}mapFunction(e){let t=e.nodeDef,n=[],a=[],r={};t!=null&&(r=t.reduce((u,p)=>(u[p.name]=this.mapNode(p),p.op==="Const"&&a.push(u[p.name]),u),{}));let s=[],i=[];e.signature.inputArg.forEach(u=>{let[p]=lr(u.name),d={name:p,op:"Placeholder",inputs:[],inputNames:[],category:"graph",inputParams:{},attrParams:{dtype:{value:m0(u.type),type:"dtype"}},children:[]};d.signatureKey=u.name,s.push(d),r[p]=d}),Object.keys(r).forEach(u=>{let p=r[u];p.inputNames.forEach((d,c)=>{let[h,,m]=lr(d),f=r[h];if(f.outputs!=null){let g=f.outputs.indexOf(m);if(g!==-1){let y=`${h}:${g}`;p.inputNames[c]=y}}p.inputs.push(f),f.children.push(p)})});let o=e.ret;e.signature.outputArg.forEach(u=>{let[p,d]=lr(o[u.name]),c=r[p];c!=null&&(c.defaultOutput=d,i.push(c))});let l=this.mapArgsToSignature(e);return{nodes:r,inputs:s,outputs:i,weights:a,placeholders:n,signature:l}}mapArgsToSignature(e){return{methodName:e.signature.name,inputs:e.signature.inputArg.reduce((t,n)=>(t[n.name]=this.mapArgToTensorInfo(n),t),{}),outputs:e.signature.outputArg.reduce((t,n)=>(t[n.name]=this.mapArgToTensorInfo(n,e.ret),t),{})}}mapArgToTensorInfo(e,t){let n=e.name;return t!=null&&(n=t[n]),{name:n,dtype:e.type}}};function q6(e){let t=X().global;if(typeof t.atob!="undefined")return t.atob(e);if(typeof Buffer!="undefined")return new Buffer(e,"base64").toString();throw new Error("Unable to decode base64 in this environment. Missing built-in atob() or Buffer()")}function lT(e,t){let n=Array.isArray(e)?String.fromCharCode.apply(null,e):q6(e);return t?n:n.toLowerCase()}function ax(e,t,n,a=!1){let r=e[t];return r!=null?lT(r.s,a):n}function rx(e,t,n){let a=e[t];return a?a.b:n}function sx(e,t,n){let a=e[t]||{},r=a.i!=null?a.i:a.f!=null?a.f:n;return typeof r=="number"?r:parseInt(r,10)}function m0(e){switch(typeof e=="string"&&(e=Ia[e]),e){case Ia.DT_FLOAT:case Ia.DT_HALF:return"float32";case Ia.DT_INT32:case Ia.DT_INT64:case Ia.DT_INT8:case Ia.DT_UINT8:return"int32";case Ia.DT_BOOL:return"bool";case Ia.DT_DOUBLE:return"float32";case Ia.DT_STRING:return"string";default:return null}}function Mk(e,t,n){let a=e[t];return a&&a.func?a.func.name:n}function ix(e,t,n){let a=e[t];return a&&a.type?m0(a.type):n}function ox(e,t,n){let a=e[t];return a&&a.list&&a.list.type?a.list.type.map(r=>m0(r)):n}function uT(e){if(!e.unknownRank)return e.dim!=null?e.dim.map(t=>typeof t.size=="number"?t.size:parseInt(t.size,10)):[]}function lx(e,t,n){let a=e[t];return a&&a.shape?uT(a.shape):n}function ux(e,t,n){let a=e[t];return a?((a.list.f&&a.list.f.length?a.list.f:a.list.i)||[]).map(r=>typeof r=="number"?r:parseInt(r,10)):n}function px(e,t,n,a=!1){let r=e[t];return r&&r.list&&r.list.s?r.list.s.map(s=>lT(s,a)):n}function cx(e,t,n){let a=e[t];return a&&a.list&&a.list.shape?a.list.shape.map(r=>uT(r)):n}function dx(e,t,n){let a=e[t];return a&&a.list&&a.list.b?a.list.b:n}var K6=class{constructor(e,t,n){this.node=e,this.tensorMap=t,this.context=n,this.inputs=[],this.attrs={},this.inputs=e.inputNames.map(a=>this.getInput(a)),e.rawAttrs!=null&&(this.attrs=Object.keys(e.rawAttrs).reduce((a,r)=>(a[r]=this.getAttr(r),a),{}))}getInput(e){return Sn(e,this.tensorMap,this.context)}getAttr(e,t){let n=this.node.rawAttrs[e];if(n.tensor!=null)return Sn(e,this.tensorMap,this.context);if(n.i!=null||n.f!=null)return sx(this.node.rawAttrs,e,t);if(n.s!=null)return ax(this.node.rawAttrs,e,t);if(n.b!=null)return rx(this.node.rawAttrs,e,t);if(n.shape!=null)return lx(this.node.rawAttrs,e,t);if(n.type!=null)return ix(this.node.rawAttrs,e,t);if(n.list!=null){if(n.list.i!=null||n.list.f!=null)return ux(this.node.rawAttrs,e,t);if(n.list.s!=null)return px(this.node.rawAttrs,e,t);if(n.list.shape!=null)return cx(this.node.rawAttrs,e,t);if(n.list.b!=null)return dx(this.node.rawAttrs,e,t);if(n.list.type!=null)return ox(this.node.rawAttrs,e,t)}return t}},X6=(e,t,n)=>{switch(e.op){case"BiasAdd":case"AddV2":case"Add":return[J(I("a",e,t,n),I("b",e,t,n))];case"AddN":return[NS(I("tensors",e,t,n))];case"FloorMod":case"Mod":return[Nv(I("a",e,t,n),I("b",e,t,n))];case"Mul":return[B(I("a",e,t,n),I("b",e,t,n))];case"RealDiv":case"Div":return[fe(I("a",e,t,n),I("b",e,t,n))];case"DivNoNan":return[dv(I("a",e,t,n),I("b",e,t,n))];case"FloorDiv":return[rf(I("a",e,t,n),I("b",e,t,n))];case"Sub":return[ce(I("a",e,t,n),I("b",e,t,n))];case"Minimum":return[Uu(I("a",e,t,n),I("b",e,t,n))];case"Maximum":return[fr(I("a",e,t,n),I("b",e,t,n))];case"Pow":return[Dr(I("a",e,t,n),I("b",e,t,n))];case"SquaredDifference":return[Nf(I("a",e,t,n),I("b",e,t,n))];default:throw TypeError(`Node type ${e.op} is not implemented`)}},Y6=(e,t,n)=>{switch(e.op){case"Abs":case"ComplexAbs":return[zt(I("x",e,t,n))];case"Acos":return[Yx(I("x",e,t,n))];case"Acosh":return[Qx(I("x",e,t,n))];case"Asin":return[Zx(I("x",e,t,n))];case"Asinh":return[ev(I("x",e,t,n))];case"Atan":return[tv(I("x",e,t,n))];case"Atan2":return[nv(I("x",e,t,n),I("y",e,t,n))];case"Atanh":return[av(I("x",e,t,n))];case"Ceil":return[ov(I("x",e,t,n))];case"Complex":return[Ar(I("real",e,t,n),I("imag",e,t,n))];case"Cos":return[Vc(I("x",e,t,n))];case"Cosh":return[uf(I("x",e,t,n))];case"Elu":return[Wu(I("x",e,t,n))];case"Erf":return[hv(I("x",e,t,n))];case"Exp":return[gn(I("x",e,t,n))];case"Expm1":return[yv(I("x",e,t,n))];case"Floor":return[Vu(I("x",e,t,n))];case"Log":return[ea(I("x",e,t,n))];case"Log1p":return[Hc(I("x",e,t,n))];case"Imag":return[zc(I("x",e,t,n))];case"Neg":return[It(I("x",e,t,n))];case"Reciprocal":return[Cv(I("x",e,t,n))];case"Real":return[Nl(I("x",e,t,n))];case"Relu":return[Xe(I("x",e,t,n))];case"Round":return[bf(I("x",e,t,n))];case"Selu":return[vf(I("x",e,t,n))];case"Sigmoid":return[ma(I("x",e,t,n))];case"Sin":return[wf(I("x",e,t,n))];case"Sign":return[Ev(I("x",e,t,n))];case"Sinh":return[kf(I("x",e,t,n))];case"Softplus":return[xo(I("x",e,t,n))];case"Sqrt":return[pn(I("x",e,t,n))];case"Square":return[ut(I("x",e,t,n))];case"Tanh":return[ui(I("x",e,t,n))];case"Tan":return[$v(I("x",e,t,n))];case"ClipByValue":return[an(I("x",e,t,n),I("clipValueMin",e,t,n),I("clipValueMax",e,t,n))];case"Relu6":return[yf(I("x",e,t,n))];case"Rsqrt":return[xf(Sn(e.inputNames[0],t,n))];case"Prod":return[gf(I("x",e,t,n),I("axes",e,t,n))];case"LeakyRelu":return[Gc(I("x",e,t,n),I("alpha",e,t,n))];case"Prelu":return[Kc(I("x",e,t,n),I("alpha",e,t,n))];case"IsNan":return[xv(Sn(e.inputNames[0],t,n))];default:throw TypeError(`Node type ${e.op} is not implemented`)}};function Na(e,t,n=""){if(!(typeof e=="number"||typeof t=="number")){w.assert(e.length===t.length,()=>n+` Shapes ${e} and ${t} must match`);for(let a=0;a<e.length;a++){let r=e[a],s=t[a];w.assert(r<0||s<0||r===s,()=>n+` Shapes ${e} and ${t} must match`)}}}function Pk(e){return!(typeof e=="number"||e.some(t=>t<0))}function Op(e,t,n){let a=hx(e,n),r=!Pk(a);if(r&&t.length===0)throw new Error(`Tried to calculate elements of an empty list with non-fully-defined elementShape: ${a}`);if(r&&t.forEach(s=>{a=hx(s.shape,a)}),!Pk(a))throw new Error(`Non-fully-defined elementShape: ${a}`);return a}function hx(e,t){if(typeof e=="number")return t;if(typeof t=="number")return e;if(e.length!==t.length)throw new Error(`Incompatible ranks during merge: ${e} vs. ${t}`);let n=[];for(let a=0;a<e.length;++a){let r=e[a],s=t[a];if(r>=0&&s>=0&&r!==s)throw new Error(`Incompatible shape during merge: ${e} vs. ${t}`);n[a]=r>=0?r:s}return n}var Q6=class{constructor(e,t,n,a,r,s,i){this.name=e,this.dtype=t,this.maxSize=n,this.elementShape=a,this.identicalElementShapes=r,this.dynamicSize=s,this.clearAfterRead=i,this.tensors=[],this.closed_=!1,this.idTensor=we(0),tn(this.idTensor)}get id(){return this.idTensor.id}get closed(){return this.closed_}clearAndClose(e){this.tensors.forEach(t=>{(e==null||!e.has(t.tensor.id))&&t.tensor.dispose()}),this.tensors=[],this.closed_=!0,this.idTensor.dispose()}size(){return this.tensors.length}read(e){if(this.closed_)throw new Error(`TensorArray ${this.name} has already been closed.`);if(e<0||e>=this.size())throw new Error(`Tried to read from index ${e}, but array size is: ${this.size()}`);let t=this.tensors[e];if(t.cleared)throw new Error(`TensorArray ${this.name}: Could not read index ${e} twice because it was cleared after a previous read (perhaps try setting clear_after_read = false?).`);return this.clearAfterRead&&(t.cleared=!0),t.read=!0,t.tensor}readMany(e){return e.map(t=>this.read(t))}write(e,t){if(this.closed_)throw new Error(`TensorArray ${this.name} has already been closed.`);if(e<0||!this.dynamicSize&&e>=this.maxSize)throw new Error(`Tried to write to index ${e}, but array is not resizeable and size is: ${this.maxSize}`);let n=this.tensors[e]||{};if(t.dtype!==this.dtype)throw new Error(`TensorArray ${this.name}: Could not write to TensorArray index ${e},
because the value dtype is ${t.dtype}, but TensorArray dtype is ${this.dtype}.`);if(this.size()===0&&(this.elementShape==null||this.elementShape.length===0)&&(this.elementShape=t.shape),Na(this.elementShape,t.shape,`TensorArray ${this.name}: Could not write to TensorArray index ${e}.`),n.read)throw new Error(`TensorArray ${this.name}: Could not write to TensorArray index ${e}, because it has already been read.`);if(n.written)throw new Error(`TensorArray ${this.name}: Could not write to TensorArray index ${e}, because it has already been written.`);n.tensor=t,tn(t),n.written=!0,this.tensors[e]=n}writeMany(e,t){if(e.length!==t.length)throw new Error(`TensorArray ${this.name}: could not write multiple tensors,because the index size: ${e.length} is not the same as tensors size: ${t.length}.`);e.forEach((n,a)=>this.write(n,t[a]))}gather(e,t){if(!!t&&t!==this.dtype)throw new Error(`TensorArray dtype is ${this.dtype} but gather requested dtype ${t}`);if(e)e=e.slice(0,this.size());else{e=[];for(let a=0;a<this.size();a++)e.push(a)}if(e.length===0)return Jn([],[0].concat(this.elementShape));let n=this.readMany(e);return Na(this.elementShape,n[0].shape,"TensorArray shape mismatch: "),Mt(n,0)}concat(e){if(!!e&&e!==this.dtype)throw new Error(`TensorArray dtype is ${this.dtype} but concat requested dtype ${e}`);if(this.size()===0)return Jn([],[0].concat(this.elementShape));let t=[];for(let a=0;a<this.size();a++)t.push(a);let n=this.readMany(t);return Na(this.elementShape,n[0].shape,`TensorArray shape mismatch: tensor array shape (${this.elementShape}) vs first tensor shape (${n[0].shape})`),Ze(n,0)}scatter(e,t){if(t.dtype!==this.dtype)throw new Error(`TensorArray dtype is ${this.dtype} but tensor has dtype ${t.dtype}`);if(e.length!==t.shape[0])throw new Error(`Expected len(indices) == tensor.shape[0], but saw: ${e.length} vs. ${t.shape[0]}`);let n=Math.max(...e);if(!this.dynamicSize&&n>=this.maxSize)throw new Error(`Max index must be < array size (${n} vs. ${this.maxSize})`);this.writeMany(e,ht(t,0))}split(e,t){if(t.dtype!==this.dtype)throw new Error(`TensorArray dtype is ${this.dtype} but tensor has dtype ${t.dtype}`);let n=0,a=e.map(o=>(n+=o,n));if(n!==t.shape[0])throw new Error(`Expected sum of lengths to be equal to
tensor.shape[0], but sum of lengths is
${n}, and tensor's shape is: ${t.shape}`);if(!this.dynamicSize&&e.length!==this.maxSize)throw new Error(`TensorArray's size is not equal to the size of lengths (${this.maxSize} vs. ${e.length}), and the TensorArray is not marked as dynamically resizeable`);let r=n===0?0:t.size/n,s=[];O(()=>{t=W(t,[1,n,r]);for(let o=0;o<e.length;++o){let l=o===0?0:a[o-1],u=[0,l,0],p=[1,e[o],r];s[o]=W(He(t,u,p),this.elementShape)}return s});let i=[];for(let o=0;o<e.length;o++)i[o]=o;this.writeMany(i,s)}},Al=class{constructor(e,t,n,a=-1){this.tensors=e,this.elementShape=t,this.elementDtype=n,e!=null&&e.forEach(r=>{if(n!==r.dtype)throw new Error(`Invalid data types; op elements ${n}, but list elements ${r.dtype}`);Na(t,r.shape,"TensorList shape mismatch: "),tn(r)}),this.idTensor=we(0),this.maxNumElements=a,tn(this.idTensor)}get id(){return this.idTensor.id}copy(){return new Al([...this.tensors],this.elementShape,this.elementDtype)}clearAndClose(e){this.tensors.forEach(t=>{(e==null||!e.has(t.id))&&t.dispose()}),this.tensors.length=0,this.idTensor.dispose()}size(){return this.tensors.length}stack(e,t,n=-1){if(t!==this.elementDtype)throw new Error(`Invalid data types; op elements ${t}, but list elements ${this.elementDtype}`);if(n!==-1&&this.tensors.length!==n)throw new Error(`Operation expected a list with ${n} elements but got a list with ${this.tensors.length} elements.`);Na(e,this.elementShape,"TensorList shape mismatch: ");let a=Op(this.elementShape,this.tensors,e);return O(()=>{let r=this.tensors.map(s=>W(s,a));return Mt(r,0)})}popBack(e,t){if(t!==this.elementDtype)throw new Error(`Invalid data types; op elements ${t}, but list elements ${this.elementDtype}`);if(this.size()===0)throw new Error("Trying to pop from an empty list.");let n=Op(this.elementShape,this.tensors,e),a=this.tensors.pop();return Na(a.shape,e,"TensorList shape mismatch: "),W(a,n)}pushBack(e){if(e.dtype!==this.elementDtype)throw new Error(`Invalid data types; op elements ${e.dtype}, but list elements ${this.elementDtype}`);if(Na(e.shape,this.elementShape,"TensorList shape mismatch: "),this.maxNumElements===this.size())throw new Error("Trying to push element into a full list.");tn(e),this.tensors.push(e)}resize(e){if(e<0)throw new Error(`TensorListResize expects size to be non-negative. Got: ${e}`);if(this.maxNumElements!==-1&&e>this.maxNumElements)throw new Error(`TensorListResize input size ${e} is greater maxNumElement ${this.maxNumElements}.`);let t=new Al([],this.elementShape,this.elementDtype,this.maxNumElements);t.tensors.length=e;for(let n=0;n<Math.min(this.tensors.length,e);++n)t.tensors[n]=this.tensors[n];return t}getItem(e,t,n){if(n!==this.elementDtype)throw new Error(`Invalid data types; op elements ${n}, but list elements ${this.elementDtype}`);if(e<0||e>this.tensors.length)throw new Error(`Trying to access element ${e} in a list with ${this.tensors.length} elements.`);if(this.tensors[e]==null)throw new Error(`element at index ${e} is null.`);Na(this.tensors[e].shape,t,"TensorList shape mismatch: ");let a=Op(this.elementShape,this.tensors,t);return W(this.tensors[e],a)}setItem(e,t){if(t.dtype!==this.elementDtype)throw new Error(`Invalid data types; op elements ${t.dtype}, but list elements ${this.elementDtype}`);if(e<0||this.maxNumElements!==-1&&e>=this.maxNumElements)throw new Error(`Trying to set element ${e} in a list with max ${this.maxNumElements} elements.`);Na(this.elementShape,t.shape,"TensorList shape mismatch: "),tn(t),this.tensors[e]=t}gather(e,t,n){if(t!==this.elementDtype)throw new Error(`Invalid data types; op elements ${t}, but list elements ${this.elementDtype}`);Na(this.elementShape,n,"TensorList shape mismatch: "),e=e.slice(0,this.size());let a=Op(this.elementShape,this.tensors,n);return e.length===0?Jn([],[0].concat(a)):O(()=>{let r=e.map(s=>W(this.tensors[s],a));return Mt(r,0)})}concat(e,t){if(!!e&&e!==this.elementDtype)throw new Error(`TensorList dtype is ${this.elementDtype} but concat requested dtype ${e}`);Na(this.elementShape,t,"TensorList shape mismatch: ");let n=Op(this.elementShape,this.tensors,t);return this.size()===0?Jn([],[0].concat(n)):O(()=>{let a=this.tensors.map(r=>W(r,n));return Ze(a,0)})}};function J6(e,t,n){let a=e.dtype;if(e.shape.length<1)throw new Error(`Tensor must be at least a vector, but saw shape: ${e.shape}`);if(e.dtype!==n)throw new Error(`Invalid data types; op elements ${e.dtype}, but list elements ${n}`);let r=e.shape.slice(1);Na(r,t,"TensorList shape mismatch: ");let s=ht(e);return new Al(s,t,a)}function Z6(e,t,n){return new Al([],e,t,n)}function eH(e,t,n,a){if(t.length!==e.shape[0])throw new Error(`Expected len(indices) == tensor.shape[0], but saw: ${t.length} vs. ${e.shape[0]}`);let r=Math.max(...t);if(a!=null&&a!==-1&&r>=a)throw new Error(`Max index must be < array size (${r} vs. ${a})`);let s=new Al([],n,e.dtype,a),i=ht(e,0);return t.forEach((o,l)=>{s.setItem(o,i[l])}),s}function tH(e,t,n){let a=0,r=t.map(p=>(a+=p,a));if(a!==e.shape[0])throw new Error(`Expected sum of lengths to be equal to
tensor.shape[0], but sum of lengths is
${a}, and tensor's shape is: ${e.shape}`);let s=e.shape.slice(1),i=hx(s,n),o=a===0?0:e.size/a,l=O(()=>{let p=[];e=W(e,[1,a,o]);for(let d=0;d<t.length;++d){let c=d===0?0:r[d-1],h=[0,c,0],m=[1,t[d],o];p[d]=W(He(e,h,m),i)}return e.dispose(),p}),u=new Al([],n,e.dtype,t.length);for(let p=0;p<l.length;p++)u.setItem(p,l[p]);return u}var nH=async(e,t,n)=>{switch(e.op){case"If":case"StatelessIf":{let a=I("thenBranch",e,t,n),r=I("elseBranch",e,t,n),s=I("cond",e,t,n),i=I("args",e,t,n);return(await s.data())[0]?n.functionMap[a].executeFunctionAsync(i,n.tensorArrayMap,n.tensorListMap):n.functionMap[r].executeFunctionAsync(i,n.tensorArrayMap,n.tensorListMap)}case"While":case"StatelessWhile":{let a=I("body",e,t,n),r=I("cond",e,t,n),s=I("args",e,t,n),i=await n.functionMap[r].executeFunctionAsync(s,n.tensorArrayMap,n.tensorListMap),o=s.map(p=>p.id),l=await i[0].data();i.forEach(p=>{!p.kept&&o.indexOf(p.id)===-1&&p.dispose()});let u=s;for(;l[0];){let p=u;u=await n.functionMap[a].executeFunctionAsync(u,n.tensorArrayMap,n.tensorListMap);let d=u.map(h=>h.id);p.forEach(h=>{!h.kept&&o.indexOf(h.id)===-1&&d.indexOf(h.id)===-1&&h.dispose()});let c=await n.functionMap[r].executeFunctionAsync(u,n.tensorArrayMap,n.tensorListMap);l=await c[0].data(),c.forEach(h=>{!h.kept&&o.indexOf(h.id)===-1&&d.indexOf(h.id)===-1&&h.dispose()})}return u}case"LoopCond":{let a=I("pred",e,t,n);return[Nr(a)]}case"Switch":{let a=I("pred",e,t,n),r=I("data",e,t,n);return r.kept||(r=Nr(r)),(await a.data())[0]?[void 0,r]:[r,void 0]}case"Merge":{let a=e.inputNames.find(r=>Sn(r,t,n)!==void 0);if(a){let r=Sn(a,t,n);return[Nr(r)]}return}case"Enter":{let a=I("frameName",e,t,n),r=I("tensor",e,t,n);return n.enterFrame(a),[Nr(r)]}case"Exit":{let a=I("tensor",e,t,n);return n.exitFrame(),[Nr(a)]}case"NextIteration":{let a=I("tensor",e,t,n);return n.nextIteration(),[Nr(a)]}case"TensorArrayV3":{let a=I("size",e,t,n),r=I("dtype",e,t,n),s=I("elementShape",e,t,n),i=I("dynamicSize",e,t,n),o=I("clearAfterRead",e,t,n),l=I("identicalElementShapes",e,t,n),u=I("name",e,t,n),p=new Q6(u,r,a,s,l,i,o);return n.addTensorArray(p),[p.idTensor,we(1)]}case"TensorArrayWriteV3":{let a=I("tensorArrayId",e,t,n),r=I("index",e,t,n),s=I("tensor",e,t,n),i=n.getTensorArray(a.id);return i.write(r,s),[i.idTensor]}case"TensorArrayReadV3":{let a=I("tensorArrayId",e,t,n),r=I("index",e,t,n);return[n.getTensorArray(a.id).read(r)]}case"TensorArrayGatherV3":{let a=I("tensorArrayId",e,t,n),r=I("indices",e,t,n),s=I("dtype",e,t,n);return[n.getTensorArray(a.id).gather(r,s)]}case"TensorArrayScatterV3":{let a=I("tensorArrayId",e,t,n),r=I("indices",e,t,n),s=I("tensor",e,t,n),i=n.getTensorArray(a.id);return i.scatter(r,s),[i.idTensor]}case"TensorArrayConcatV3":{let a=I("tensorArrayId",e,t,n),r=n.getTensorArray(a.id),s=I("dtype",e,t,n);return[r.concat(s)]}case"TensorArraySplitV3":{let a=I("tensorArrayId",e,t,n),r=I("tensor",e,t,n),s=I("lengths",e,t,n),i=n.getTensorArray(a.id);return i.split(s,r),[i.idTensor]}case"TensorArraySizeV3":{let a=I("tensorArrayId",e,t,n),r=n.getTensorArray(a.id);return[we(r.size(),"int32")]}case"TensorArrayCloseV3":{let a=I("tensorArrayId",e,t,n),r=n.getTensorArray(a.id);return r.clearAndClose(),[r.idTensor]}case"TensorListSetItem":{let a=I("tensorListId",e,t,n),r=I("index",e,t,n),s=I("tensor",e,t,n),i=n.getTensorList(a.id);return i.setItem(r,s),[i.idTensor]}case"TensorListGetItem":{let a=I("tensorListId",e,t,n),r=I("index",e,t,n),s=I("elementShape",e,t,n),i=I("elementDType",e,t,n);return[n.getTensorList(a.id).getItem(r,s,i)]}case"TensorListScatterV2":case"TensorListScatter":{let a=I("indices",e,t,n),r=I("tensor",e,t,n),s=I("elementShape",e,t,n),i=I("numElements",e,t,n),o=eH(r,a,s,i);return n.addTensorList(o),[o.idTensor]}case"TensorListReserve":case"EmptyTensorList":{let a=I("elementShape",e,t,n),r=I("elementDType",e,t,n),s;e.op==="TensorListReserve"?s="numElements":s="maxNumElements";let i=I(s,e,t,n),o=Z6(a,r,i);return n.addTensorList(o),[o.idTensor]}case"TensorListGather":{let a=I("tensorListId",e,t,n),r=I("indices",e,t,n),s=I("elementShape",e,t,n),i=I("elementDType",e,t,n);return[n.getTensorList(a.id).gather(r,i,s)]}case"TensorListStack":{let a=I("tensorListId",e,t,n),r=I("elementShape",e,t,n),s=I("elementDType",e,t,n),i=I("numElements",e,t,n);return[n.getTensorList(a.id).stack(r,s,i)]}case"TensorListFromTensor":{let a=I("tensor",e,t,n),r=I("elementShape",e,t,n),s=I("elementDType",e,t,n),i=J6(a,r,s);return n.addTensorList(i),[i.idTensor]}case"TensorListConcat":case"TensorListConcatV2":{let a=I("tensorListId",e,t,n),r=n.getTensorList(a.id),s=I("dtype",e,t,n),i=I("elementShape",e,t,n);return[r.concat(s,i)]}case"TensorListPushBack":{let a=I("tensorListId",e,t,n),r=I("tensor",e,t,n),s=n.getTensorList(a.id);return s.pushBack(r),[s.idTensor]}case"TensorListPopBack":{let a=I("tensorListId",e,t,n),r=I("elementShape",e,t,n),s=I("elementDType",e,t,n);return[n.getTensorList(a.id).popBack(r,s)]}case"TensorListSplit":{let a=I("tensor",e,t,n),r=I("elementShape",e,t,n),s=I("lengths",e,t,n),i=tH(a,s,r);return n.addTensorList(i),[i.idTensor]}case"TensorListLength":{let a=I("tensorListId",e,t,n),r=n.getTensorList(a.id);return[we(r.size(),"int32")]}case"TensorListResize":{let a=I("tensorListId",e,t,n),r=I("size",e,t,n),s=n.getTensorList(a.id).resize(r);return n.addTensorList(s),[s.idTensor]}default:throw TypeError(`Node type ${e.op} is not implemented`)}};function Ok(e,t,n){let[a,r]=I("fusedOps",e,t,n),s=a==="biasadd",i=!s,o=r==="prelu",l=a==="fusedbatchnorm",u=I("numArgs",e,t,n);if(s){if(o&&u!==2)throw new Error("FusedConv2d and DepthwiseConv2d with BiasAdd and Prelu must have two extra arguments: bias and alpha.");if(!o&&s&&u!==1)throw new Error("FusedConv2d and DepthwiseConv2d with BiasAdd must have one extra argument: bias.")}if(l)throw new Error("FusedConv2d and DepthwiseConv2d with FusedBatchNorm is not supported");let p=I("strides",e,t,n),d=Dh(e,t,n),c=I("dataFormat",e,t,n).toUpperCase(),h=I("dilations",e,t,n),[m,f]=I("args",e,t,n);i&&(f=m,m=void 0);let g=I("leakyreluAlpha",e,t,n);return{stride:p,pad:d,dataFormat:c,dilations:h,biasArg:m,preluArg:f,activationFunc:r,leakyreluAlpha:g}}var aH=(e,t,n)=>{switch(e.op){case"Conv1D":{let a=I("stride",e,t,n),r=I("pad",e,t,n),s=I("dataFormat",e,t,n).toUpperCase(),i=I("dilation",e,t,n);return[of(I("x",e,t,n),I("filter",e,t,n),a,r,s,i)]}case"Conv2D":{let a=I("strides",e,t,n),r=Dh(e,t,n),s=I("dataFormat",e,t,n).toUpperCase(),i=I("dilations",e,t,n);return[Rt(I("x",e,t,n),I("filter",e,t,n),[a[1],a[2]],r,s,[i[1],i[2]])]}case"_FusedConv2D":{let{stride:a,pad:r,dataFormat:s,dilations:i,biasArg:o,preluArg:l,activationFunc:u,leakyreluAlpha:p}=Ok(e,t,n);return[ps.conv2d({x:I("x",e,t,n),filter:I("filter",e,t,n),strides:[a[1],a[2]],pad:r,dataFormat:s,dilations:[i[1],i[2]],bias:o,activation:u,preluActivationWeights:l,leakyreluAlpha:p})]}case"FusedDepthwiseConv2dNative":{let{stride:a,pad:r,dataFormat:s,dilations:i,biasArg:o,preluArg:l,activationFunc:u,leakyreluAlpha:p}=Ok(e,t,n);return[ps.depthwiseConv2d({x:I("x",e,t,n),filter:I("filter",e,t,n),strides:[a[1],a[2]],pad:r,dataFormat:s,dilations:[i[1],i[2]],bias:o,activation:u,preluActivationWeights:l,leakyreluAlpha:p})]}case"Conv2DBackpropInput":case"Conv2dTranspose":{let a=I("outputShape",e,t,n),r=I("strides",e,t,n),s=Dh(e,t,n);return[lf(I("x",e,t,n),I("filter",e,t,n),a,[r[1],r[2]],s)]}case"DepthwiseConv2dNative":case"DepthwiseConv2d":{let a=I("strides",e,t,n),r=Dh(e,t,n),s=I("dilations",e,t,n),i=I("dataFormat",e,t,n).toUpperCase();return[Ss(I("input",e,t,n),I("filter",e,t,n),[a[1],a[2]],r,i,[s[1],s[2]])]}case"Conv3D":{let a=I("strides",e,t,n),r=I("pad",e,t,n),s=I("dataFormat",e,t,n).toUpperCase(),i=I("dilations",e,t,n);return[uv(I("x",e,t,n),I("filter",e,t,n),[a[1],a[2],a[3]],r,s,[i[1],i[2],i[3]])]}case"AvgPool":{let a=I("strides",e,t,n),r=I("pad",e,t,n),s=I("kernelSize",e,t,n);return[ga(I("x",e,t,n),[s[1],s[2]],[a[1],a[2]],r)]}case"MaxPool":{let a=I("strides",e,t,n),r=I("pad",e,t,n),s=I("kernelSize",e,t,n);return[Pt(I("x",e,t,n),[s[1],s[2]],[a[1],a[2]],r)]}case"MaxPoolWithArgmax":{let a=I("strides",e,t,n),r=I("pad",e,t,n),s=I("kernelSize",e,t,n),i=I("includeBatchInIndex",e,t,n),{result:o,indexes:l}=ZS(I("x",e,t,n),[s[1],s[2]],[a[1],a[2]],r,i);return[o,l]}case"AvgPool3D":{let a=I("strides",e,t,n),r=I("pad",e,t,n),s=I("kernelSize",e,t,n);return[sv(I("x",e,t,n),[s[1],s[2],s[3]],[a[1],a[2],a[3]],r)]}case"MaxPool3D":{let a=I("strides",e,t,n),r=I("pad",e,t,n),s=I("kernelSize",e,t,n);return[Iv(I("x",e,t,n),[s[1],s[2],s[3]],[a[1],a[2],a[3]],r)]}case"Dilation2D":{let a=I("strides",e,t,n),r=I("pad",e,t,n),s=I("dilations",e,t,n),i=a[1],o=a[2],l=s[1],u=s[2];return[cv(I("x",e,t,n),I("filter",e,t,n),[i,o],r,[l,u],"NHWC")]}default:throw TypeError(`Node type ${e.op} is not implemented`)}},rH=(e,t,n)=>{switch(e.op){case"Fill":{let a=I("shape",e,t,n),r=I("dtype",e,t,n),s=I("value",e,t,n);return[Cn(a,s,r)]}case"LinSpace":{let a=I("start",e,t,n),r=I("stop",e,t,n),s=I("num",e,t,n);return[KS(a,r,s)]}case"Multinomial":{let a=I("logits",e,t,n),r=I("numSamples",e,t,n),s=I("seed",e,t,n);return[e2(a,r,s)]}case"OneHot":{let a=I("indices",e,t,n),r=I("depth",e,t,n),s=I("onValue",e,t,n),i=I("offValue",e,t,n);return[Sl(a,r,s,i)]}case"Ones":return[Qn(I("shape",e,t,n),I("dtype",e,t,n))];case"OnesLike":return[ta(I("x",e,t,n))];case"RandomUniform":return[Gu(I("shape",e,t,n),I("minval",e,t,n),I("maxval",e,t,n),I("dtype",e,t,n))];case"Range":{let a=I("start",e,t,n),r=I("stop",e,t,n),s=I("step",e,t,n);return[Tl(a,r,s,I("dtype",e,t,n))]}case"TruncatedNormal":{let a=I("shape",e,t,n),r=I("mean",e,t,n),s=I("stdDev",e,t,n),i=I("seed",e,t,n);return[Tf(a,r,s,I("dtype",e,t,n),i)]}case"Zeros":return[St(I("shape",e,t,n),I("dtype",e,t,n))];case"ZerosLike":return[Ke(I("x",e,t,n))];default:throw TypeError(`Node type ${e.op} is not implemented`)}};function Tb(e,t,n){let a=I("boxes",e,t,n),r=I("scores",e,t,n),s=I("maxOutputSize",e,t,n),i=I("iouThreshold",e,t,n),o=I("scoreThreshold",e,t,n),l=I("softNmsSigma",e,t,n);return{boxes:a,scores:r,maxOutputSize:s,iouThreshold:i,scoreThreshold:o,softNmsSigma:l}}var sH=async(e,t,n)=>{switch(e.op){case"NonMaxSuppressionV5":{let{boxes:a,scores:r,maxOutputSize:s,iouThreshold:i,scoreThreshold:o,softNmsSigma:l}=Tb(e,t,n),u=await Ln.nonMaxSuppressionWithScoreAsync(a,r,s,i,o,l);return[u.selectedIndices,u.selectedScores]}case"NonMaxSuppressionV4":{let{boxes:a,scores:r,maxOutputSize:s,iouThreshold:i,scoreThreshold:o}=Tb(e,t,n),l=I("padToMaxOutputSize",e,t,n),u=await Ln.nonMaxSuppressionPaddedAsync(a,r,s,i,o,l);return[u.selectedIndices,u.validOutputs]}case"NonMaxSuppressionV3":case"NonMaxSuppressionV2":{let{boxes:a,scores:r,maxOutputSize:s,iouThreshold:i,scoreThreshold:o}=Tb(e,t,n);return[await Ln.nonMaxSuppressionAsync(a,r,s,i,o)]}case"Where":{let a=oe(I("condition",e,t,n),"bool"),r=[await Mv(a)];return a.dispose(),r}case"ListDiff":return a2(I("x",e,t,n),I("y",e,t,n));default:throw TypeError(`Node type ${e.op} is not implemented`)}},iH=(e,t,n)=>{switch(e.op){case"LowerBound":{let a=I("sortedSequence",e,t,n),r=I("values",e,t,n);return[JS(a,r)]}case"TopKV2":{let a=I("x",e,t,n),r=I("k",e,t,n),s=I("sorted",e,t,n),i=Dv(a,r,s);return[i.values,i.indices]}case"UpperBound":{let a=I("sortedSequence",e,t,n),r=I("values",e,t,n);return[r2(a,r)]}case"Unique":{let a=I("x",e,t,n),r=Zh(a);return[r.values,r.indices]}case"UniqueV2":{let a=I("x",e,t,n),r=I("axis",e,t,n),s=Zh(a,r);return[s.values,s.indices]}default:throw TypeError(`Node type ${e.op} is not implemented`)}},oH=(e,t,n)=>{switch(e.op){case"Const":return t[e.name];case"PlaceholderWithDefault":let a=I("default",e,t,n);return[Sn(e.name,t,n)||a];case"Placeholder":return[Sn(e.name,t,n)];case"Identity":case"StopGradient":case"FakeQuantWithMinMaxVars":{let u=I("x",e,t,n);return[Nr(u)]}case"IdentityN":return I("x",e,t,n).map(u=>Nr(u));case"Snapshot":let r=I("x",e,t,n);return[Nr(r)];case"Shape":return[qe(I("x",e,t,n).shape,"int32")];case"ShapeN":return I("x",e,t,n).map(u=>qe(u.shape));case"Size":return[we(I("x",e,t,n).size,"int32")];case"Rank":return[we(I("x",e,t,n).rank,"int32")];case"NoOp":return[we(1)];case"Print":let s=I("x",e,t,n),i=I("data",e,t,n),o=I("message",e,t,n),l=I("summarize",e,t,n);console.warn("The graph has a tf.print() operation,usually used for debugging, which slows down performance."),console.log(o);for(let u=0;u<i.length;u++)console.log(Array.prototype.slice.call(i[u].dataSync()).slice(0,l));return[s];default:throw TypeError(`Node type ${e.op} is not implemented`)}},lH=class{constructor(e,t){this.keyDType=e,this.valueDType=t,this.handle=we(0),this.tensorMap=new Map,tn(this.handle)}get id(){return this.handle.id}clearAndClose(){this.tensorMap.forEach(e=>e.dispose()),this.tensorMap.clear(),this.handle.dispose()}size(){return this.tensorMap.size}tensorSize(){return we(this.size(),"int32")}async import(e,t){this.checkKeyAndValueTensor(e,t);let n=await e.data();return this.tensorMap.forEach(a=>a.dispose()),this.tensorMap.clear(),O(()=>{let a=ht(t),r=n.length,s=a.length;w.assert(r===s,()=>`The number of elements doesn't match, keys has ${r} elements, the values has ${s} elements.`);for(let i=0;i<r;i++){let o=n[i],l=a[i];tn(l),this.tensorMap.set(o,l)}return this.handle})}async find(e,t){this.checkKeyAndValueTensor(e,t);let n=await e.data();return O(()=>{let a=[];for(let r=0;r<n.length;r++){let s=n[r],i=this.findWithDefault(s,t);a.push(i)}return Mt(a)})}findWithDefault(e,t){let n=this.tensorMap.get(e);return n!=null?n:t}checkKeyAndValueTensor(e,t){if(e.dtype!==this.keyDType)throw new Error(`Expect key dtype ${this.keyDType}, but got ${e.dtype}`);if(t.dtype!==this.valueDType)throw new Error(`Expect value dtype ${this.valueDType}, but got ${t.dtype}`)}},uH=async(e,t,n,a)=>{switch(e.op){case"HashTable":case"HashTableV2":{let r=I("keyDType",e,t,n),s=I("valueDType",e,t,n),i=new lH(r,s);return a.addHashTable(e.name,i),[i.handle]}case"LookupTableImport":case"LookupTableImportV2":{let r=I("tableHandle",e,t,n,a),s=I("keys",e,t,n),i=I("values",e,t,n);return[await a.getHashTableById(r.id).import(s,i)]}case"LookupTableFind":case"LookupTableFindV2":{let r=I("tableHandle",e,t,n,a),s=I("keys",e,t,n),i=I("defaultValue",e,t,n);return[await a.getHashTableById(r.id).find(s,i)]}case"LookupTableSize":case"LookupTableSizeV2":{let r=I("tableHandle",e,t,n,a);return[a.getHashTableById(r.id).tensorSize()]}default:throw TypeError(`Node type ${e.op} is not implemented`)}},pH=(e,t,n)=>{switch(e.op){case"ResizeBilinear":{let a=I("images",e,t,n),r=I("size",e,t,n),s=I("alignCorners",e,t,n),i=I("halfPixelCenters",e,t,n);return[Ln.resizeBilinear(a,[r[0],r[1]],s,i)]}case"ResizeNearestNeighbor":{let a=I("images",e,t,n),r=I("size",e,t,n),s=I("alignCorners",e,t,n),i=I("halfPixelCenters",e,t,n);return[Ln.resizeNearestNeighbor(a,[r[0],r[1]],s,i)]}case"CropAndResize":{let a=I("image",e,t,n),r=I("boxes",e,t,n),s=I("boxInd",e,t,n),i=I("cropSize",e,t,n),o=I("method",e,t,n),l=I("extrapolationValue",e,t,n);return[Ln.cropAndResize(a,r,s,i,o,l)]}case"ImageProjectiveTransformV3":{let a=I("images",e,t,n),r=I("transforms",e,t,n),s=I("outputShape",e,t,n),i=I("fillValue",e,t,n),o=I("interpolation",e,t,n),l=I("fillMode",e,t,n);return[Ln.transform(a,r,o.toLowerCase(),l.toLowerCase(),i,s)]}default:throw TypeError(`Node type ${e.op} is not implemented`)}},cH=(e,t,n)=>{switch(e.op){case"Equal":return[Zn(I("a",e,t,n),I("b",e,t,n))];case"NotEqual":return[di(I("a",e,t,n),I("b",e,t,n))];case"Greater":return[Un(I("a",e,t,n),I("b",e,t,n))];case"GreaterEqual":return[Ns(I("a",e,t,n),I("b",e,t,n))];case"Less":return[cf(I("a",e,t,n),I("b",e,t,n))];case"LessEqual":return[Ts(I("a",e,t,n),I("b",e,t,n))];case"LogicalAnd":return[Ca(I("a",e,t,n),I("b",e,t,n))];case"LogicalNot":return[jc(I("a",e,t,n))];case"LogicalOr":return[mf(I("a",e,t,n),I("b",e,t,n))];case"Select":case"SelectV2":return[fn(I("condition",e,t,n),I("a",e,t,n),I("b",e,t,n))];default:throw TypeError(`Node type ${e.op} is not implemented`)}},dH=(e,t,n)=>{switch(e.op){case"BatchMatMul":case"BatchMatMulV2":case"MatMul":return[De(I("a",e,t,n),I("b",e,t,n),I("transposeA",e,t,n),I("transposeB",e,t,n))];case"Einsum":return[WS(I("equation",e,t,n),...I("tensors",e,t,n))];case"Transpose":return[Ae(I("x",e,t,n),I("perm",e,t,n))];case"_FusedMatMul":let[a,r]=I("fusedOps",e,t,n),s=a==="biasadd",i=r==="prelu",o=I("numArgs",e,t,n),l=I("leakyreluAlpha",e,t,n);if(s){if(i&&o!==2)throw new Error("Fused MatMul with BiasAdd and Prelu must have two extra arguments: bias and alpha.");if(!i&&o!==1)throw new Error("Fused MatMul with BiasAdd must have one extra argument: bias.")}let[u,p]=I("args",e,t,n);return[ps.matMul({a:I("a",e,t,n),b:I("b",e,t,n),transposeA:I("transposeA",e,t,n),transposeB:I("transposeB",e,t,n),bias:u,activation:r,preluActivationWeights:p,leakyreluAlpha:l})];default:throw TypeError(`Node type ${e.op} is not implemented`)}},hH=(e,t,n)=>{switch(e.op){case"EuclideanNorm":return[gv(I("x",e,t,n),I("axis",e,t,n),I("keepDims",e,t,n))];case"FusedBatchNorm":case"FusedBatchNormV2":return[$r(I("x",e,t,n),I("mean",e,t,n),I("variance",e,t,n),I("offset",e,t,n),I("scale",e,t,n),I("epsilon",e,t,n))];case"FusedBatchNormV3":return[$r(I("x",e,t,n),I("mean",e,t,n),I("variance",e,t,n),I("offset",e,t,n),I("scale",e,t,n),I("epsilon",e,t,n))];case"LRN":return[vv(I("x",e,t,n),I("radius",e,t,n),I("bias",e,t,n),I("alpha",e,t,n),I("beta",e,t,n))];case"Softmax":return[Qa(I("x",e,t,n))];case"LogSoftmax":return[hf(I("x",e,t,n))];case"SparseToDense":return[Pv(I("sparseIndices",e,t,n),I("outputShape",e,t,n),I("sparseValues",e,t,n),I("defaultValue",e,t,n))];default:throw TypeError(`Node type ${e.op} is not implemented`)}},mH=(e,t,n)=>{switch(e.op){case"Max":{let i=I("axis",e,t,n),o=I("keepDims",e,t,n);return[Ta(I("x",e,t,n),i,o)]}case"Mean":{let i=I("axis",e,t,n),o=I("keepDims",e,t,n);return[Et(I("x",e,t,n),i,o)]}case"Min":{let i=I("axis",e,t,n),o=I("keepDims",e,t,n);return[ic(I("x",e,t,n),i,o)]}case"Sum":{let i=I("axis",e,t,n),o=I("keepDims",e,t,n);return[be(I("x",e,t,n),i,o)]}case"All":{let i=I("axis",e,t,n),o=I("keepDims",e,t,n);return[sf(I("x",e,t,n),i,o)]}case"Any":{let i=I("axis",e,t,n),o=I("keepDims",e,t,n);return[rc(I("x",e,t,n),i,o)]}case"ArgMax":{let i=I("axis",e,t,n);return[li(I("x",e,t,n),i)]}case"ArgMin":{let i=I("axis",e,t,n);return[Jx(I("x",e,t,n),i)]}case"Prod":{let i=I("axis",e,t,n),o=I("keepDims",e,t,n);return[gf(I("x",e,t,n),i,o)]}case"Cumprod":{let i=I("axis",e,t,n),o=I("exclusive",e,t,n),l=I("reverse",e,t,n);return[sc(I("x",e,t,n),i,o,l)]}case"Cumsum":{let i=I("axis",e,t,n),o=I("exclusive",e,t,n),l=I("reverse",e,t,n);return[pf(I("x",e,t,n),i,o,l)]}case"Bincount":let a=I("x",e,t,n),r=I("weights",e,t,n),s=I("size",e,t,n);return[iv(a,r,s)];case"DenseBincount":{let i=I("x",e,t,n),o=I("weights",e,t,n),l=I("size",e,t,n),u=I("binaryOutput",e,t,n);return[zS(i,o,l,u)]}default:throw TypeError(`Node type ${e.op} is not implemented`)}},fH=(e,t,n)=>{switch(e.op){case"ConcatV2":case"Concat":{let a=I("n",e,t,n),r=I("axis",e,t,n),s=I("tensors",e,t,n);return s=s.slice(0,a),[Ze(s,r)]}case"Gather":{let a=I("x",e,t,n),r=I("indices",e,t,n);return[ci(a,oe(r,"int32"),0)]}case"GatherV2":{let a=I("axis",e,t,n),r=I("batchDims",e,t,n),s=I("x",e,t,n),i=I("indices",e,t,n);return[ci(s,oe(i,"int32"),a,r)]}case"Reverse":{let a=I("dims",e,t,n),r=[];for(let i=0;i<a.length;i++)a[i]&&r.push(i);let s=I("x",e,t,n);return[na(s,r)]}case"ReverseV2":{let a=I("axis",e,t,n),r=I("x",e,t,n);return[na(r,a)]}case"Slice":{let a=I("begin",e,t,n),r=I("size",e,t,n);return[He(I("x",e,t,n),a,r)]}case"StridedSlice":{let a=I("begin",e,t,n),r=I("end",e,t,n),s=I("strides",e,t,n),i=I("beginMask",e,t,n),o=I("endMask",e,t,n),l=I("ellipsisMask",e,t,n),u=I("newAxisMask",e,t,n),p=I("shrinkAxisMask",e,t,n),d=I("x",e,t,n);return[Av(d,a,r,s,i,o,l,u,p)]}case"Pack":return O(()=>{let a=I("axis",e,t,n),r=I("tensors",e,t,n),s=r[0].shape,i=dr(r[0]).shape,o=r.map(l=>{let u=w.arraysEqual(l.shape,s);if(!u&&!w.arraysEqual(dr(l).shape,i))throw new Error("the input tensors shape does not match");return u?l:W(l,s)});return[Mt(o,a)]});case"Unpack":{let a=I("axis",e,t,n),r=I("tensor",e,t,n);return ht(r,a)}case"Tile":{let a=I("reps",e,t,n);return[On(I("x",e,t,n),a)]}case"Split":case"SplitV":{let a=I("axis",e,t,n),r=I("numOrSizeSplits",e,t,n),s=I("x",e,t,n);return zn(s,r,a)}case"ScatterNd":{let a=I("indices",e,t,n),r=I("values",e,t,n),s=I("shape",e,t,n);return[o2(a,r,s)]}case"GatherNd":{let a=I("x",e,t,n),r=I("indices",e,t,n);return[l2(a,r)]}case"SparseToDense":{let a=I("sparseIndices",e,t,n),r=I("outputShape",e,t,n),s=I("sparseValues",e,t,n),i=I("defaultValue",e,t,n);return[Pv(a,s,r,s.dtype===i.dtype?i:oe(i,s.dtype))]}default:throw TypeError(`Node type ${e.op} is not implemented`)}},gH=(e,t,n)=>{switch(e.op){case"SparseFillEmptyRows":{let{outputIndices:a,outputValues:r,emptyRowIndicator:s,reverseIndexMap:i}=Wp.sparseFillEmptyRows(I("indices",e,t,n),I("values",e,t,n),I("denseShape",e,t,n),I("defaultValue",e,t,n));return[a,r,s,i]}case"SparseReshape":{let{outputIndices:a,outputShape:r}=Wp.sparseReshape(I("inputIndices",e,t,n),I("inputShape",e,t,n),I("newShape",e,t,n));return[a,r]}case"SparseSegmentMean":return[Wp.sparseSegmentMean(I("data",e,t,n),I("indices",e,t,n),I("segmentIds",e,t,n))];case"SparseSegmentSum":return[Wp.sparseSegmentSum(I("data",e,t,n),I("indices",e,t,n),I("segmentIds",e,t,n))];default:throw TypeError(`Node type ${e.op} is not implemented`)}},yH=(e,t,n)=>{switch(e.op){case"FFT":return[Xc(I("x",e,t,n))];case"IFFT":return[_l(I("x",e,t,n))];case"RFFT":return[Yc(I("x",e,t,n))];case"IRFFT":return[Sf(I("x",e,t,n))];default:throw TypeError(`Node type ${e.op} is not implemented`)}},bH=(e,t,n)=>{switch(e.op){case"StringNGrams":{let{nGrams:a,nGramsSplits:r}=$h.stringNGrams(I("data",e,t,n),I("dataSplits",e,t,n),I("separator",e,t,n),I("nGramWidths",e,t,n),I("leftPad",e,t,n),I("rightPad",e,t,n),I("padWidth",e,t,n),I("preserveShortSequences",e,t,n));return[a,r]}case"StringSplit":{let{indices:a,values:r,shape:s}=$h.stringSplit(I("input",e,t,n),I("delimiter",e,t,n),I("skipEmpty",e,t,n));return[a,r,s]}case"StringToHashBucketFast":return[$h.stringToHashBucketFast(I("input",e,t,n),I("numBuckets",e,t,n))];default:throw TypeError(`Node type ${e.op} is not implemented`)}},xH=(e,t,n)=>{switch(e.op){case"Cast":return[oe(I("x",e,t,n),I("dtype",e,t,n))];case"ExpandDims":{let a=I("axis",e,t,n);return[mn(I("x",e,t,n),a)]}case"Squeeze":{let a=I("axis",e,t,n);return[dr(I("x",e,t,n),a)]}case"Reshape":return[W(I("x",e,t,n),I("shape",e,t,n))];case"MirrorPad":return[Sv(I("x",e,t,n),I("padding",e,t,n),I("mode",e,t,n))];case"PadV2":case"Pad":return[ya(I("x",e,t,n),I("padding",e,t,n),I("constantValue",e,t,n))];case"SpaceToBatchND":{let a=I("blockShape",e,t,n),r=I("paddings",e,t,n);return[qc(I("x",e,t,n),a,r)]}case"BatchToSpaceND":{let a=I("blockShape",e,t,n),r=I("crops",e,t,n);return[Wc(I("x",e,t,n),a,r)]}case"DepthToSpace":{let a=I("blockSize",e,t,n),r=I("dataFormat",e,t,n).toUpperCase();return[pv(I("x",e,t,n),a,r)]}case"BroadcastTo":return[xl(I("x",e,t,n),I("shape",e,t,n))];case"BroadcastArgs":return[$S(I("s0",e,t,n),I("s1",e,t,n))];default:throw TypeError(`Node type ${e.op} is not implemented`)}};function Lk(e,t,n,a){let r=((s,i,o)=>{switch(s.category){case"arithmetic":return O(()=>X6(s,i,o));case"basic_math":return O(()=>Y6(s,i,o));case"control":return nH(s,i,o);case"convolution":return O(()=>aH(s,i,o));case"creation":return O(()=>rH(s,i,o));case"dynamic":return sH(s,i,o);case"evaluation":return O(()=>iH(s,i,o));case"image":return O(()=>pH(s,i,o));case"graph":return O(()=>oH(s,i,o));case"logical":return O(()=>cH(s,i,o));case"matrices":return O(()=>dH(s,i,o));case"normalization":return O(()=>hH(s,i,o));case"reduction":return O(()=>mH(s,i,o));case"slice_join":return O(()=>fH(s,i,o));case"sparse":return O(()=>gH(s,i,o));case"spectral":return O(()=>yH(s,i,o));case"string":return O(()=>bH(s,i,o));case"transformation":return O(()=>xH(s,i,o));case"hash_table":return uH(s,i,o,a);case"custom":let l=VN(s.op);if(l&&l.customExecutor)return l.customExecutor(new K6(s,i,o));throw TypeError(`Custom op ${s.op} is not registered.`);default:throw TypeError(`Unknown op '${s.op}'. File an issue at https://github.com/tensorflow/tfjs/issues so we can add it, or register a custom execution with tf.registerOp()`)}})(e,t,n);return w.isPromise(r)?r.then(s=>[].concat(s)):[].concat(r)}var zk=class{constructor(e={},t={},n={},a={}){this.weightMap=e,this.tensorArrayMap=t,this.tensorListMap=n,this.functionMap=a,this.rootContext={id:0,frameName:"",iterationId:0},this.contexts=[this.rootContext],this.lastId=0,this.generateCurrentContextIds()}newFrame(e,t){return{id:e,frameName:t,iterationId:0}}set currentContext(e){this.contexts!==e&&(this.contexts=e,this.generateCurrentContextIds())}get currentContext(){return this.contexts}get currentContextId(){return this._currentContextIds[0]}get currentContextIds(){return this._currentContextIds}generateCurrentContextIds(){let e=[];for(let t=0;t<this.contexts.length-1;t++){let n=this.contexts.slice(0,this.contexts.length-t);e.push(this.contextIdforContexts(n))}e.push(""),this._currentContextIds=e}contextIdforContexts(e){return e?e.map(t=>t.id===0&&t.iterationId===0?"":`${t.frameName}-${t.iterationId}`).join("/"):""}enterFrame(e){this.contexts&&(this.lastId++,this.contexts=this.contexts.slice(),this.contexts.push(this.newFrame(this.lastId,e)),this._currentContextIds.unshift(this.contextIdforContexts(this.contexts)))}exitFrame(){if(this.contexts&&this.contexts.length>1)this.contexts=this.contexts.slice(),this.contexts.splice(-1),this.currentContextIds.shift();else throw new Error("Cannot exit frame, the context is empty")}nextIteration(){if(this.contexts&&this.contexts.length>0){this.contexts=this.contexts.slice(),this.lastId++;let e=Object.assign({},this.contexts[this.contexts.length-1]);e.iterationId+=1,e.id=this.lastId,this.contexts.splice(-1,1,e),this._currentContextIds.splice(0,1,this.contextIdforContexts(this.contexts))}else throw new Error("Cannot increase frame iteration, the context is empty")}getWeight(e){return this.weightMap[e]}addTensorArray(e){this.tensorArrayMap[e.id]=e}getTensorArray(e){return this.tensorArrayMap[e]}addTensorList(e){this.tensorListMap[e.id]=e}getTensorList(e){return this.tensorListMap[e]}dispose(e){for(let t in this.tensorArrayMap)this.tensorArrayMap[t].clearAndClose(e);for(let t in this.tensorListMap)this.tensorListMap[t].clearAndClose(e)}};function Bk(e,t,n,a){let r=new Set,s=[],i=null,o=null,l=new Set,u=Object.keys(e).map(c=>Xn(c)[0]),p=[];a!=null&&(p=a.map(c=>Xn(c.name)[0]));let d=[...t];for(;d.length>0;){let c=d.pop();if((pT(c)||SH(c)||NH(c))&&i==null&&(i=c,o=i.children.map(h=>h.name).filter(h=>r.has(h))),r.add(c.name),n[c.name]==null&&u.indexOf(c.name)===-1&&p.indexOf(c.name)===-1){if(c.inputs.length===0){s.push(c.name);continue}c.inputs.forEach(h=>{l.has(h.name)||(l.add(h.name),d.push(h))})}}return{inputs:e,outputs:t,usedNodes:r,missingInputs:s,dynamicNode:i,syncInputs:o}}function vH(e,t,n){let{usedNodes:a,inputs:r}=n,s=[],i=Object.keys(r).map(p=>Xn(p)[0]).map(p=>e.nodes[p]),o=e.initNodes;i.forEach(p=>{a.has(p.name)&&s.push(p)}),e.weights.forEach(p=>{a.has(p.name)&&s.push(p)}),o!=null&&o.forEach(p=>{a.has(p.name)&&s.push(p)});let l=new Set,u=[];for(;s.length>0;){let p=s.pop();l.add(p.name),t[p.name]||u.push(p),p.children.forEach(d=>{!l.has(d.name)&&a.has(d.name)&&d.inputs.every(c=>l.has(c.name))&&s.push(d)})}return u}var wH=["Switch","Merge","Enter","Exit","NextIteration","StatelessIf","StatelessWhile","if","While"],kH=["NonMaxSuppressionV2","NonMaxSuppressionV3","NonMaxSuppressionV5","Where"],IH=["HashTable","HashTableV2","LookupTableImport","LookupTableImportV2","LookupTableFind","LookupTableFindV2","LookupTableSize","LookupTableSizeV2"];function pT(e){return wH.indexOf(e.op)>=0}function SH(e){return kH.indexOf(e.op)>=0}function NH(e){return IH.indexOf(e.op)>=0}var mx=class{constructor(e,t){this.graph=e,this.parent=t,this.compiledMap=new Map,this._weightMap={},this.SEPERATOR=",",this._functions={},this._functionExecutorMap={},this.intermediateTensors={},this.keepTensorForDebug=!1,this._outputs=e.outputs,this._inputs=e.inputs,this._initNodes=e.initNodes,this._signature=e.signature,this._functions=e.functions,e.functions!=null&&Object.keys(e.functions).forEach(n=>{this._functionExecutorMap[n]=new mx(e.functions[n],this)})}get weightIds(){return this.parent?this.parent.weightIds:this._weightIds}get functionExecutorMap(){return this.parent?this.parent.functionExecutorMap:this._functionExecutorMap}get weightMap(){return this.parent?this.parent.weightMap:this._weightMap}set weightMap(e){let t=Object.keys(e).map(n=>e[n].map(a=>a.id));this._weightIds=[].concat(...t),this._weightMap=e}set resourceManager(e){this._resourceManager=e}get inputs(){return this._inputs.map(e=>({name:e.name,shape:e.attrParams.shape?e.attrParams.shape.value:void 0,dtype:e.attrParams.dtype?e.attrParams.dtype.value:void 0}))}get outputs(){return this._outputs.map(e=>({name:e.name,shape:e.attrParams.shape?e.attrParams.shape.value:void 0,dtype:e.attrParams.dtype?e.attrParams.dtype.value:void 0}))}get inputNodes(){return this._inputs.map(e=>e.signatureKey||e.name)}get outputNodes(){return this._outputs.map(e=>{let t=e.signatureKey||e.name;return e.defaultOutput?`${t}:${e.defaultOutput}`:t})}get functions(){return Object.keys(this._functions).reduce((e,t)=>(e[t]=this._functions[t].signature,e),{})}getCompilationKey(e,t){let n=e.map(r=>r.name).sort(),a=t.map(r=>r.name).sort();return n.join(this.SEPERATOR)+"--"+a.join(this.SEPERATOR)}compile(e,t){let n=Bk(e,t,this.weightMap,this._initNodes),{missingInputs:a,dynamicNode:r,syncInputs:s}=n;if(r!=null)throw new Error(`This execution contains the node '${r.name}', which has the dynamic op '${r.op}'. Please use model.executeAsync() instead. Alternatively, to avoid the dynamic ops, specify the inputs [${s}]`);if(a.length>0){let i=t.map(l=>l.name),o=Object.keys(e);throw new Error(`Cannot compute the outputs [${i}] from the provided inputs [${o}]. Missing the following inputs: [${a}]`)}return vH(this.graph,this.weightMap,n)}execute(e,t){e=this.mapInputs(e);let n=Object.keys(e).sort();this.checkInputs(e),this.checkInputShapeAndType(e),t=this.mapOutputs(t),this.checkOutputs(t);let a=n.map(p=>this.graph.nodes[Xn(p)[0]]),r=t.map(p=>Xn(p)[0]),s=r.map(p=>this.graph.nodes[p]);this.resetIntermediateTensors(),s.length===0&&(s=this._outputs);let i=this.getCompilationKey(a,s),o=this.compiledMap.get(i);o==null&&(o=this.compile(e,s),this.compiledMap.set(i,o));let l={},u={};return O(()=>{let p=new zk(this.weightMap,l,u,this.functionExecutorMap),d=Object.assign({},this.weightMap);Object.keys(e).forEach(m=>{let[f,g]=Xn(m),y=[];y[g]=e[m],d[f]=y});let c=this.getFrozenTensorIds(d),h={};for(let m=0;m<o.length;m++){let f=o[m];if(!d[f.name]){let g=Lk(f,d,p,this._resourceManager);if(w.isPromise(g))throw new Error(`The execution of the op '${f.op}' returned a promise. Please use model.executeAsync() instead.`);d[f.name]=g,this.checkTensorForDisposal(f.name,f,d,p,c,r,h)}}return this.parent==null&&p.dispose(c),t.map(m=>Sn(m,d,p))})}getFrozenTensorIds(e){let t=[].concat.apply([],Object.keys(e).map(n=>e[n]).map(n=>n.map(a=>a.id)));return new Set(t)}checkTensorForDisposal(e,t,n,a,r,s,i){t.category==="control"||s.indexOf(e)!==-1||(n[e].forEach(o=>{o!=null&&(i[o.id]=(i[o.id]||0)+t.children.length)}),t.inputs.forEach(o=>{if(o.category!=="control"){let l=_6(o.name,n,a);l!=null&&l.forEach(u=>{if(u&&!u.kept&&!r.has(u.id)){let p=i[u.id];if(p===1){if(!this.keepTensorForDebug)u.dispose();else{let[d,c]=lr(t.name,a);this.intermediateTensors[d]?this.intermediateTensors[d][c]=u:(this.intermediateTensors[d]=[],this.intermediateTensors[d][c]=u)}delete i[u.id]}else p!=null&&i[u.id]--}})}}))}async executeAsync(e,t){return this._executeAsync(e,t)}disposeIntermediateTensors(){!this.intermediateTensors||(Object.keys(this.intermediateTensors).forEach(e=>this.intermediateTensors[e].forEach(t=>t.dispose())),this.disposeTensorsMap())}disposeTensorsMap(){!this.tensorsMap||Object.keys(this.tensorsMap).forEach(e=>{this.tensorsMap[e].forEach(t=>{t&&!t.kept&&!t.isDisposed&&!this.keepIds.has(t.id)&&t.dispose()})})}getIntermediateTensors(){return this.tensorsMap}resetIntermediateTensors(){for(let e in this.intermediateTensors)this.intermediateTensors[e].forEach(t=>t.dispose()),delete this.intermediateTensors[e]}async _executeAsync(e,t,n=!1,a={},r={}){n||(e=this.mapInputs(e),this.checkInputs(e),this.checkInputShapeAndType(e),t=this.mapOutputs(t),this.checkOutputs(t));try{this.keepTensorForDebug=X().getBool("KEEP_INTERMEDIATE_TENSORS")}catch(u){console.warn(u.message)}this.resetIntermediateTensors();let s=new zk(this.weightMap,a,r,this.functionExecutorMap);this.tensorsMap=await this.executeWithControlFlow(e,s,t,n);let i=t.map(u=>Sn(u,this.tensorsMap,s)),o=i.map(u=>u.id),l=Object.keys(e).map(u=>e[u].id);return this.keepIds=new Set([...o,...l,...this.weightIds]),this.keepTensorForDebug||this.disposeTensorsMap(),this.parent==null&&s.dispose(this.keepIds),i}async executeFunctionAsync(e,t,n){let a=e.reduce((r,s,i)=>(r[this.inputs[i].name]=s,r),{});return this._executeAsync(a,this.outputNodes,!0,t,n)}async executeWithControlFlow(e,t,n,a){let r=Object.keys(e),s=r.map(b=>this.graph.nodes[Xn(b)[0]]),i=n.map(b=>Xn(b)[0]),o=i.map(b=>this.graph.nodes[b]);o.length===0&&(o=this._outputs);let{usedNodes:l,missingInputs:u,dynamicNode:p,syncInputs:d}=Bk(e,o,this.weightMap,this._initNodes),c=[...s,...this.graph.weights,...this._initNodes||[]].map(b=>({node:b,contexts:t.currentContext})),h=Object.assign({},this.weightMap);Object.keys(e).forEach(b=>{let[x,v]=Xn(b),k=[];k[v]=e[b],h[x]=k});let m={},f=this.getFrozenTensorIds(h),g={};for(;c.length>0;){let b=this.processStack(s,c,t,h,g,f,i,m,l);await Promise.all(b)}p==null&&!a&&console.warn("This model execution did not contain any nodes with control flow or dynamic output shapes. You can use model.execute() instead.");let y=o.filter(b=>!pT(b)&&!Sn(b.name,h,t)).map(b=>b.name);if(y.length>0){let b="";throw p!=null&&(b=`Alternatively, to avoid the dynamic ops, use model.execute() and specify the inputs [${d}]`),new Error(`Cannot compute the outputs [${y}] from the provided inputs [${r}]. Consider providing the following inputs: [${u}]. ${b}`)}return h}processStack(e,t,n,a,r,s,i,o,l){let u=[];for(;t.length>0;){let p=t.pop();n.currentContext=p.contexts;let d="";if(p.node.op==="Enter"&&I("isConstant",p.node,a,n)&&([d]=lr(p.node.name,n)),a[p.node.name]==null){let c=Lk(p.node,a,n,this._resourceManager);d||([d]=lr(p.node.name,n));let h=n.currentContext;w.isPromise(c)?u.push(c.then(m=>(a[d]=m,n.currentContext=h,this.checkTensorForDisposal(d,p.node,a,n,s,i,o),this.processChildNodes(p.node,t,n,a,r,l),m))):(a[d]=c,this.checkTensorForDisposal(d,p.node,a,n,s,i,o),this.processChildNodes(p.node,t,n,a,r,l))}else this.processChildNodes(p.node,t,n,a,r,l)}return u}processChildNodes(e,t,n,a,r,s){e.children.forEach(i=>{let[o]=lr(i.name,n);r[o]||!s.has(i.name)||(i.op==="Merge"?i.inputNames.some(l=>!!Sn(l,a,n))&&(r[o]=!0,t.push({contexts:n.currentContext,node:i})):i.inputNames.every(l=>!!Sn(l,a,n))&&(r[o]=!0,t.push({contexts:n.currentContext,node:i})))})}dispose(){Object.keys(this.weightMap).forEach(e=>this.weightMap[e].forEach(t=>t.dispose()))}checkInputShapeAndType(e){Object.keys(e).forEach(t=>{let n=e[t],[a]=Xn(t),r=this.graph.nodes[a];if(r.attrParams.shape&&r.attrParams.shape.value){let s=r.attrParams.shape.value,i=s.length===n.shape.length&&n.shape.every((o,l)=>s[l]===-1||s[l]===o);w.assert(i,()=>`The shape of dict['${r.name}'] provided in model.execute(dict) must be [${s}], but was [${n.shape}]`)}r.attrParams.dtype&&r.attrParams.dtype.value&&w.assert(n.dtype===r.attrParams.dtype.value,()=>`The dtype of dict['${r.name}'] provided in model.execute(dict) must be ${r.attrParams.dtype.value}, but was ${n.dtype}`)})}mapInputs(e){let t={};for(let n in e)if(this._signature!=null&&this._signature.inputs!=null&&this._signature.inputs[n]!=null){let a=this._signature.inputs[n];t[a.name]=e[n]}else t[n]=e[n];return t}checkInputs(e){let t=Object.keys(e).filter(n=>{let[a]=Xn(n);return this.graph.nodes[a]==null});if(t.length>0)throw new Error(`The dict provided in model.execute(dict) has keys: [${t}] that are not part of graph`)}mapOutputs(e){return e.map(t=>this._signature!=null&&this._signature.outputs!=null&&this._signature.outputs[t]!=null?this._signature.outputs[t].name:t,{})}checkOutputs(e){e.forEach(t=>{let[n]=Xn(t);if(!this.graph.nodes[n])throw new Error(`The output '${t}' is not found in the graph`)})}},TH=class{constructor(e={},t={}){this.hashTableNameToHandle=e,this.hashTableMap=t}addHashTable(e,t){this.hashTableNameToHandle[e]=t.handle,this.hashTableMap[t.id]=t}getHashTableHandleByName(e){return this.hashTableNameToHandle[e]}getHashTableById(e){return this.hashTableMap[e]}dispose(){for(let e in this.hashTableMap)this.hashTableMap[e].clearAndClose(),delete this.hashTableMap[e];for(let e in this.hashTableNameToHandle)this.hashTableNameToHandle[e].dispose(),delete this.hashTableNameToHandle[e]}},_H="?tfjs-format=file",CH="model.json",f0=class{constructor(e,t={}){this.modelUrl=e,this.loadOptions=t,this.version="n/a",t==null&&(this.loadOptions={}),this.resourceManager=new TH}get modelVersion(){return this.version}get inputNodes(){return this.executor.inputNodes}get outputNodes(){return this.executor.outputNodes}get inputs(){return this.executor.inputs}get outputs(){return this.executor.outputs}get weights(){return this.executor.weightMap}get metadata(){return this.artifacts.userDefinedMetadata}get modelSignature(){return this.signature}findIOHandler(){let e=this.modelUrl;if(e.load!=null)this.handler=e;else if(this.loadOptions.requestInit!=null)this.handler=en.browserHTTPRequest(e,this.loadOptions);else{let t=en.getLoadHandlers(e,this.loadOptions);if(t.length===0)t.push(en.browserHTTPRequest(e,this.loadOptions));else if(t.length>1)throw new Error(`Found more than one (${t.length}) load handlers for URL '${[e]}'`);this.handler=t[0]}}load(){if(this.findIOHandler(),this.handler.load==null)throw new Error("Cannot proceed with model loading because the IOHandler provided does not have the `load` method implemented.");let e=this.handler.load();return w.isPromise(e)?e.then(t=>this.loadSync(t)):this.loadSync(e)}loadSync(e){this.artifacts=e;let t=this.artifacts.modelTopology,n;this.artifacts.userDefinedMetadata!=null&&this.artifacts.userDefinedMetadata.signature!=null?n=this.artifacts.userDefinedMetadata.signature:n=this.artifacts.signature,this.signature=n,this.version=`${t.versions.producer}.${t.versions.minConsumer}`;let a=en.decodeWeights(this.artifacts.weightData,this.artifacts.weightSpecs);if(this.executor=new mx(Rk.Instance.transformGraph(t,this.signature)),this.executor.weightMap=this.convertTensorMapToTensorsMap(a),this.executor.resourceManager=this.resourceManager,e.modelInitializer!=null&&e.modelInitializer.node!=null){let r=Rk.Instance.transformGraph(e.modelInitializer);this.initializer=new mx(r),this.initializer.weightMap=this.executor.weightMap,this.initializer.resourceManager=this.resourceManager,this.initializer.executeAsync({},[])}return!0}async save(e,t){if(typeof e=="string"){let n=en.getSaveHandlers(e);if(n.length===0)throw new Error(`Cannot find any save handlers for URL '${e}'`);if(n.length>1)throw new Error(`Found more than one (${n.length}) save handlers for URL '${e}'`);e=n[0]}if(e.save==null)throw new Error("GraphModel.save() cannot proceed because the IOHandler provided does not have the `save` attribute defined.");return e.save(this.artifacts)}predict(e,t){return this.execute(e,this.outputNodes)}normalizeInputs(e){if(!(e instanceof Fe)&&!Array.isArray(e))return e;if(e=Array.isArray(e)?e:[e],e.length!==this.inputNodes.length)throw new Error(`Input tensor count mismatch,the graph model has ${this.inputNodes.length} placeholders, while there are ${e.length} input tensors.`);return this.inputNodes.reduce((t,n,a)=>(t[n]=e[a],t),{})}normalizeOutputs(e){return e=e||this.outputNodes,Array.isArray(e)?e:[e]}execute(e,t){e=this.normalizeInputs(e),t=this.normalizeOutputs(t);let n=this.executor.execute(e,t);return n.length>1?n:n[0]}async executeAsync(e,t){e=this.normalizeInputs(e),t=this.normalizeOutputs(t);let n=await this.executor.executeAsync(e,t);return n.length>1?n:n[0]}getIntermediateTensors(){return this.executor.getIntermediateTensors()}disposeIntermediateTensors(){this.executor.disposeIntermediateTensors()}convertTensorMapToTensorsMap(e){return Object.keys(e).reduce((t,n)=>(t[n]=[e[n]],t),{})}dispose(){this.executor.dispose(),this.initializer&&this.initializer.dispose(),this.resourceManager.dispose()}};async function EH(e,t={}){if(e==null)throw new Error("modelUrl in loadGraphModel() cannot be null. Please provide a url or an IOHandler that loads the model");t==null&&(t={}),t.fromTFHub&&typeof e=="string"&&(e=AH(e));let n=new f0(e,t);return await n.load(),n}function FH(e){if(e==null)throw new Error("modelUrl in loadGraphModelSync() cannot be null. Please provide a url or an IOHandler that loads the model");if(!e.load)throw new Error(`modelUrl IO Handler ${e} has no load function`);let t=new f0(e);return t.load(),t}function AH(e){return e.endsWith("/")||(e=e+"/"),`${e}${CH}${_H}`}var $H="3.18.0",cT={};Me(cT,{CSVDataset:()=>vT,Dataset:()=>Yu,FileDataSource:()=>_T,TextLineDataset:()=>xT,URLDataSource:()=>CT,array:()=>tj,csv:()=>dj,func:()=>hj,generator:()=>mj,microphone:()=>gj,version_data:()=>yj,webcam:()=>fj,zip:()=>nj});var DH=xi(xI()),RH=xi(xI());function MH(e,t){return pm(e,t)}function pm(e,t,n=new Map,a=new Set){if(e==null)return null;if(typeof Blob=="function"&&e instanceof Blob)return e.slice();if(a.has(e))throw new Error("Circular references are not supported.");if(n.has(e))return n.get(e);let r=t(e);if(r.recurse&&r.value!==null)throw new Error("A deep map function may not return both a value and recurse=true.");if(r.recurse)if($l(e)){let s=Array.isArray(e)?[]:{};a.add(e);for(let i in e){let o=e[i],l=pm(o,t,n,a);s[i]=l}return a.delete(e),e.__proto__&&(s.__proto__=e.__proto__),s}else throw new Error(`Can't recurse into non-iterable type: ${e}`);else return n.set(e,r.value),r.value}function PH(e,t=hT){return dT(e,t)}function dT(e,t,n=new Set){let a=e[0];if(n.has(a))throw new Error("Circular references are not supported.");let r=t(e);if(r.recurse&&r.value!==null)throw new Error("A deep zip function may not return both a value and recurse=true.");if(r.recurse)if($l(a)){let s=Array.isArray(a)?[]:{};n.add(a);for(let i in a){let o=e.map(u=>u[i]),l=dT(o,t,n);s[i]=l}return n.delete(a),s}else throw new Error(`Can't recurse into non-iterable type: ${a}`);else return r.value}function hT(e){return e===null?null:$l(e[0])?{value:null,recurse:!0}:{value:e,recurse:!1}}async function mT(e,t){let n=new Map;pm(e,t,n);for(let a of Array.from(n.keys())){let r=n.get(a);if(w.isPromise(r)){let s=await r;n.set(a,s)}}return pm(e,t,n)}function $l(e){let t=!1;if(X().get("IS_BROWSER"))t=e instanceof TextDecoder;else{let{StringDecoder:n}=vI();t=e instanceof n}return e!=null&&!ArrayBuffer.isView(e)&&(Array.isArray(e)||typeof e=="object"&&!(e instanceof Fe)&&!(e instanceof Promise)&&!t)}function OH(e){return e==null||LH(e)||Array.isArray(e)||typeof e=="object"&&e instanceof Fe||w.isTypedArray(e)}function LH(e){return e===null||typeof e!="object"&&typeof e!="function"}function zH(e){return MH(e,BH)}function BH(e){return e instanceof Fe?{value:e.clone(),recurse:!1}:$l(e)?{value:null,recurse:!0}:{value:e,recurse:!1}}var fT=class{constructor(e){if(this.capacity=e,this.begin=0,this.end=0,e==null)throw new RangeError("Can't create a ring buffer of unknown capacity.");if(e<1)throw new RangeError("Can't create ring buffer of capacity < 1.");this.data=new Array(e),this.doubledCapacity=2*e}wrap(e){for(;e<0;)e+=this.doubledCapacity;return e%this.doubledCapacity}get(e){if(e<0)throw new RangeError("Can't get item at a negative index.");return this.data[e%this.capacity]}set(e,t){if(e<0)throw new RangeError("Can't set item at a negative index.");this.data[e%this.capacity]=t}length(){let e=this.end-this.begin;return e<0&&(e=this.doubledCapacity+e),e}isFull(){return this.length()===this.capacity}isEmpty(){return this.length()===0}push(e){if(this.isFull())throw new RangeError("Ring buffer is full.");this.set(this.end,e),this.end=this.wrap(this.end+1)}pushAll(e){for(let t of e)this.push(t)}pop(){if(this.isEmpty())throw new RangeError("Ring buffer is empty.");this.end=this.wrap(this.end-1);let e=this.get(this.end);return this.set(this.end,void 0),e}unshift(e){if(this.isFull())throw new RangeError("Ring buffer is full.");this.begin=this.wrap(this.begin-1),this.set(this.begin,e)}shift(){if(this.isEmpty())throw new RangeError("Ring buffer is empty.");let e=this.get(this.begin);return this.set(this.begin,void 0),this.begin=this.wrap(this.begin+1),e}shuffleExcise(e){if(this.isEmpty())throw new RangeError("Ring buffer is empty.");let t=this.wrap(this.begin+e),n=this.get(t);return this.set(t,this.pop()),n}},g0=class extends fT{constructor(){super(g0.INITIAL_CAPACITY)}isFull(){return!1}push(e){super.isFull()&&this.expand(),super.push(e)}unshift(e){super.isFull()&&this.expand(),super.unshift(e)}expand(){let e=this.capacity*2,t=new Array(e),n=this.length();for(let a=0;a<n;a++)t[a]=this.get(this.wrap(this.begin+a));this.data=t,this.capacity=e,this.doubledCapacity=2*this.capacity,this.begin=0,this.end=n}};g0.INITIAL_CAPACITY=32;function gT(e){return new UH(e)}function y0(e){return new GH(e)}function WH(e,t){return new yT(e,t)}function VH(e,t=as.FAIL){return new ZH(e,t)}var rn=class{async toArray(){let e=[],t=await this.next();for(;!t.done;)e.push(t.value),t=await this.next();return e}async toArrayForTest(){let e=this.prefetch(100),t=[],n=await e.next();for(;!n.done;)t.push(n.value),n=await e.next();return t}async resolveFully(){let e=await this.next();for(;!e.done;)e=await this.next()}async resolveWhile(e){let t=await this.next(),n=e(t.value);for(;!t.done&&n;)t=await this.next(),n=e(t.value)}handleErrors(e){return new QH(this,e)}filter(e){return new XH(this,e)}map(e){return new YH(this,e)}mapAsync(e){return new Wk(this,e)}serialMapAsync(e){return new Wk(this,e).serial()}flatmap(e){return new JH(this,e)}async forEachAsync(e){return this.map(e).resolveFully()}async serialForEach(e){return this.serialMapAsync(e).resolveWhile(t=>t===!0)}rowMajorBatch(e,t=!0){return new KH(this,e,t)}columnMajorBatch(e,t=!0,n=hT){return this.rowMajorBatch(e,t).map(a=>PH(a,n))}concatenate(e,t){return new yT(gT([this,e]),t)}take(e){return e<0||e==null?this:new qH(this,e)}skip(e){return e<0||e==null?this:new jH(this,e)}prefetch(e){return new bT(this,e)}shuffle(e,t){return new ej(this,e,t)}serial(){return new HH(this)}},UH=class extends rn{constructor(e){super(),this.items=e,this.trav=0}summary(){return`Array of ${this.items.length} items`}async next(){if(this.trav>=this.items.length)return{value:null,done:!0};let e=this.items[this.trav];return this.trav++,{value:zH(e),done:!1}}},GH=class extends rn{constructor(e){super(),this.nextFn=e}summary(){return"Function call"}async next(){try{return this.nextFn()}catch(e){throw e.message=`Error thrown while iterating through a dataset: ${e.message}`,e}}},HH=class extends rn{constructor(e){super(),this.upstream=e,this.lastRead=Promise.resolve({value:null,done:!1})}summary(){return`${this.upstream.summary()} -> Serial`}async next(){return this.lastRead=this.lastRead.then(()=>this.serialNext()),this.lastRead}async serialNext(){return this.upstream.next()}},jH=class extends rn{constructor(e,t){super(),this.upstream=e,this.maxCount=t,this.count=0,this.lastRead=Promise.resolve({value:null,done:!1})}summary(){return`${this.upstream.summary()} -> Skip`}async next(){return this.lastRead=this.lastRead.then(()=>this.serialNext()),this.lastRead}async serialNext(){for(;this.count++<this.maxCount;){let e=await this.upstream.next();if(e.done)return e;Re(e.value)}return this.upstream.next()}},qH=class extends rn{constructor(e,t){super(),this.upstream=e,this.maxCount=t,this.count=0}summary(){return`${this.upstream.summary()} -> Take`}async next(){return this.count++>=this.maxCount?{value:null,done:!0}:this.upstream.next()}},KH=class extends rn{constructor(e,t,n=!0){super(),this.upstream=e,this.batchSize=t,this.enableSmallLastBatch=n,this.lastRead=Promise.resolve({value:null,done:!1})}summary(){return`${this.upstream.summary()} -> RowMajorBatch`}async next(){return this.lastRead=this.lastRead.then(()=>this.serialNext()),this.lastRead}async serialNext(){let e=[];for(;e.length<this.batchSize;){let t=await this.upstream.next();if(t.done)return this.enableSmallLastBatch&&e.length>0?{value:e,done:!1}:{value:null,done:!0};e.push(t.value)}return{value:e,done:!1}}},XH=class extends rn{constructor(e,t){super(),this.upstream=e,this.predicate=t,this.lastRead=Promise.resolve({value:null,done:!1})}summary(){return`${this.upstream.summary()} -> Filter`}async next(){return this.lastRead=this.lastRead.then(()=>this.serialNext()),this.lastRead}async serialNext(){for(;;){let e=await this.upstream.next();if(e.done||this.predicate(e.value))return e;Re(e.value)}}},YH=class extends rn{constructor(e,t){super(),this.upstream=e,this.transform=t}summary(){return`${this.upstream.summary()} -> Map`}async next(){let e=await this.upstream.next();if(e.done)return{value:null,done:!0};let t=Ga.getTensorsInContainer(e.value),n=this.transform(e.value),a=Ga.getTensorsInContainer(n);for(let r of t)Ga.isTensorInList(r,a)||r.dispose();return{value:n,done:!1}}},QH=class extends rn{constructor(e,t){super(),this.upstream=e,this.handler=t,this.count=0,this.lastRead=Promise.resolve({value:null,done:!1})}summary(){return`${this.upstream.summary()} -> handleErrors`}async next(){return this.lastRead=this.lastRead.then(()=>this.serialNext()),this.lastRead}async serialNext(){for(;;)try{return await this.upstream.next()}catch(e){if(!this.handler(e))return{value:null,done:!0}}}},Wk=class extends rn{constructor(e,t){super(),this.upstream=e,this.transform=t}summary(){return`${this.upstream.summary()} -> AsyncMap`}async next(){let e=await this.upstream.next();if(e.done)return{value:null,done:!0};let t=Ga.getTensorsInContainer(e.value),n=await this.transform(e.value),a=Ga.getTensorsInContainer(n);for(let r of t)Ga.isTensorInList(r,a)||r.dispose();return{value:n,done:!1}}},b0=class extends rn{constructor(){super(),this.outputQueue=new g0,this.lastRead=Promise.resolve({value:null,done:!1})}async next(){return this.lastRead=this.lastRead.then(()=>this.serialNext()),this.lastRead}async serialNext(){for(;this.outputQueue.length()===0;)if(!await this.pump())return{value:null,done:!0};return{value:this.outputQueue.shift(),done:!1}}},JH=class extends b0{constructor(e,t){super(),this.upstream=e,this.transform=t}summary(){return`${this.upstream.summary()} -> Flatmap`}async pump(){let e=await this.upstream.next();if(e.done)return!1;let t=Ga.getTensorsInContainer(e.value),n=this.transform(e.value),a=Ga.getTensorsInContainer(n);this.outputQueue.pushAll(n);for(let r of t)Ga.isTensorInList(r,a)||r.dispose();return!0}},yT=class extends rn{constructor(e,t){super(),this.baseErrorHandler=t,this.lastRead=null,this.iterator=null,this.moreIterators=e}summary(){return"TODO: fill in upstream of chained summaries -> Chained"}async next(){return this.lastRead=this.readFromChain(this.lastRead),this.lastRead}async readFromChain(e){if(await e,this.iterator==null){let n=await this.moreIterators.next();if(n.done)return{value:null,done:!0};this.iterator=n.value,this.baseErrorHandler!=null&&(this.iterator=this.iterator.handleErrors(this.baseErrorHandler))}let t=await this.iterator.next();return t.done?(this.iterator=null,this.readFromChain(e)):t}},as;(function(e){e[e.FAIL=0]="FAIL",e[e.SHORTEST=1]="SHORTEST",e[e.LONGEST=2]="LONGEST"})(as||(as={}));var ZH=class extends rn{constructor(e,t=as.FAIL){super(),this.iterators=e,this.mismatchMode=t,this.count=0,this.currentPromise=null}summary(){return"{TODO: fill in upstream of zip summaries} -> Zip"}async nextState(e){await e;let t=0,n=0;function a(s){return s instanceof rn?{value:s.next().then(i=>(t++,i.done&&n++,i.value)),recurse:!1}:{value:null,recurse:!0}}let r=await mT(this.iterators,a);if(t===n)return{value:null,done:!0};if(n>0)switch(this.mismatchMode){case as.FAIL:throw new Error(`Zipped streams should have the same length. Mismatched at element ${this.count}.`);case as.SHORTEST:return{value:null,done:!0};case as.LONGEST:default:}return this.count++,{value:r,done:!1}}async next(){return this.currentPromise=this.nextState(this.currentPromise),this.currentPromise}},bT=class extends rn{constructor(e,t){super(),this.upstream=e,this.bufferSize=t,this.buffer=new fT(t)}summary(){return`${this.upstream.summary()} -> Prefetch`}refill(){for(;!this.buffer.isFull();){let e=this.upstream.next();this.buffer.push(e)}}next(){return this.refill(),this.buffer.shift()}},ej=class extends bT{constructor(e,t,n){super(e,t),this.upstream=e,this.windowSize=t,this.upstreamExhausted=!1,this.random=RH.alea(n||w.now().toString()),this.lastRead=Promise.resolve({value:null,done:!1})}async next(){return this.lastRead=this.lastRead.then(()=>this.serialNext()),this.lastRead}randomInt(e){return Math.floor(this.random()*e)}chooseIndex(){return this.randomInt(this.buffer.length())}async serialNext(){for(this.upstreamExhausted||this.refill();!this.buffer.isEmpty();){let e=this.chooseIndex(),t=await this.buffer.shuffleExcise(e);if(t.done)this.upstreamExhausted=!0;else return this.refill(),t}return{value:null,done:!0}}},Yu=class{constructor(){this.size=null}batch(e,t=!0){let n=this;w.assert(e>0,()=>`batchSize needs to be positive, but it is
${e}`);let a;return this.size===1/0||this.size==null?a=this.size:t?a=Math.ceil(this.size/e):a=Math.floor(this.size/e),Kn(async()=>(await n.iterator()).columnMajorBatch(e,t,aj),a)}concatenate(e){let t=this,n;return this.size===1/0||e.size===1/0?n=1/0:this.size!=null&&e.size!=null?n=this.size+e.size:n=null,Kn(async()=>(await t.iterator()).concatenate(await e.iterator()),n)}filter(e){let t=this,n;return this.size===1/0?n=1/0:n=null,Kn(async()=>(await t.iterator()).filter(a=>O(()=>e(a))),n)}async forEachAsync(e){return(await this.iterator()).forEachAsync(e)}map(e){let t=this;return Kn(async()=>(await t.iterator()).map(n=>O(()=>e(n))),this.size)}mapAsync(e){let t=this;return Kn(async()=>(await t.iterator()).mapAsync(e),this.size)}prefetch(e){if(e==null)throw new RangeError("`Dataset.prefetch()` requires bufferSize to be specified.");let t=this;return Kn(async()=>(await t.iterator()).prefetch(e),this.size)}repeat(e){let t=this,n;return this.size!=null&&e>0?n=this.size*e:e===0?n=0:this.size!=null&&(e===void 0||e<0)?n=1/0:n=null,Kn(async()=>{let a=y0(async()=>({value:await t.iterator(),done:!1}));return WH(a.take(e))},n)}skip(e){let t=this,n;return this.size!=null&&e>=0&&this.size>=e?n=this.size-e:this.size!=null&&(this.size<e||e===void 0||e<0)?n=0:n=null,Kn(async()=>(await t.iterator()).skip(e),n)}shuffle(e,t,n=!0){if(e==null||e<0)throw this.size==null?new RangeError("`Dataset.shuffle()` requires bufferSize to be specified."):new RangeError(`\`Dataset.shuffle()\` requires bufferSize to be specified. If your data fits in main memory (for regular JS objects), and/or GPU memory (for \`tf.Tensor\`s), consider setting bufferSize to the dataset size (${this.size} elements)`);let a=this,r=DH.alea(t||w.now().toString());return Kn(async()=>{let s=r.int32();return n&&(s+=r.int32()),(await a.iterator()).shuffle(e,s.toString())},this.size)}take(e){let t=this,n;return this.size!=null&&this.size>e?n=e:this.size!=null&&this.size<=e?n=this.size:n=null,Kn(async()=>(await t.iterator()).take(e),n)}async toArray(){if(this.size===1/0)throw new Error("Can not convert infinite data stream to array.");return(await this.iterator()).toArray()}async toArrayForTest(){if(this.size===1/0)throw new Error("Can not convert infinite data stream to array.");return(await this.iterator()).toArrayForTest()}};Yu.MAX_BUFFER_SIZE=1e4;function Kn(e,t=null){return new class extends Yu{constructor(){super(...arguments),this.size=t}async iterator(){return e()}}}function tj(e){return Kn(async()=>gT(e),e.length)}function nj(e){if(!$l(e))throw new Error("The argument to zip() must be an object or array.");let t;if(Array.isArray(e))for(let n=0;n<e.length;n++)t=t==null?e[n].size:Math.min(t,e[n].size);else if(e instanceof Object)for(let n in e)t=t==null?e[n].size:Math.min(t,e[n].size);return Kn(async()=>{let n=await mT(e,a=>{if(a instanceof Yu)return{value:a.iterator(),recurse:!1};if($l(a))return{value:null,recurse:!0};throw new Error("Leaves of the structure passed to zip() must be Datasets, not primitives.")});return VH(n,as.SHORTEST)},t)}function aj(e){if(e===null)return null;let t=e[0];return OH(t)?{value:rj(e),recurse:!1}:{value:null,recurse:!0}}function rj(e){if(e.length===0)throw new Error("Can't make a batch of zero elements.");return e[0]instanceof Fe?Mt(e):Jn(e)}var xT=class extends Yu{constructor(e){super(),this.input=e}async iterator(){return(await this.input.iterator()).decodeUTF8().split(`
`).map(e=>(e.endsWith("\r")&&(e=e.slice(0,-1)),e))}},Nh='"',Lp=Symbol("out"),Vk=Symbol("field"),Th=Symbol("quote"),_b=Symbol("quoteafterquote"),Uk=Symbol("quoteinquote"),vT=class extends Yu{constructor(e,t){super(),this.input=e,this.hasHeader=!0,this.fullColumnNames=null,this.columnNamesValidated=!1,this.columnConfigs=null,this.configuredColumnsOnly=!1,this.delimiter=",",this.delimWhitespace=!1,this.base=new xT(e),t||(t={}),this.hasHeader=t.hasHeader!==!1,this.fullColumnNames=t.columnNames,this.columnConfigs=t.columnConfigs,this.configuredColumnsOnly=t.configuredColumnsOnly,t.delimWhitespace?(w.assert(t.delimiter==null,()=>"Delimiter should not be provided when delimWhitespace is true."),this.delimWhitespace=!0,this.delimiter=" "):this.delimiter=t.delimiter?t.delimiter:","}async columnNames(){return this.columnNamesValidated||await this.setColumnNames(),this.configuredColumnsOnly?Object.keys(this.columnConfigs):this.fullColumnNames}async setColumnNames(){let e=await this.maybeReadHeaderLine();if(!this.fullColumnNames&&!e)throw new Error("Column names must be provided if there is no header line.");this.fullColumnNames&&e&&w.assert(e.length===this.fullColumnNames.length,()=>"The length of provided columnNames ("+this.fullColumnNames.length.toString()+") does not match the length of the header line read from file ("+e.length.toString()+")."),this.fullColumnNames||(this.fullColumnNames=e);let t=this.fullColumnNames.reduce((a,r)=>(a[r]=a[r]+1||1,a),{}),n=Object.keys(t).filter(a=>t[a]>1);if(w.assert(n.length===0,()=>"Duplicate column names found: "+n.toString()),this.columnConfigs){for(let a of Object.keys(this.columnConfigs))if(this.fullColumnNames.indexOf(a)===-1)throw new Error('The key "'+a+'" provided in columnConfigs does not match any of the column names ('+this.fullColumnNames.toString()+").")}this.columnNamesValidated=!0}async maybeReadHeaderLine(){if(this.hasHeader){let e=await(await this.base.iterator()).next();if(e.done)throw new Error("No data was found for CSV parsing.");let t=e.value;return this.parseRow(t,!1)}else return null}async iterator(){this.columnNamesValidated||await this.setColumnNames();let e=await this.base.iterator();return this.hasHeader&&(e=e.skip(1)),e.map(t=>this.makeDataElement(t))}makeDataElement(e){let t=this.parseRow(e),n={},a={};for(let r=0;r<this.fullColumnNames.length;r++){let s=this.fullColumnNames[r],i=this.columnConfigs?this.columnConfigs[s]:null;if(!(this.configuredColumnsOnly&&!i)){let o=t[r],l=null;if(o==="")if(i&&i.default!==void 0)l=i.default;else{if(i&&(i.required||i.isLabel))throw new Error(`Required column ${s} is empty in this line: ${e}`);l=void 0}else{let u=Number(o);if(isNaN(u))i&&i.dtype==="bool"?l=this.getBoolean(o):l=o;else if(!i||!i.dtype)l=u;else switch(i.dtype){case"float32":l=u;break;case"int32":l=Math.floor(u);break;case"bool":l=this.getBoolean(o);break;default:l=u}}i&&i.isLabel?a[s]=l:n[s]=l}}return Object.keys(a).length===0?n:{xs:n,ys:a}}getBoolean(e){return e==="1"||e.toLowerCase()==="true"?1:0}parseRow(e,t=!0){let n=[],a=0,r=e.length,s=Lp;for(let i=0;i<r;i++)switch(s){case Lp:switch(e.charAt(i)){case Nh:a=i+1,s=Th;break;case this.delimiter:if(a=i+1,this.delimiter===" "&&this.delimWhitespace)break;n.push(""),s=Lp;break;default:s=Vk,a=i;break}break;case Vk:switch(e.charAt(i)){case this.delimiter:n.push(e.substring(a,i)),s=Lp,a=i+1;break;default:}break;case Th:switch(e.charAt(i)){case Nh:s=_b;break;default:}break;case _b:switch(e.charAt(i)){case this.delimiter:n.push(e.substring(a,i-1)),s=Lp,a=i+1;break;case Nh:s=Th;break;default:s=Uk;break}break;case Uk:switch(e.charAt(i)){case Nh:s=Th;break;default:}break;default:}if(s===_b?n.push(e.substring(a,r-1)):n.push(e.substring(a)),t&&n.length!==this.fullColumnNames.length)throw new Error(`Invalid row in csv file. Should have ${this.fullColumnNames.length} elements in a row, but got ${n}`);return n}},wT=class extends rn{constructor(e){super(),this.microphoneConfig=e,this.isClosed=!1,this.fftSize=e.fftSize||1024;let t=Math.log2(this.fftSize);if(this.fftSize<0||t<4||t>14||!Number.isInteger(t))throw new Error(`Invalid fftSize: it must be a power of 2 between 2 to 4 and 2 to 14, but got ${this.fftSize}`);if(this.numFrames=e.numFramesPerSpectrogram||43,this.sampleRateHz=e.sampleRateHz,this.columnTruncateLength=e.columnTruncateLength||this.fftSize,this.audioTrackConstraints=e.audioTrackConstraints,this.smoothingTimeConstant=e.smoothingTimeConstant||0,this.includeSpectrogram=e.includeSpectrogram!==!1,this.includeWaveform=e.includeWaveform===!0,!this.includeSpectrogram&&!this.includeWaveform)throw new Error("Both includeSpectrogram and includeWaveform are false. At least one type of data should be returned.")}summary(){return"microphone"}static async create(e={}){if(!X().get("IS_BROWSER"))throw new Error("microphone API is only supported in browser environment.");let t=new wT(e);return await t.start(),t}async start(){try{this.stream=await navigator.mediaDevices.getUserMedia({audio:this.audioTrackConstraints==null?!0:this.audioTrackConstraints,video:!1})}catch(n){throw new Error(`Error thrown while initializing video stream: ${n.message}`)}if(!this.stream)throw new Error("Could not obtain audio from microphone.");let e=window.AudioContext||window.webkitAudioContext;if(this.audioContext=new e,!this.sampleRateHz)this.sampleRateHz=this.audioContext.sampleRate;else if(this.audioContext.sampleRate!==this.sampleRateHz)throw new Error(`Mismatch in sampling rate: Expected: ${this.sampleRateHz}; Actual: ${this.audioContext.sampleRate}`);let t=this.audioContext.createMediaStreamSource(this.stream);this.analyser=this.audioContext.createAnalyser(),this.analyser.fftSize=this.fftSize*2,this.analyser.smoothingTimeConstant=this.smoothingTimeConstant,t.connect(this.analyser),this.freqData=new Float32Array(this.fftSize),this.timeData=new Float32Array(this.fftSize)}async next(){if(this.isClosed)return{value:null,done:!0};let e,t,n=await this.getAudioData();if(this.includeSpectrogram){let a=this.flattenQueue(n.freqDataQueue);e=this.getTensorFromAudioDataArray(a,[this.numFrames,this.columnTruncateLength,1])}if(this.includeWaveform){let a=this.flattenQueue(n.timeDataQueue);t=this.getTensorFromAudioDataArray(a,[this.numFrames*this.fftSize,1])}return{value:{spectrogram:e,waveform:t},done:!1}}async capture(){return(await this.next()).value}async getAudioData(){let e=[],t=[],n=0;return new Promise(a=>{let r=setInterval(()=>{this.includeSpectrogram&&(this.analyser.getFloatFrequencyData(this.freqData),this.freqData[0]===-1/0&&a({freqDataQueue:e,timeDataQueue:t}),e.push(this.freqData.slice(0,this.columnTruncateLength))),this.includeWaveform&&(this.analyser.getFloatTimeDomainData(this.timeData),t.push(this.timeData.slice())),++n===this.numFrames&&(clearInterval(r),a({freqDataQueue:e,timeDataQueue:t}))},this.fftSize/this.sampleRateHz*1e3)})}stop(){this.isClosed||(this.isClosed=!0,this.analyser.disconnect(),this.audioContext.close(),this.stream!=null&&this.stream.getTracks().length>0&&this.stream.getTracks()[0].stop())}toArray(){throw new Error("Can not convert infinite audio stream to array.")}getSampleRate(){return this.sampleRateHz}flattenQueue(e){let t=e[0].length,n=new Float32Array(e.length*t);return e.forEach((a,r)=>n.set(a,r*t)),n}getTensorFromAudioDataArray(e,t){let n=new Float32Array(w.sizeFromShape(t));return n.set(e,n.length-e.length),Jn(n,t)}},kT=class extends rn{constructor(e,t){if(super(),this.webcamVideoElement=e,this.webcamConfig=t,this.isClosed=!0,this.resize=!1,this.needToResize())if(this.resize=!0,this.cropSize=[this.webcamConfig.resizeHeight,this.webcamConfig.resizeWidth],this.cropBoxInd=qe([0],"int32"),this.webcamConfig.centerCrop){let n=this.webcamConfig.resizeWidth*1/this.webcamVideoElement.width,a=this.webcamConfig.resizeHeight*1/this.webcamVideoElement.height,r=(1-n)/2,s=(1-a)/2,i=r+n,o=a+s;this.cropBox=Ha([s,r,o,i],[1,4])}else this.cropBox=Ha([0,0,1,1],[1,4])}summary(){return"webcam"}static async create(e,t={}){if(!X().get("IS_BROWSER"))throw new Error("tf.data.webcam is only supported in browser environment.");if(!e){if(e=document.createElement("video"),!t.resizeWidth||!t.resizeHeight)throw new Error("Please provide webcam video element, or resizeWidth and resizeHeight to create a hidden video element.");e.width=t.resizeWidth,e.height=t.resizeHeight}let n=new kT(e,t);return await n.start(),n}async start(){this.webcamConfig.facingMode&&w.assert(this.webcamConfig.facingMode==="user"||this.webcamConfig.facingMode==="environment",()=>`Invalid webcam facing mode: ${this.webcamConfig.facingMode}. Please provide 'user' or 'environment'`);try{this.stream=await navigator.mediaDevices.getUserMedia({video:{deviceId:this.webcamConfig.deviceId,facingMode:this.webcamConfig.facingMode?this.webcamConfig.facingMode:"user",width:this.webcamVideoElement.width,height:this.webcamVideoElement.height}})}catch(e){throw e.message=`Error thrown while initializing video stream: ${e.message}`,e}if(!this.stream)throw new Error("Could not obtain video from webcam.");try{this.webcamVideoElement.srcObject=this.stream}catch(e){console.log(e),this.webcamVideoElement.src=window.URL.createObjectURL(this.stream)}return this.webcamVideoElement.play(),this.isClosed=!1,new Promise(e=>{this.webcamVideoElement.onloadedmetadata=()=>{e()}})}async next(){if(this.isClosed)return{value:null,done:!0};let e;try{e=bo.fromPixels(this.webcamVideoElement)}catch(t){throw new Error(`Error thrown converting video to pixels: ${JSON.stringify(t)}`)}if(this.resize)try{return{value:this.cropAndResizeFrame(e),done:!1}}catch(t){throw new Error(`Error thrown cropping the video: ${t.message}`)}finally{e.dispose()}else return{value:e,done:!1}}needToResize(){return!!(this.webcamConfig.resizeWidth&&this.webcamConfig.resizeHeight&&(this.webcamVideoElement.width!==this.webcamConfig.resizeWidth||this.webcamVideoElement.height!==this.webcamConfig.resizeHeight))}cropAndResizeFrame(e){return O(()=>{let t=mn(oe(e,"float32"),0),n;n=Ln.cropAndResize(t,this.cropBox,this.cropBoxInd,this.cropSize,"bilinear");let a=n.shape;return W(n,a.slice(1))})}async capture(){return(await this.next()).value}stop(){this.stream.getTracks().forEach(e=>e.stop());try{this.webcamVideoElement.srcObject=null}catch(e){console.log(e),this.webcamVideoElement.src=null}this.isClosed=!0}toArray(){throw new Error("Can not convert infinite video stream to array.")}},IT=class{},ST=class extends rn{split(e){return new sj(this,e)}},sj=class extends ST{constructor(e,t){super(),this.upstream=e,this.impl=new ij(e,t)}summary(){return this.impl.summary()}async next(){return this.impl.next()}},ij=class extends b0{constructor(e,t){super(),this.upstream=e,this.separator=t,this.carryover=""}summary(){return`${this.upstream.summary()} -> Split('${this.separator}')`}async pump(){let e=await this.upstream.next();if(e.done)return this.carryover===""?!1:(this.outputQueue.push(this.carryover),this.carryover="",!0);let t=e.value.split(this.separator);t[0]=this.carryover+t[0];for(let n of t.slice(0,-1))this.outputQueue.push(n);return this.carryover=t[t.length-1],!0}},oj=class extends rn{decodeUTF8(){return new lj(this)}},lj=class extends ST{constructor(e){super(),this.upstream=e,this.impl=new uj(e)}summary(){return this.impl.summary()}async next(){return this.impl.next()}},uj=class extends b0{constructor(e){if(super(),this.upstream=e,X().get("IS_BROWSER"))this.decoder=new TextDecoder("utf-8");else{let{StringDecoder:t}=vI();this.decoder=new t("utf8")}}summary(){return`${this.upstream.summary()} -> Utf8`}async pump(){let e=await this.upstream.next(),t;if(e.done)return!1;t=e.value;let n;return X().get("IS_BROWSER")?n=this.decoder.decode(t,{stream:!0}):n=this.decoder.write(Buffer.from(t.buffer)),this.outputQueue.push(n),!0}},NT=class extends oj{constructor(e,t={}){super(),this.file=e,this.options=t,w.assert(e instanceof Uint8Array||(X().get("IS_BROWSER")?e instanceof File||e instanceof Blob:!1),()=>"FileChunkIterator only supports File, Blob and Uint8Array right now."),this.offset=t.offset||0,this.chunkSize=t.chunkSize||1024*1024}summary(){return`FileChunks ${this.file}`}async next(){return this.offset>=(this.file instanceof Uint8Array?this.file.byteLength:this.file.size)?{value:null,done:!0}:{value:await new Promise((e,t)=>{let n=this.offset+this.chunkSize;if(this.file instanceof Uint8Array)e(new Uint8Array(this.file.slice(this.offset,n)));else{let a=new FileReader;a.onload=s=>{let i=a.result;if(i instanceof ArrayBuffer&&(i=new Uint8Array(i)),!(i instanceof Uint8Array))return t(new TypeError("FileReader returned unknown type."));e(i)},a.onabort=s=>t(new Error("Aborted")),a.onerror=s=>t(new Error(s.type));let r=this.file.slice(this.offset,n);a.readAsArrayBuffer(r)}this.offset=n}),done:!1}}};async function pj(e,t={},n){let a,r;typeof e=="string"?a=e:(a=e.url,r=cj(e));let s=await(n||w.fetch)(a,r);if(s.ok){let i=new Uint8Array(await s.arrayBuffer());return new NT(i,t)}else throw new Error(s.statusText)}var cj=e=>({method:e.method,headers:e.headers,body:e.body,mode:e.mode,credentials:e.credentials,cache:e.cache,redirect:e.redirect,referrer:e.referrer,integrity:e.integrity});function TT(e){return typeof e=="string"&&e.slice(0,7)==="file://"}var _T=class extends IT{constructor(e,t={}){super(),this.input=e,this.options=t}async iterator(){if(TT(this.input)&&X().get("IS_NODE")){let e=Fx();this.input=e.readFileSync(this.input.slice(7))}return new NT(this.input,this.options)}},CT=class extends IT{constructor(e,t={}){super(),this.url=e,this.fileOptions=t}async iterator(){return TT(this.url)?new _T(this.url,this.fileOptions).iterator():pj(this.url,this.fileOptions)}};function dj(e,t={}){return new vT(new CT(e),t)}function hj(e){let t=y0(e);return Kn(async()=>t)}function mj(e){return Kn(async()=>{let t=await e();return y0(()=>t.next())})}async function fj(e,t){return kT.create(e,t)}async function gj(e){return wT.create(e)}var yj="3.18.0";function xe(e,t){Array.isArray(e)||(e=[e]),e.forEach(n=>{n!=null&&w.assert(n.dtype!=="complex64",()=>`${t} does not support complex64 tensors in the CPU backend.`)})}var bj=gr.whereImpl,x0=class extends gc{constructor(){super(),this.blockSize=48,this.firstUse=!0,this.data=new xm(this,sr())}nextDataId(){return x0.nextDataId++}write(e,t,n){this.firstUse&&(this.firstUse=!1,X().get("IS_NODE")&&C.warn(`
============================
Hi, looks like you are running TensorFlow.js in Node.js. To speed things up dramatically, install our node backend, visit https://github.com/tensorflow/tfjs-node for more details.
============================`));let a={id:this.nextDataId()};return this.data.set(a,{values:e,dtype:n,refCount:1}),a}makeTensorInfo(e,t,n){let a;if(t==="string"&&n!=null&&n.length>0&&w.isString(n[0])){let r=n.map(s=>w.encodeString(s));a=this.write(r,e,t)}else a=this.write(n,e,t);return{dataId:a,shape:e,dtype:t}}refCount(e){return this.data.has(e)?this.data.get(e).refCount:0}incRef(e){let t=this.data.get(e);t.refCount++}decRef(e){if(this.data.has(e)){let t=this.data.get(e);t.refCount--}}move(e,t,n,a,r){this.data.set(e,{values:t,dtype:a,refCount:r})}numDataIds(){return this.data.numDataIds()}async read(e){return this.readSync(e)}readSync(e){let{dtype:t,complexTensorInfos:n}=this.data.get(e);if(t==="complex64"){let a=this.readSync(n.real.dataId),r=this.readSync(n.imag.dataId);return C.mergeRealAndImagArrays(a,r)}return this.data.get(e).values}bufferSync(e){let t=this.readSync(e.dataId);if(e.dtype==="string")try{let n=t.map(a=>w.decodeString(a));return Ve(e.shape,e.dtype,n)}catch(n){throw new Error("Failed to decode encoded string bytes into utf-8")}return Ve(e.shape,e.dtype,t)}makeOutput(e,t,n){return sr().makeTensorFromTensorInfo(this.makeTensorInfo(t,n,e),this)}disposeData(e,t=!1){if(this.data.has(e)){if(this.data.get(e).refCount--,!t&&this.data.get(e).refCount>0)return!1;let{complexTensorInfos:n}=this.data.get(e);n!=null&&(this.disposeData(n.real.dataId,!0),this.disposeData(n.imag.dataId,!0)),this.data.delete(e)}return!0}disposeIntermediateTensorInfo(e){this.disposeData(e.dataId)}async time(e){let t=w.now();return e(),{kernelMs:w.now()-t}}memory(){return{unreliable:!0,reasons:["The reported memory is an upper bound. Due to automatic garbage collection, the true allocated memory may be less."]}}where(e){xe([e],"where");let t=this.readSync(e.dataId);return bj(e.shape,t)}dispose(){}floatPrecision(){return 32}epsilon(){return super.epsilon()}};x0.nextDataId=0;var ET={};Me(ET,{addImpl:()=>AT,bincountImpl:()=>w0,bincountReduceImpl:()=>$T,ceilImpl:()=>DT,concatImpl:()=>k0,equalImpl:()=>RT,expImpl:()=>PT,expm1Impl:()=>LT,floorImpl:()=>zT,gatherNdImpl:()=>BT,gatherV2Impl:()=>WT,greaterEqualImpl:()=>UT,greaterImpl:()=>VT,lessEqualImpl:()=>HT,lessImpl:()=>GT,linSpaceImpl:()=>jT,logImpl:()=>qT,maxImpl:()=>KT,maximumImpl:()=>XT,minimumImpl:()=>YT,multiplyImpl:()=>I0,negImpl:()=>QT,notEqualImpl:()=>JT,prodImpl:()=>ZT,rangeImpl:()=>N0,rsqrtImpl:()=>e_,scatterImpl:()=>fl,sigmoidImpl:()=>i5,simpleAbsImpl:()=>FT,sliceImpl:()=>dm,sparseFillEmptyRowsImpl:()=>n_,sparseReshapeImpl:()=>a_,sparseSegmentReductionImpl:()=>T0,sqrtImpl:()=>u5,squaredDifferenceImpl:()=>r_,stridedSliceImpl:()=>s_,stringNGramsImpl:()=>i_,stringSplitImpl:()=>o_,stringToHashBucketFastImpl:()=>l_,subImpl:()=>u_,tileImpl:()=>p_,topKImpl:()=>d_,transposeImpl:()=>S0,uniqueImpl:()=>h_});function FT(e){let t=new Float32Array(e.length);for(let n=0;n<e.length;++n)t[n]=Math.abs(e[n]);return t}var xj=e=>{let{x:t}=e.inputs,n=e.backend;xe(t,"abs");let a=new Float32Array(w.sizeFromShape(t.shape)),r=n.data.get(t.dataId).values;return a=FT(r),n.makeOutput(a,t.shape,t.dtype)},vj={kernelName:Ll,backendName:"cpu",kernelFunc:xj};function Ut(e){return(t,n,a,r,s)=>{let i=C.assertAndGetBroadcastShape(t,n),o=i.length,l=w.computeStrides(i),u=w.sizeFromShape(i),p=w.getTypedArrayFromDType(s,u),d=t.length,c=n.length,h=w.computeStrides(t),m=w.computeStrides(n),f=C.getBroadcastDims(t,i),g=C.getBroadcastDims(n,i);if(f.length+g.length===0)for(let y=0;y<p.length;++y)p[y]=e(a[y%a.length],r[y%r.length]);else for(let y=0;y<p.length;++y){let b=w.indexToLoc(y,o,l),x=b.slice(-d);f.forEach(_=>x[_]=0);let v=w.locToIndex(x,d,h),k=b.slice(-c);g.forEach(_=>k[_]=0);let T=w.locToIndex(k,c,m);p[y]=e(a[v],r[T])}return[p,i]}}function Yn(e){let{inputs:t,backend:n}=e,{real:a,imag:r}=t,s=n.data.get(a.dataId).values,i=n.data.get(r.dataId).values,o=n.makeTensorInfo(a.shape,"complex64"),l=n.data.get(o.dataId);return l.complexTensorInfos={real:n.makeTensorInfo(a.shape,"float32",s),imag:n.makeTensorInfo(r.shape,"float32",i)},o}var wj={kernelName:Tm,backendName:"cpu",kernelFunc:Yn};function cm(e,t,n="float32"){if(n==="complex64"){let r=cm(e,t,"float32"),s=cm(e,t,"float32");return Yn({inputs:{real:r,imag:s},backend:e})}let a=w.makeZerosTypedArray(w.sizeFromShape(t),n);return e.makeTensorInfo(t,n,a)}function hr(e){let{inputs:t,backend:n}=e,{x:a}=t;return n.incRef(a.dataId),{dataId:a.dataId,shape:a.shape,dtype:a.dtype}}var kj={kernelName:Bi,backendName:"cpu",kernelFunc:hr};function mi(e){let{inputs:t,backend:n}=e,{input:a}=t,r=n.data.get(a.dataId).complexTensorInfos.real,s=n.data.get(r.dataId).values;return n.makeTensorInfo(r.shape,r.dtype,s)}var Ij={kernelName:Hm,backendName:"cpu",kernelFunc:mi};function fs(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{dtype:s}=a;if(s==="complex64"){if(r.dtype==="complex64")return hr({inputs:{x:r},backend:n});let i=cm(n,r.shape,r.dtype),o=fs({inputs:{x:r},backend:n,attrs:{dtype:"float32"}}),l=Yn({inputs:{real:o,imag:i},backend:n});return n.disposeIntermediateTensorInfo(i),n.disposeIntermediateTensorInfo(o),l}if(r.dtype==="complex64"){let i=mi({inputs:{input:r},backend:n}),o=fs({inputs:{x:i},backend:n,attrs:{dtype:s}});return n.disposeIntermediateTensorInfo(i),o}if(!w.hasEncodingLoss(r.dtype,s)){let i=hr({inputs:{x:r},backend:n});return{dataId:i.dataId,shape:i.shape,dtype:s}}if(s==="int32"){let i=n.data.get(r.dataId).values,o=Int32Array.from(i);return n.makeTensorInfo(r.shape,"int32",o)}if(s==="bool"){let i=n.data.get(r.dataId).values,o=w.toTypedArray([0],r.dtype),[l,u]=Ut((p,d)=>p!==d?1:0)(r.shape,[],i,o,"bool");return n.makeTensorInfo(u,"bool",l)}throw new Error(`Error in Cast: failed to cast ${r.dtype} to ${s}`)}var Sj={kernelName:Ni,backendName:"cpu",kernelFunc:fs};function sn(e,t,n,a){return n==null?({inputs:r,backend:s})=>{let{a:i,b:o}=r,l=s;xe([i,o],e);let u=l.data.get(i.dataId).values,p=l.data.get(o.dataId).values,d=i.dtype==="string"?C.fromUint8ToStringArray(u):u,c=i.dtype==="string"?C.fromUint8ToStringArray(p):p,h=a||i.dtype,[m,f]=t(i.shape,o.shape,d,c,h);return l.makeTensorInfo(f,h,m)}:({inputs:r,backend:s})=>{let{a:i,b:o}=r,l=s;if(i.dtype==="complex64"||o.dtype==="complex64"){let u=fs({inputs:{x:i},backend:l,attrs:{dtype:"complex64"}}),p=l.data.get(u.dataId),d=p.complexTensorInfos.real,c=p.complexTensorInfos.imag,h=l.data.get(d.dataId).values,m=l.data.get(c.dataId).values,f=fs({inputs:{x:o},backend:l,attrs:{dtype:"complex64"}}),g=l.data.get(f.dataId),y=g.complexTensorInfos.real,b=g.complexTensorInfos.imag,x=l.data.get(y.dataId).values,v=l.data.get(b.dataId).values,[k,T,_]=n(i.shape,o.shape,h,m,x,v),E=l.makeTensorInfo(_,"float32",k),A=l.makeTensorInfo(_,"float32",T),M=Yn({inputs:{real:E,imag:A},backend:l});return l.disposeIntermediateTensorInfo(u),l.disposeIntermediateTensorInfo(f),l.disposeIntermediateTensorInfo(E),l.disposeIntermediateTensorInfo(A),M}else{let u=l.data.get(i.dataId).values,p=l.data.get(o.dataId).values,d=a||i.dtype,[c,h]=t(i.shape,o.shape,u,p,d);return l.makeTensorInfo(h,d,c)}}}function v0(e){return(t,n,a,r,s,i)=>{let o=C.assertAndGetBroadcastShape(t,n),l=w.sizeFromShape(o),u=o.length,p=w.computeStrides(o),d=w.getTypedArrayFromDType("float32",l),c=w.getTypedArrayFromDType("float32",l),h=C.getBroadcastDims(t,o),m=C.getBroadcastDims(n,o),f=C.mergeRealAndImagArrays(a,r),g=C.mergeRealAndImagArrays(s,i),y=t.length,b=w.computeStrides(t),x=n.length,v=w.computeStrides(n);if(h.length+m.length===0)for(let k=0;k<d.length;k++){let T=k%f.length,_=k%g.length,E=e(f[T*2],f[T*2+1],g[_*2],g[_*2+1]);d[k]=E.real,c[k]=E.imag}else for(let k=0;k<d.length;k++){let T=w.indexToLoc(k,u,p),_=T.slice(-y);h.forEach(S=>_[S]=0);let E=w.locToIndex(_,y,b),A=T.slice(-x);m.forEach(S=>A[S]=0);let M=w.locToIndex(A,x,v),$=e(f[E*2],f[E*2+1],g[M*2],g[M*2+1]);d[k]=$.real,c[k]=$.imag}return[d,c,o]}}var AT=Ut((e,t)=>e+t),Nj=v0((e,t,n,a)=>({real:e+n,imag:t+a})),Dl=sn(bs,AT,Nj),Tj={kernelName:bs,backendName:"cpu",kernelFunc:Dl};function w0(e,t,n,a,r){let s=w.sizeFromShape(a),i=w.makeZerosTypedArray(r,n);for(let o=0;o<e.length;o++){let l=e[o];if(l<0)throw new Error("Input x must be non-negative!");l>=r||(s>0?i[l]+=t[o]:i[l]+=1)}return i}function $T(e,t,n,a=!1){let r=e.shape[0],s=e.shape[1],i=Ve([r,n],t.dtype);for(let o=0;o<r;o++)for(let l=0;l<s;l++){let u=e.get(o,l);if(u<0)throw new Error("Input x must be non-negative!");u>=n||(a?i.set(1,o,u):t.size>0?i.set(i.get(o,u)+t.get(o,l),o,u):i.set(i.get(o,u)+1,o,u))}return i}function _s(e){return(t,n,a)=>{let r=w.getTypedArrayFromDType(n,t.length);for(let s=0;s<t.length;++s)r[s]=e(t[s],a);return r}}function ot(e,t,n){return({inputs:a,attrs:r,backend:s})=>{let{x:i}=a;if(xe(i,e),i.dtype==="string"||n==="string")throw new Error("unaryKernelFunc does not support string input/output");let o=s,l=o.data.get(i.dataId).values,u=w.sizeFromShape(i.shape),p=n||i.dtype,d=w.getArrayFromDType(p,u);for(let c=0;c<u;++c)d[c]=t(l[c],r);return o.makeTensorInfo(i.shape,p,d)}}function Qu(e,t,n){return({inputs:a,attrs:r,backend:s})=>{let{x:i}=a;if(xe(i,e),i.dtype==="string"||n==="string")throw new Error("unaryKernelFunc does not support string input/output");let o=s,l=o.data.get(i.dataId).values,u=n||i.dtype,p=t(l,u,r);return o.makeTensorInfo(i.shape,u,p)}}var DT=_s(e=>Math.ceil(e)),_j=Qu(Ti,DT),Cj={kernelName:Ti,backendName:"cpu",kernelFunc:_j};function k0(e,t,n,a){let r=w.getArrayFromDType(n,w.sizeFromShape(t));if(a&&n!=="string"){let s=0;e.forEach(i=>{let o=w.sizeFromShape(i.shape);r.set(i.vals,s),s+=o})}else{let s=0;e.forEach(i=>{let o=n==="string"?C.fromUint8ToStringArray(i.vals):i.vals,l=0;for(let u=0;u<i.shape[0];++u){let p=u*t[1]+s;for(let d=0;d<i.shape[1];++d)r[p+d]=o[l++]}s+=i.shape[1]})}return r}var RT=Ut((e,t)=>e===t?1:0),MT=sn(eu,RT,null,"bool"),Ej={kernelName:eu,backendName:"cpu",kernelFunc:MT},PT=_s(e=>Math.exp(e)),OT=Qu(Mi,PT,"float32"),Fj={kernelName:Mi,backendName:"cpu",kernelFunc:OT},LT=_s(e=>Math.expm1(e)),Aj=Qu(nu,LT),$j={kernelName:nu,backendName:"cpu",kernelFunc:Aj},zT=_s(e=>Math.floor(e)),Dj=Qu(Pi,zT),Rj={kernelName:Pi,backendName:"cpu",kernelFunc:Dj};function BT(e,t,n,a,r,s,i,o,l){let u=Ve([a,s],n);for(let p=0;p<a;p++){let d=[],c=0;for(let h=0;h<r;h++){let m=e[p*r+h];c+=m*i[h],d.push(m)}if(c<0||c>=l/s)throw new Error(`Invalid indices: ${d} does not index into ${o}`);for(let h=0;h<s;h++)u.values[p*s+h]=t.get(...t.indexToLoc(c*s+h))}return u}function WT(e,t,n){let a=Ve(n,e.dtype);for(let r=0;r<a.size;++r){let s=a.indexToLoc(r).slice(),i=s[0],o=s[2],l=t.locToIndex([i,o]);s[2]=t.values[l];let u=e.locToIndex(s);0<=u&&u<e.values.length&&(a.values[r]=e.values[u])}return a}var VT=Ut((e,t)=>e>t?1:0),Mj=sn(iu,VT,null,"bool"),Pj={kernelName:iu,backendName:"cpu",kernelFunc:Mj},UT=Ut((e,t)=>e>=t?1:0),Oj=sn(zi,UT,null,"bool"),Lj={kernelName:zi,backendName:"cpu",kernelFunc:Oj},GT=Ut((e,t)=>e<t?1:0),zj=sn(pu,GT,null,"bool"),Bj={kernelName:pu,backendName:"cpu",kernelFunc:zj},HT=Ut((e,t)=>e<=t?1:0),Wj=sn(cu,HT,null,"bool"),Vj={kernelName:cu,backendName:"cpu",kernelFunc:Wj};function jT(e,t,n){let a=(t-e)/(n-1),r=w.makeZerosTypedArray(n,"float32");r[0]=e;for(let s=1;s<r.length;s++)r[s]=r[s-1]+a;return r}var qT=_s(e=>Math.log(e)),Uj=Qu(Vi,qT),Gj={kernelName:Vi,backendName:"cpu",kernelFunc:Uj};function KT(e,t,n,a){let r=w.getTypedArrayFromDType(a,w.sizeFromShape(n));for(let s=0;s<r.length;++s){let i=s*t,o=e[i];for(let l=0;l<t;++l){let u=e[i+l];(Number.isNaN(u)||u>o)&&(o=u)}r[s]=o}return r}var XT=Ut((e,t)=>Math.max(e,t)),Hj=sn(Gi,XT),jj={kernelName:Gi,backendName:"cpu",kernelFunc:Hj},YT=Ut((e,t)=>Math.min(e,t)),qj=sn(Ki,YT),Kj={kernelName:Ki,backendName:"cpu",kernelFunc:qj},I0=Ut((e,t)=>e*t),Xj=v0((e,t,n,a)=>({real:e*n-t*a,imag:e*a+t*n})),ag=sn(Yi,I0,Xj),Yj={kernelName:Yi,backendName:"cpu",kernelFunc:ag};function QT(e,t,n){let a=w.createScalarValue(-1,n);return I0([],t,a,e,n)}function Qj(e){let{inputs:t,backend:n}=e,{x:a}=t;xe(a,"neg");let r=n.data.get(a.dataId).values,[s,i]=QT(r,a.shape,a.dtype);return n.makeTensorInfo(i,a.dtype,s)}var Jj={kernelName:fu,backendName:"cpu",kernelFunc:Qj},JT=Ut((e,t)=>e!==t?1:0),Zj=sn(gu,JT,null,"bool"),e5={kernelName:gu,backendName:"cpu",kernelFunc:Zj};function S0(e,t,n,a,r){let s=t.length,i=w.sizeFromShape(t),o=w.computeStrides(t),l=w.computeStrides(r),u=w.getTypedArrayFromDType(n,w.sizeFromShape(r));for(let p=0;p<i;++p){let d=w.indexToLoc(p,s,o),c=new Array(d.length);for(let m=0;m<c.length;m++)c[m]=d[a[m]];let h=w.locToIndex(c,s,l);u[h]=e[p]}return u}function Vn(e){let{inputs:t,attrs:n,backend:a}=e,{x:r}=t,{perm:s}=n;xe(r,"transpose");let i=r.shape.length,o=new Array(i);for(let p=0;p<o.length;p++)o[p]=r.shape[s[p]];let l=a.data.get(r.dataId).values,u=S0(l,r.shape,r.dtype,s,o);return{dataId:a.write(u,o,r.dtype),shape:o,dtype:r.dtype}}var t5={kernelName:Cr,backendName:"cpu",kernelFunc:Vn};function ZT(e,t,n,a){let[r,s]=C.computeOutAndReduceShapes(e,a),i=fa(t,"int32"),o=w.makeZerosTypedArray(w.sizeFromShape(r),i),l=w.sizeFromShape(s);for(let u=0;u<o.length;++u){let p=u*l,d=1;for(let c=0;c<l;++c)d*=n[p+c];o[u]=d}return{outVals:o,outShape:r,outDtype:i}}function n5(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{axis:s,keepDims:i}=a;xe(r,"prod");let o=r.shape.length,l=w.parseAxisParam(s,r.shape),u=C.getAxesPermutation(l,o),p=l,d=r,c=[];u!=null&&(d=Vn({inputs:{x:r},backend:n,attrs:{perm:u}}),c.push(d),p=C.getInnerMostAxes(p.length,o));let h=n.data.get(d.dataId).values,{outVals:m,outShape:f,outDtype:g}=ZT(d.shape,d.dtype,h,p),y=f;return i&&(y=C.expandShapeToKeepDim(f,l)),c.forEach(b=>n.disposeIntermediateTensorInfo(b)),n.makeTensorInfo(y,g,m)}var a5={kernelName:to,backendName:"cpu",kernelFunc:n5};function N0(e,t,n,a){let r=e===t,s=e<t&&n<0,i=t<e&&n>1;if(r||s||i)return w.makeZerosTypedArray(0,a);let o=Math.abs(Math.ceil((t-e)/n)),l=w.makeZerosTypedArray(o,a);t<e&&n===1&&(n=-1),l[0]=e;for(let u=1;u<l.length;u++)l[u]=l[u-1]+n;return l}var e_=_s(e=>1/Math.sqrt(e)),r5=Qu(oo,e_),s5={kernelName:oo,backendName:"cpu",kernelFunc:r5};function fl(e,t,n,a,r,s,i,o,l,u){let p=[a/r,r],d=e.values,c=t.values;if(a===0)return Ve(n,t.dtype);let h=Ve(p,t.dtype);typeof l=="string"||typeof l=="number"?h.values.fill(l):typeof l=="boolean"&&h.values.fill(+l);for(let m=0;m<s;m++){let f=[],g=0;for(let y=0;y<i;y++){let b=d[m*i+y];f.push(b),g+=b*o[y]}if(g<0||g>=a/r)throw new Error(`Invalid indices: ${f} does not index into ${n}`);for(let y=0;y<r;y++)u?h.values[g*r+y]+=c[m*r+y]:h.values[g*r+y]=t.rank===0?c[0]:c[m*r+y]}return h}var i5=_s(e=>1/(1+Math.exp(-e))),t_=ot(uo,e=>1/(1+Math.exp(-e))),o5={kernelName:uo,backendName:"cpu",kernelFunc:t_};function dm(e,t,n,a,r){let s=Kt.isSliceContinous(a,t,n),i=w.sizeFromShape(n),o=w.computeStrides(a);if(s){let d=Kt.computeFlatOffset(t,o);return r==="string"?e.slice(d,d+i):e.subarray(d,d+i)}let l=r==="string"?C.fromUint8ToStringArray(e):e,u=Ve(a,r,l),p=Ve(n,r);for(let d=0;d<p.size;++d){let c=p.indexToLoc(d),h=c.map((m,f)=>m+t[f]);p.set(u.get(...h),...c)}return r==="string"?C.fromStringArrayToUint8(p.values):p.values}function fi(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{begin:s,size:i}=a;xe(r,"slice");let[o,l]=Kt.parseSliceParams(r,s,i);Kt.assertParamsValid(r,o,l);let u=n.data.get(r.dataId).values,p=dm(u,o,l,r.shape,r.dtype);return n.makeTensorInfo(l,r.dtype,p)}var l5={kernelName:_u,backendName:"cpu",kernelFunc:fi};function n_(e,t,n,a,r,s,i){let o=t[0],l=s[0],u=new Array(l),p=new Array(o),d=t[1];if(l===0){if(o!==0)throw new Error(C.getSparseFillEmptyRowsIndicesDenseShapeMismatch(o));let g=w.getArrayFromDType(n,0),y=w.getArrayFromDType(r,0);return[g,[0,d],y,u,p]}let c=!0,h=0,m=new Array(l).fill(0);for(let g=0;g<o;++g){let y=e[g*d];if(y<0)throw new Error(C.getSparseFillEmptyRowsNegativeIndexErrorMessage(g,y));if(y>=l)throw new Error(C.getSparseFillEmptyRowsOutOfRangeIndexErrorMessage(g,y,l));++m[y],c=c&&y>=h,h=y}let f=!0;for(let g=0;g<l;++g){let y=m[g]===0;u[g]=y,f=f&&!y,m[g]=Math.max(m[g],1),g>0&&(m[g]+=m[g-1])}if(f&&c){let g=e,y=a;for(let b=0;b<o;++b)p[b]=b;return[g,[o,d],y,u,p]}else{let g=m[l-1],y=w.getArrayFromDType(n,g*d),b=w.getArrayFromDType(r,g),x=new Array(l).fill(0);for(let v=0;v<o;++v){let k=e[v*d],T=x[k],_=(k===0?0:m[k-1])+T;x[k]++;for(let E=0;E<d;++E)y[_*d+E]=e[v*d+E];b[_]=a[v],p[v]=_}for(let v=0;v<l;++v)if(x[v]===0){let k=v===0?0:m[v-1];y[k*d+0]=v;for(let T=1;T<d;++T)y[k*d+T]=0;b[k]=i}return[y,[g,d],b,u,p]}}function a_(e,t,n,a,r){let s=w.sizeFromShape(a),i=t[0],o=r.length,l=[],u=1,p=-1;for(let f=0;f<o;++f){let g=r[f];if(g===-1){if(p!==-1)throw new Error(C.getSparseReshapeMultipleNegativeOneOutputDimErrorMessage(p,f));p=f,l.push(1)}else{if(g<0)throw new Error(C.getSparseReshapeNegativeOutputDimErrorMessage(f,g));u*=g,l.push(g)}}if(p!==-1){if(u<=0)throw new Error(C.getSparseReshapeEmptyTensorZeroOutputDimErrorMessage());let f=Math.trunc(s/u);if(u*f!==s)throw new Error(C.getSparseReshapeInputOutputMultipleErrorMessage(a,l));l[p]=f}if(w.sizeFromShape(l)!==s)throw new Error(C.getSparseReshapeInputOutputMismatchErrorMessage(a,l));let d=a.length,c=[];if(d>0){c[d-1]=1;for(let f=d-2;f>=0;--f)c[f]=c[f+1]*a[f+1]}let h=[];if(o>0){h[o-1]=1;for(let f=o-2;f>=0;--f)h[f]=h[f+1]*l[f+1]}let m=w.getArrayFromDType(n,i*o);for(let f=0;f<i;++f){let g=0;for(let y=0;y<d;++y)g+=e[f*d+y]*c[y];for(let y=0;y<o;++y)m[f*o+y]=Math.trunc(g/h[y]),g%=h[y]}return[m,[i,o],l]}function T0(e,t,n,a,r,s=!1,i=0){let o=a.length,l=[t[0],e.length/t[0]],u=l[1],p=o>0?r[o-1]+1:0;if(p<0)throw new Error(C.getSparseSegmentReductionNegativeSegmentIdsErrorMessage());let d=t.slice();d[0]=p;let c=d.reduce((b,x)=>b*x,1),h=w.getArrayFromDType(n,c);if(o===0)return p>0&&h.fill(i),[h,d];if(p<=0)throw new Error(C.getSparseSegmentReductionNegativeSegmentIdsErrorMessage());let m=0,f=1,g=0,y=r[m];for(;;){let b=0;if(f<o){if(b=r[f],y===b){++f;continue}if(y>=b)throw new Error(C.getSparseSegmentReductionNonIncreasingSegmentIdsErrorMessage())}if(y<0||y>=p)throw new Error(C.getSparseSegmentReductionSegmentIdOutOfRangeErrorMessage(y,p));y>g&&h.fill(i,g*u,y*u);for(let x=m;x<f;++x){let v=a[x];if(v<0||v>=l[0])throw new Error(C.getSparseSegmentReductionIndicesOutOfRangeErrorMessage(x,a[x],l[0]));for(let k=0;k<u;k++)h[y*u+k]+=e[v*u+k]}if(s)for(let x=0;x<u;x++)h[y*u+x]/=f-m;if(m=f,++f,g=y+1,y=b,f>o)break}return g<p&&h.fill(i,g*u,p*u),[h,d]}var u5=_s(e=>Math.sqrt(e)),p5=ot(po,e=>Math.sqrt(e)),c5={kernelName:po,backendName:"cpu",kernelFunc:p5},r_=Ut((e,t)=>{let n=e-t;return n*n}),d5=sn(mo,r_),h5={kernelName:mo,backendName:"cpu",kernelFunc:d5};function s_(e,t,n,a){let r=Ve(e,t.dtype);for(let s=0;s<r.size;s++){let i=r.indexToLoc(s),o=new Array(i.length);for(let l=0;l<o.length;l++)o[l]=i[l]*n[l]+a[l];r.set(t.get(...o),...i)}return r}var m5=class{constructor(e,t,n,a,r,s){this.separator=w.encodeString(e),this.nGramWidths=t,this.leftPad=w.encodeString(n),this.rightPad=w.encodeString(a),this.padWidth=r,this.preserveShort=s}getPadWidth(e){return Math.min(this.padWidth<0?e-1:this.padWidth,e-1)}getNumNGrams(e,t){let n=this.getPadWidth(t);return Math.max(0,e+2*n-t+1)}createNGrams(e,t,n,a,r,s){for(let i=0;i<r;++i){let o=this.getPadWidth(s),l=Math.max(0,o-i),u=Math.max(0,o-(r-(i+1))),p=s-(l+u),d=t+(l>0?0:i-o),c=0;c+=l*this.leftPad.length;for(let g=0;g<p;++g)c+=e[d+g].length;c+=u*this.rightPad.length,c+=(l+u+p-1)*this.separator.length,n[a+i]=new Uint8Array(c);let h=n[a+i],m=0,f=g=>g.forEach(y=>h[m++]=y);for(let g=0;g<l;++g)f(this.leftPad),f(this.separator);for(let g=0;g<p-1;++g)f(e[d+g]),f(this.separator);if(p>0){f(e[d+p-1]);for(let g=0;g<u;++g)f(this.separator),f(this.rightPad)}else{for(let g=0;g<u-1;++g)f(this.rightPad),f(this.separator);f(this.rightPad)}}}compute(e,t){let n=e.length,a=t.length;if(a>0){let o=t[0];if(o!==0)throw new Error(`First split value must be 0, got ${o}`);for(let l=1;l<a;++l){let u=t[l]>=o;if(u=u&&t[l]<=n,!u)throw new Error(`Invalid split value ${t[l]}, must be in [${o}, ${n}]`);o=t[l]}if(o!==n)throw new Error(`Last split value must be data size. Expected ${n}, got ${o}`)}let r=a-1,s=w.getArrayFromDType("int32",a);if(n===0||a===0){let o=new Array(n);for(let l=0;l<=r;++l)s[l]=0;return[o,s]}s[0]=0;for(let o=1;o<=r;++o){let l=t[o]-t[o-1],u=0;this.nGramWidths.forEach(p=>{u+=this.getNumNGrams(l,p)}),this.preserveShort&&l>0&&u===0&&(u=1),s[o]=s[o-1]+u}let i=new Array(s[r]);for(let o=0;o<r;++o){let l=t[o],u=s[o];if(this.nGramWidths.forEach(p=>{let d=t[o+1]-t[o],c=this.getNumNGrams(d,p);this.createNGrams(e,l,i,u,c,p),u+=c}),this.preserveShort&&u===s[o]){let p=t[o+1]-t[o];if(p===0)continue;let d=p+2*this.padWidth,c=1;this.createNGrams(e,l,i,u,c,d)}}return[i,s]}};function i_(e,t,n,a,r,s,i,o){return new m5(n,a,r,s,i,o).compute(e,t)}function f5(e,t,n,a){if(!e.length)return;if(t.length===0){for(let s=0;s<e.length;++s)a.push(e.subarray(s,s+1));return}if(t.length===1){let s=t[0],i=e.indexOf(s);for(;i!==-1;){let o=e.subarray(0,i);(!n||o.length!==0)&&a.push(o),e=e.subarray(i+1),i=e.indexOf(s)}(!n||e.length!==0)&&a.push(e);return}let r=0;for(let s=0;s<e.length+1;s++)if(s===e.length||t.indexOf(e[s])!==-1){let i=e.subarray(r,s);(!n||i.length!==0)&&a.push(i),r=s+1}}function o_(e,t,n){let a=e.length,r=[],s=0,i=0,o=new Array(a);for(let c=0;c<a;++c){let h=r.length;f5(e[c],t,n,r);let m=r.length-h;o[c]=m,s+=m,i=Math.max(i,m)}let l=w.getArrayFromDType("int32",s*2),u=new Array(s),p=[a,i],d=0;for(let c=0;c<a;++c)for(let h=0;h<o[c];++h)l[d*2]=c,l[d*2+1]=h,u[d]=r[d],++d;return[l,u,p]}function l_(e,t){let n=w.getArrayFromDType("int32",e.length);for(let a=0;a<e.length;++a)n[a]=w.fingerPrint64(e[a]).modulo(t).getLowBitsUnsigned();return n}var u_=Ut((e,t)=>e-t),g5=v0((e,t,n,a)=>({real:e-n,imag:t-a})),_0=sn(fo,u_,g5),y5={kernelName:fo,backendName:"cpu",kernelFunc:_0};function p_(e,t){let n=new Array(e.rank);for(let r=0;r<n.length;r++)n[r]=e.shape[r]*t[r];let a=Ve(n,e.dtype);for(let r=0;r<a.values.length;++r){let s=a.indexToLoc(r),i=new Array(e.rank);for(let l=0;l<i.length;l++)i[l]=s[l]%e.shape[l];let o=e.locToIndex(i);a.values[r]=e.values[o]}return a}var Gp=(e,t)=>{let n=t.value-e.value;return n===0?e.index-t.index:n};function c_(e,t,n=0,a=e.length-1){for(;a>n;){if(a-n>600){let o=a-n+1,l=t-n+1,u=Math.log(o),p=.5*Math.exp(2*u/3),d=.5*Math.sqrt(u*p*(o-p)/o)*Math.sign(l-o/2),c=Math.max(n,Math.floor(t-l*p/o+d)),h=Math.min(a,Math.floor(t+(o-l)*p/o+d));c_(e,t,c,h)}let r=e[t],s=n,i=a;for(w.swap(e,n,t),Gp(e[a],r)>0&&w.swap(e,n,a);s<i;){for(w.swap(e,s,i),s++,i--;Gp(e[s],r)<0;)s=s+1;for(;Gp(e[i],r)>0;)i=i-1}Gp(e[n],r)===0?w.swap(e,n,i):(i=i+1,w.swap(e,i,a)),i<=t&&(n=i+1),t<=i&&(a=i-1)}}function d_(e,t,n,a,r){let s=t[t.length-1],[i,o]=[e.length/s,s],l=w.getTypedArrayFromDType(n,i*a),u=w.getTypedArrayFromDType("int32",i*a);for(let d=0;d<i;d++){let c=d*o,h=e.subarray(c,c+o),m=new Array(h.length);h.forEach((b,x)=>m[x]={value:b,index:x}),a<m.length&&(c_(m,a),m=m.slice(0,a)),r&&m.sort(Gp);let f=d*a,g=l.subarray(f,f+a),y=u.subarray(f,f+a);for(let b=0;b<a;b++)g[b]=m[b].value,y[b]=m[b].index}let p=t.slice();return p[p.length-1]=a,[Ve(p,n,l),Ve(p,"int32",u)]}function h_(e,t,n,a){let r=w.parseAxisParam(t,n)[0],s=[1,n[0],1];for(let m=0;m<r;m++)s[0]*=n[m];s[1]=n[r];for(let m=r+1;m<n.length;m++)s[2]*=n[m];let i={},o=new Int32Array(n[r]),l=new qt(s,a,e),u=[],p=s[0]===1&&s[2]===1;for(let m=0;m<n[r];m++){let f;if(p)f=e[m].toString();else{let g=[];for(let y=0;y<s[0];y++)for(let b=0;b<s[2];b++)g.push(l.get(y,m,b));f=g.join(",")}if(i[f]!==void 0)o[m]=i[f];else{let g=Object.keys(i).length;i[f]=g,o[m]=g,u.push(m)}}let d=s.slice();d[1]=Object.keys(i).length;let c=new qt(d,a);u.forEach((m,f)=>{for(let g=0;g<s[0];g++)for(let y=0;y<s[2];y++)c.set(l.get(g,m,y),g,f,y)});let h=n.slice();return h[r]=d[1],{outputValues:c.values,outputShape:h,indices:o}}nf("cpu",()=>new x0,1);var m_=ot(Ri,e=>e>=0?e:Math.exp(e)-1),b5={kernelName:Ri,backendName:"cpu",kernelFunc:m_};function f_(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{alpha:s}=a;xe([r],"leakyRelu");let i=w.sizeFromShape(r.shape),o=n.data.get(r.dataId).values,l=w.getTypedArrayFromDType("float32",i);for(let u=0;u<o.length;u++)l[u]=o[u]<0?s*o[u]:o[u];return n.makeTensorInfo(r.shape,"float32",l)}var x5={kernelName:Wi,backendName:"cpu",kernelFunc:f_},v5=Ut((e,t)=>e<0?t*e:e);function g_(e){let{inputs:t,backend:n}=e,{x:a,alpha:r}=t;xe([a,r],"prelu");let s=n.data.get(a.dataId).values,i=n.data.get(r.dataId).values,[o,l]=v5(a.shape,r.shape,s,i,"float32");return n.makeTensorInfo(l,"float32",o)}var w5={kernelName:eo,backendName:"cpu",kernelFunc:g_},y_=ot(no,e=>Math.max(0,e)),k5={kernelName:no,backendName:"cpu",kernelFunc:y_},b_=ot(ro,e=>Math.min(Math.max(0,e),6)),I5={kernelName:ro,backendName:"cpu",kernelFunc:b_};function hm(e,t,n,a,r){if(n==="linear")return hr({inputs:{x:t},backend:e});if(n==="relu")return y_({inputs:{x:t},backend:e});if(n==="elu")return m_({inputs:{x:t},backend:e});if(n==="relu6")return b_({inputs:{x:t},backend:e});if(n==="prelu")return g_({inputs:{x:t,alpha:a},backend:e});if(n==="leakyrelu")return f_({inputs:{x:t},backend:e,attrs:{alpha:r}});if(n==="sigmoid")return t_({inputs:{x:t},backend:e});throw new Error(`Activation ${n} has not been implemented for the CPU backend.`)}function xt(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{shape:s}=a,i=w.sizeFromShape(r.shape),o=w.inferFromImplicitShape(s,i),l=w.sizeFromShape(o);w.assert(i===l,()=>`The new shape (${o}) has ${l} elements and the old shape (${r.shape}) has ${i} elements. The new shape and old shape must have the same number of elements.`),n.incRef(r.dataId);let u=n.data.get(r.dataId);if(u.complexTensorInfos!=null){let p=u.complexTensorInfos.real,d=u.complexTensorInfos.imag;p.shape=o,d.shape=o}return{dataId:r.dataId,shape:o,dtype:r.dtype}}var S5={kernelName:Iu,backendName:"cpu",kernelFunc:xt};function x_(e){let{inputs:t,backend:n,attrs:a}=e,{a:r,b:s}=t,{transposeA:i,transposeB:o}=a;xe([r,s],"matMul");let l=r.shape.length,u=s.shape.length,p=i?r.shape[l-2]:r.shape[l-1],d=o?s.shape[u-1]:s.shape[u-2],c=i?r.shape[l-1]:r.shape[l-2],h=o?s.shape[u-2]:s.shape[u-1],m=r.shape.slice(0,-2),f=s.shape.slice(0,-2),g=w.sizeFromShape(m),y=w.sizeFromShape(f),b=Bu.assertAndGetBroadcastShape(r.shape.slice(0,-2),s.shape.slice(0,-2)).concat([c,h]);w.assert(p===d,()=>`Error in matMul: inner shapes (${p}) and (${d}) of Tensors with shapes ${r.shape} and ${s.shape} and transposeA=${i} and transposeB=${o} must match.`);let x=i?[g,p,c]:[g,c,p],v=o?[y,h,d]:[y,d,h],k=xt({inputs:{x:r},backend:n,attrs:{shape:x}}),T=xt({inputs:{x:s},backend:n,attrs:{shape:v}}),_=i?k.shape[1]:k.shape[2],E=i?k.shape[2]:k.shape[1],A=o?T.shape[1]:T.shape[2],M=Math.max(g,y),$=n.data.get(k.dataId).values,S=n.data.get(T.dataId).values,P=w.computeStrides(k.shape),V=w.computeStrides(T.shape),[j,q,K]=i?[P[0],1,P[1]]:[P[0],P[1],1],[Z,ee,re]=o?[1,V[1],V[0]]:[V[1],1,V[0]],Y=E*A,ie=Ve([M,E,A],k.dtype),ae=ie.values,le=n.blockSize;for(let ue=0;ue<M;ue++)for(let ke=0;ke<E;ke+=le)for(let ye=0;ye<A;ye+=le)for(let Ie=0;Ie<_;Ie+=le){let Ee=Math.min(ke+le,E),$e=Math.min(ye+le,A),Be=Math.min(Ie+le,_);for(let je=ke;je<Ee;je++)for(let st=ye;st<$e;st++){let nt=0;for(let at=Ie;at<Be;at++){let Te=Math.min(ue,g-1)*j,gt=Math.min(ue,y-1)*re,ct=$[Te+je*q+at*K],bn=S[at*Z+st*ee+gt];nt+=ct*bn}ae[ue*Y+(je*A+st)]+=nt}}return n.disposeIntermediateTensorInfo(k),n.disposeIntermediateTensorInfo(T),n.makeTensorInfo(b,ie.dtype,ie.values)}var N5={kernelName:Si,backendName:"cpu",kernelFunc:x_};function T5(e){let{inputs:t,backend:n,attrs:a}=e,{a:r,b:s,bias:i,preluActivationWeights:o}=t,{transposeA:l,transposeB:u,activation:p,leakyreluAlpha:d}=a,c,h,m,f=[];c=x_({inputs:{a:r,b:s},attrs:{transposeA:l,transposeB:u},backend:n}),i&&(h=Dl({inputs:{a:c,b:i},backend:n}),f.push(c),c=h),p&&(m=hm(n,c,p,o,d),f.push(c),c=m);for(let g of f)n.disposeIntermediateTensorInfo(g);return c}var _5={kernelName:ai,backendName:"cpu",kernelFunc:T5},C5=ot(zl,e=>Math.acos(e)),E5={kernelName:zl,backendName:"cpu",kernelFunc:C5},F5=ot(Bl,e=>Math.acosh(e)),A5={kernelName:Bl,backendName:"cpu",kernelFunc:F5};function $5(e){let{inputs:t,backend:n}=e,a=t;xe(t,"addN");let r=a.map(o=>n.data.get(o.dataId).values),s=Ve(a[0].shape,a[0].dtype),i=s.values;for(let o=0;o<a.length;o++){let l=r[o];for(let u=0;u<i.length;u++)i[u]+=l[u]}return n.makeTensorInfo(s.shape,s.dtype,s.values)}var D5={kernelName:wi,backendName:"cpu",kernelFunc:$5};function R5(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{axis:s,keepDims:i}=a;xe(r,"all");let o=w.parseAxisParam(s,r.shape),l=o,u=C.getAxesPermutation(l,r.shape.length),p=r;u!=null&&(p=Vn({inputs:{x:r},backend:n,attrs:{perm:u}}),l=C.getInnerMostAxes(l.length,r.shape.length)),C.assertAxesAreInnerMostDims("all",l,p.shape.length);let[d,c]=C.computeOutAndReduceShapes(p.shape,l),h=w.sizeFromShape(c),m=w.makeZerosTypedArray(w.sizeFromShape(d),p.dtype),f=n.data.get(p.dataId).values;for(let y=0;y<m.length;++y){let b=y*h,x=f[b];for(let v=0;v<h;++v){let k=f[b+v];x=x&&k}m[y]=x}u!=null&&n.disposeIntermediateTensorInfo(p);let g=n.makeTensorInfo(d,p.dtype,m);if(i){let y=C.expandShapeToKeepDim(d,o),b=xt({inputs:{x:g},backend:n,attrs:{shape:y}});return n.disposeIntermediateTensorInfo(g),b}return g}var M5={kernelName:Wl,backendName:"cpu",kernelFunc:R5};function P5(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{axis:s,keepDims:i}=a;xe(r,"any");let o=w.parseAxisParam(s,r.shape),l=o,u=C.getAxesPermutation(l,r.shape.length),p=r;u!=null&&(p=Vn({inputs:{x:r},backend:n,attrs:{perm:u}}),l=C.getInnerMostAxes(l.length,r.shape.length)),C.assertAxesAreInnerMostDims("any",l,p.shape.length);let[d,c]=C.computeOutAndReduceShapes(p.shape,l),h=w.sizeFromShape(c),m=w.makeZerosTypedArray(w.sizeFromShape(d),p.dtype),f=n.data.get(p.dataId).values;for(let y=0;y<m.length;++y){let b=y*h,x=f[b];for(let v=0;v<h;++v){let k=f[b+v];x=x||k}m[y]=x}u!=null&&n.disposeIntermediateTensorInfo(p);let g=n.makeTensorInfo(d,p.dtype,m);if(i){let y=C.expandShapeToKeepDim(d,o),b=xt({inputs:{x:g},backend:n,attrs:{shape:y}});return n.disposeIntermediateTensorInfo(g),b}return g}var O5={kernelName:Vl,backendName:"cpu",kernelFunc:P5};function L5(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{axis:s}=a;xe(r,"argMax");let i=w.parseAxisParam(s,r.shape),o=C.getAxesPermutation(i,r.shape.length),l=r,u=[];o!=null&&(l=Vn({inputs:{x:r},backend:n,attrs:{perm:o}}),u.push(l),i=C.getInnerMostAxes(i.length,l.shape.length)),i=[i[0]],C.assertAxesAreInnerMostDims("argMax",i,l.shape.length);let[p,d]=C.computeOutAndReduceShapes(l.shape,i),c=w.sizeFromShape(p),h=w.makeZerosTypedArray(c,"int32"),m=w.sizeFromShape(d),f=n.data.get(l.dataId).values;for(let g=0;g<h.length;++g){let y=g*m,b=f[y],x=0;for(let v=0;v<m;++v){let k=f[y+v];k>b&&(b=k,x=v)}h[g]=x}return u.forEach(g=>n.disposeIntermediateTensorInfo(g)),n.makeTensorInfo(p,"int32",h)}var z5={kernelName:ki,backendName:"cpu",kernelFunc:L5};function B5(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{axis:s}=a;xe(r,"argMin");let i=w.parseAxisParam(s,r.shape),o=C.getAxesPermutation(i,r.shape.length),l=r,u=[];o!=null&&(l=Vn({inputs:{x:r},backend:n,attrs:{perm:o}}),u.push(l),i=C.getInnerMostAxes(i.length,l.shape.length)),i=[i[0]],C.assertAxesAreInnerMostDims("argMin",i,l.shape.length);let[p,d]=C.computeOutAndReduceShapes(l.shape,i),c=w.sizeFromShape(p),h=w.makeZerosTypedArray(c,"int32"),m=w.sizeFromShape(d),f=n.data.get(l.dataId).values;for(let g=0;g<h.length;++g){let y=g*m,b=f[y],x=0;for(let v=0;v<m;++v){let k=f[y+v];k<b&&(b=k,x=v)}h[g]=x}return u.forEach(g=>n.disposeIntermediateTensorInfo(g)),n.makeTensorInfo(p,"int32",h)}var W5={kernelName:yc,backendName:"cpu",kernelFunc:B5},V5=ot(Ul,e=>Math.asin(e)),U5={kernelName:Ul,backendName:"cpu",kernelFunc:V5},G5=ot(Gl,e=>Math.asinh(e)),H5={kernelName:Gl,backendName:"cpu",kernelFunc:G5},j5=ot(Hl,e=>Math.atan(e)),q5={kernelName:Hl,backendName:"cpu",kernelFunc:j5},K5=Ut((e,t)=>Math.atan2(e,t)),X5=sn(ql,K5),Y5={kernelName:ql,backendName:"cpu",kernelFunc:X5},Q5=ot(jl,e=>Math.atanh(e)),J5={kernelName:jl,backendName:"cpu",kernelFunc:Q5};function C0(e,t,n,a,r,s){let i=r.strideHeight,o=r.strideWidth,l=r.dilationHeight,u=r.dilationWidth,p=r.effectiveFilterHeight,d=r.effectiveFilterWidth,c=r.padInfo.top,h=r.padInfo.left,m=s==="max"?Number.NEGATIVE_INFINITY:Number.POSITIVE_INFINITY,f=Ve(r.outShape,n),g=f.values,y=r.outShape[1]*r.outShape[2]*r.outShape[3],b=r.outShape[2]*r.outShape[3],x=r.outShape[3];for(let v=0;v<r.batchSize;++v){let k=v*y,T=v*a[0];for(let _=0;_<r.inChannels;++_)for(let E=0;E<r.outHeight;++E){let A=E*i-c,M=Math.max(0,A),$=Math.min(r.inHeight,p+A),S=k+E*b;for(let P=0;P<r.outWidth;++P){let V=P*o-h,j=Math.max(0,V),q=Math.min(r.inWidth,d+V),K=m,Z=0,ee=0;for(let Y=M;Y<$;Y+=l){let ie=T+Y*a[1];for(let ae=j;ae<q;ae+=u){let le=ie+ae*a[2],ue=e[le+_];s==="max"&&ue>K?K=ue:s==="avg"&&(Z+=ue,ee++)}if(isNaN(K))break}let re=S+P*x+_;g[re]=s==="avg"?Z/ee:K}}}return f}function v_(e,t,n,a,r=!1,s=!1){let i=Ve(a.outShape,"int32"),o=a.strideHeight,l=a.strideWidth,u=a.dilationHeight,p=a.dilationWidth,d=a.effectiveFilterHeight,c=a.effectiveFilterWidth,h=a.padInfo.top,m=a.padInfo.left,f=Ve(t,n,e);for(let g=0;g<a.batchSize;++g)for(let y=0;y<a.inChannels;++y)for(let b=0;b<a.outHeight;++b){let x=b*o-h,v=x;for(;v<0;)v+=u;let k=Math.min(a.inHeight,d+x);for(let T=0;T<a.outWidth;++T){let _=T*l-m,E=_;for(;E<0;)E+=p;let A=Math.min(a.inWidth,c+_),M=Number.NEGATIVE_INFINITY,$=-1;for(let S=v;S<k;S+=u){let P=S-x;for(let V=E;V<A;V+=p){let j=V-_,q=f.get(g,S,V,y);q>M&&(M=q,r?$=s?((g*a.inHeight+S)*a.inWidth+V)*a.inChannels+y:(S*a.inWidth+V)*a.inChannels+y:$=P*c+j)}}i.set($,g,b,T,y)}}return i}function w_(e,t,n,a,r,s){let i=r.strideDepth,o=r.strideHeight,l=r.strideWidth,u=r.dilationDepth,p=r.dilationHeight,d=r.dilationWidth,c=r.effectiveFilterDepth,h=r.effectiveFilterHeight,m=r.effectiveFilterWidth,f=r.padInfo.front,g=r.padInfo.top,y=r.padInfo.left,b=s==="max"?Number.NEGATIVE_INFINITY:Number.POSITIVE_INFINITY,x=Ve(r.outShape,n),v=x.values,k=r.outShape[1]*r.outShape[2]*r.outShape[3]*r.outShape[4],T=r.outShape[2]*r.outShape[3]*r.outShape[4],_=r.outShape[3]*r.outShape[4],E=r.outShape[4];for(let A=0;A<r.batchSize;++A){let M=A*k,$=A*a[0];for(let S=0;S<r.inChannels;++S)for(let P=0;P<r.outDepth;++P){let V=P*i-f,j=V;for(;j<0;)j+=u;let q=Math.min(r.inDepth,c+V),K=M+P*T;for(let Z=0;Z<r.outHeight;++Z){let ee=Z*o-g,re=ee;for(;re<0;)re+=p;let Y=Math.min(r.inHeight,h+ee),ie=K+Z*_;for(let ae=0;ae<r.outWidth;++ae){let le=ae*l-y,ue=le;for(;ue<0;)ue+=d;let ke=Math.min(r.inWidth,m+le),ye=ie+ae*E,Ie=b,Ee=0,$e=0;for(let je=j;je<q;je+=u){let st=$+je*a[1];for(let nt=re;nt<Y;nt+=p){let at=st+nt*a[2];for(let Te=ue;Te<ke;Te+=d){let gt=at+Te*a[3],ct=e[gt+S];if(s==="max"&&ct>Ie?Ie=ct:s==="avg"&&(Ee+=ct,$e++),isNaN(Ie))break}if(isNaN(Ie))break}if(isNaN(Ie))break}let Be=ye+S;v[Be]=s==="avg"?Ee/$e:Ie}}}}return x}function Z5(e,t){let n=Ve(t.outShape,"int32"),a=t.strideDepth,r=t.strideHeight,s=t.strideWidth,i=t.dilationDepth,o=t.dilationHeight,l=t.dilationWidth,u=t.effectiveFilterDepth,p=t.effectiveFilterHeight,d=t.effectiveFilterWidth,c=t.padInfo.front,h=t.padInfo.top,m=t.padInfo.left;for(let f=0;f<t.batchSize;++f)for(let g=0;g<t.inChannels;++g)for(let y=0;y<t.outDepth;++y){let b=y*a-c,x=b;for(;x<0;)x+=i;let v=Math.min(t.inDepth,u+b);for(let k=0;k<t.outHeight;++k){let T=k*r-h,_=T;for(;_<0;)_+=o;let E=Math.min(t.inHeight,p+T);for(let A=0;A<t.outWidth;++A){let M=A*s-m,$=M;for(;$<0;)$+=l;let S=Math.min(t.inWidth,d+M),P=Number.NEGATIVE_INFINITY,V=-1;for(let j=x;j<v;j+=i){let q=j-b;for(let K=_;K<E;K+=o){let Z=K-T;for(let ee=$;ee<S;ee+=l){let re=ee-M,Y=e.get(f,j,K,ee,g);Y>=P&&(P=Y,V=q*p*d+Z*p+re)}}}n.set(V,f,y,k,A,g)}}}return n}function eq(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t;xe(r,"avgPool");let{filterSize:s,strides:i,pad:o,dimRoundingMode:l}=a,u=1;w.assert(C.eitherStridesOrDilationsAreOne(i,u),()=>`Error in avgPool: Either strides or dilations must be 1. Got strides ${i} and dilations '${u}'`);let p=C.computePool2DInfo(r.shape,s,i,u,o,l),d;if(p.filterWidth===1&&p.filterHeight===1&&w.arraysEqual(p.inShape,p.outShape))d=hr({inputs:{x:r},backend:n});else{let c=n.data.get(r.dataId).values,h=w.computeStrides(r.shape),m=C0(c,r.shape,r.dtype,h,p,"avg");d=n.makeTensorInfo(p.outShape,r.dtype,m.values)}return d}var tq={kernelName:Ii,backendName:"cpu",kernelFunc:eq};function nq(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{filterSize:s,strides:i,pad:o,dimRoundingMode:l,dataFormat:u}=a;xe(r,"avgPool3d");let p=C.computePool3DInfo(r.shape,s,i,1,o,l,u),d=n.data.get(r.dataId).values,c=w_(d,r.shape,r.dtype,w.computeStrides(r.shape),p,"avg");return n.makeTensorInfo(c.shape,"float32",c.values)}var aq={kernelName:bc,backendName:"cpu",kernelFunc:nq};function rq(e){let{inputs:t,backend:n,attrs:a}=e,{dy:r,input:s}=t,{filterSize:i,strides:o,pad:l,dimRoundingMode:u}=a;xe([r,s],"avgPool3DGrad");let p=C.computePool3DInfo(s.shape,i,o,1,l,u),d=p.strideDepth,c=p.strideHeight,h=p.strideWidth,m=p.filterDepth,f=p.filterHeight,g=p.filterWidth,y=p.dilationDepth,b=p.dilationHeight,x=p.dilationWidth,v=p.effectiveFilterDepth,k=p.effectiveFilterHeight,T=p.effectiveFilterWidth,_=v-1-p.padInfo.front,E=T-1-p.padInfo.left,A=k-1-p.padInfo.top,M=Ve(s.shape,"float32"),$=1/(m*f*g),S=n.bufferSync(r);for(let P=0;P<p.batchSize;++P)for(let V=0;V<p.inChannels;++V)for(let j=0;j<p.inDepth;++j)for(let q=0;q<p.inHeight;++q)for(let K=0;K<p.inWidth;++K){let Z=j-_,ee=q-A,re=K-E,Y=0;for(let ie=0;ie<v;ie+=y){let ae=(Z+ie)/d;if(!(ae<0||ae>=p.outDepth||Math.floor(ae)!==ae))for(let le=0;le<k;le+=b){let ue=(ee+le)/c;if(!(ue<0||ue>=p.outHeight||Math.floor(ue)!==ue))for(let ke=0;ke<T;ke+=x){let ye=(re+ke)/h;ye<0||ye>=p.outWidth||Math.floor(ye)!==ye||(Y+=S.get(P,ae,ue,ye,V))}}}M.set(Y*$,P,j,q,K,V)}return n.makeTensorInfo(M.shape,M.dtype,M.values)}var sq={kernelName:Im,backendName:"cpu",kernelFunc:rq};function iq(e){let{inputs:t,backend:n,attrs:a}=e,{dy:r,input:s}=t,i=s;xe([r,s],"avgPoolGrad");let{filterSize:o,strides:l,pad:u}=a,p=C.computePool2DInfo(i.shape,o,l,1,u),d=p.strideHeight,c=p.strideWidth,h=p.filterHeight,m=p.filterWidth,f=p.dilationHeight,g=p.dilationWidth,y=p.effectiveFilterHeight,b=p.effectiveFilterWidth,x=b-1-p.padInfo.left,v=y-1-p.padInfo.top,k=Ve(i.shape,"float32"),T=1/(h*m),_=n.data.get(r.dataId).values,E=Ve(r.shape,"float32",_);for(let A=0;A<p.batchSize;++A)for(let M=0;M<p.inChannels;++M)for(let $=0;$<p.inHeight;++$)for(let S=0;S<p.inWidth;++S){let P=$-v,V=S-x,j=0;for(let q=0;q<y;q+=f){let K=(P+q)/d;if(!(K<0||K>=p.outHeight||Math.floor(K)!==K))for(let Z=0;Z<b;Z+=g){let ee=(V+Z)/c;ee<0||ee>=p.outWidth||Math.floor(ee)!==ee||(j+=E.get(A,K,ee,M))}}k.set(j*T,A,$,S,M)}return n.makeTensorInfo(k.shape,k.dtype,k.values)}var oq={kernelName:km,backendName:"cpu",kernelFunc:iq};function lq(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,scale:s,offset:i,mean:o,variance:l}=t;w.assert(o.shape.length===l.shape.length,()=>"Batch normalization gradient requires mean and variance to have equal ranks."),w.assert(i==null||o.shape.length===i.shape.length,()=>"Batch normalization gradient requires mean and offset to have equal ranks."),w.assert(s==null||o.shape.length===s.shape.length,()=>"Batch normalization gradient requires mean and scale to have equal ranks."),xe([r,o,l,s,i],"batchNorm");let{varianceEpsilon:u}=a;u==null&&(u=.001);let p=n.data.get(r.dataId).values,d=n.data.get(o.dataId).values,c=n.data.get(l.dataId).values,h=s?n.data.get(s.dataId).values:new Float32Array([1]),m=i?n.data.get(i.dataId).values:new Float32Array([0]),f=new Float32Array(p.length),g=m.length,y=h.length,b=c.length,x=d.length,v=0,k=0,T=0,_=0;for(let E=0;E<p.length;++E)f[E]=m[v++]+(p[E]-d[k++])*h[T++]/Math.sqrt(c[_++]+u),v>=g&&(v=0),k>=x&&(k=0),T>=y&&(T=0),_>=b&&(_=0);return n.makeTensorInfo(r.shape,r.dtype,f)}var uq={kernelName:Li,backendName:"cpu",kernelFunc:lq};function pq(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{blockShape:s,crops:i}=a;xe([r],"batchToSpaceND");let o=s.reduce((y,b)=>y*b),l=C.getReshaped(r.shape,s,o),u=C.getPermuted(l.length,s.length),p=C.getReshapedPermuted(r.shape,s,o),d=C.getSliceBeginCoords(i,s.length),c=C.getSliceSize(p,i,s.length),h=xt({inputs:{x:r},backend:n,attrs:{shape:l}}),m=Vn({inputs:{x:h},backend:n,attrs:{perm:u}}),f=xt({inputs:{x:m},backend:n,attrs:{shape:p}}),g=fi({inputs:{x:f},backend:n,attrs:{begin:d,size:c}});return n.disposeIntermediateTensorInfo(h),n.disposeIntermediateTensorInfo(m),n.disposeIntermediateTensorInfo(f),g}var cq={kernelName:Kl,backendName:"cpu",kernelFunc:pq};function dq(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,weights:s}=t,{size:i}=a,o=n.data.get(r.dataId).values,l=n.data.get(s.dataId).values,u=w0(o,l,s.dtype,s.shape,i);return n.makeTensorInfo([i],s.dtype,u)}var hq={kernelName:Sm,backendName:"cpu",kernelFunc:dq};function mq(e){let{inputs:t,backend:n}=e,{s0:a,s1:r}=t,s=n.data.get(a.dataId).values,i=n.data.get(r.dataId).values,o=C.assertAndGetBroadcastShape(Array.from(s),Array.from(i));return n.makeTensorInfo([o.length],"int32",Int32Array.from(o))}var fq={kernelName:Nm,backendName:"cpu",kernelFunc:mq},gq=ot(xs,(e,t)=>{let n=t;return e>n.clipValueMax?n.clipValueMax:e<n.clipValueMin?n.clipValueMin:e}),yq={kernelName:xs,backendName:"cpu",kernelFunc:gq},bq=e=>{let{x:t}=e.inputs,n=e.backend,a=new Float32Array(w.sizeFromShape(t.shape)),r=n.data.get(t.dataId),s=r.complexTensorInfos.real,i=r.complexTensorInfos.imag,o=n.data.get(s.dataId).values,l=n.data.get(i.dataId).values;for(let u=0;u<o.length;u++){let p=o[u],d=l[u];a[u]=Math.hypot(p,d)}return n.makeOutput(a,t.shape,"float32")},xq={kernelName:xc,backendName:"cpu",kernelFunc:bq};function Rl(e){let{inputs:t,backend:n}=e,{input:a}=t,r=n.data.get(a.dataId).complexTensorInfos.imag,s=n.data.get(r.dataId).values;return n.makeTensorInfo(r.shape,r.dtype,s)}var vq={kernelName:Lm,backendName:"cpu",kernelFunc:Rl};function Ml(e){let{inputs:t,backend:n,attrs:a}=e,{axis:r}=a,s=w.parseAxisParam(r,t[0].shape)[0],i=C.computeOutShape(t.map(f=>f.shape),s);if(w.sizeFromShape(i)===0)return n.makeTensorInfo(i,t[0].dtype,[]);let o=t.filter(f=>w.sizeFromShape(f.shape)>0);if(o.length===1)return hr({inputs:{x:o[0]},backend:n});let l=o.map(f=>f.shape);if(C.assertParamsConsistent(l,s),o[0].dtype==="complex64"){let f=o.map(v=>mi({inputs:{input:v},backend:n})),g=o.map(v=>Rl({inputs:{input:v},backend:n})),y=Ml({inputs:f,backend:n,attrs:{axis:s}}),b=Ml({inputs:g,backend:n,attrs:{axis:s}}),x=Yn({inputs:{real:y,imag:b},backend:n});return f.forEach(v=>n.disposeIntermediateTensorInfo(v)),g.forEach(v=>n.disposeIntermediateTensorInfo(v)),n.disposeIntermediateTensorInfo(y),n.disposeIntermediateTensorInfo(b),x}let u=o.map(f=>{let g=w.sizeFromShape(f.shape.slice(s));return xt({inputs:{x:f},backend:n,attrs:{shape:[-1,g]}})}),p=u.map(f=>({vals:n.data.get(f.dataId).values,shape:f.shape}));i=C.computeOutShape(u.map(f=>f.shape),1);let d=u[0].shape[0]===1,c=k0(p,i,t[0].dtype,d),h=C.computeOutShape(o.map(f=>f.shape),s),m=n.makeTensorInfo(h,t[0].dtype,c);return u.forEach(f=>n.disposeIntermediateTensorInfo(f)),m}var wq={kernelName:Xl,backendName:"cpu",kernelFunc:Ml};function k_(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,filter:s}=t,{strides:i,pad:o,dataFormat:l,dilations:u,dimRoundingMode:p}=a;xe([r,s],"conv2d");let d=C.convertConv2DDataFormat(l),c=C.computeConv2DInfo(r.shape,s.shape,i,u,o,p,!1,d),h=c.filterHeight,m=c.filterWidth,f=c.dilationHeight,g=c.dilationWidth,y=c.padInfo.left,b=c.padInfo.top,x=c.dataFormat==="channelsLast",v=new qt(c.outShape,r.dtype),k=w.computeStrides(r.shape),T=w.computeStrides(s.shape),_=k[0],E=x?k[1]:k[2],A=x?k[2]:1,M=x?1:k[1],$=v.strides[0],S=x?v.strides[1]:v.strides[2],P=x?v.strides[2]:1,V=x?1:v.strides[1],j=n.data.get(r.dataId).values,q=n.data.get(s.dataId).values,K=v.values;for(let Z=0;Z<c.batchSize;++Z){let ee=Z*_,re=Z*$;for(let Y=0;Y<c.outHeight;++Y){let ie=re+Y*S,ae=Y*c.strideHeight-b;for(let le=0;le<h;++le){let ue=ae+le*f;if(ue<0||ue>=c.inHeight)continue;let ke=le*T[0],ye=ee+ue*E;for(let Ie=0;Ie<c.outWidth;++Ie){let Ee=ie+Ie*P,$e=Ie*c.strideWidth-y;for(let Be=0;Be<m;++Be){let je=$e+Be*g;if(je<0||je>=c.inWidth)continue;let st=ke+Be*T[1],nt=ye+je*A,at=st;for(let Te=0;Te<c.inChannels;++Te){let gt=j[nt+Te*M];for(let ct=0;ct<c.outChannels;++ct)K[Ee+ct*V]+=gt*q[at+ct];at+=c.outChannels}}}}}}return n.makeTensorInfo(v.shape,v.dtype,K)}var kq={kernelName:_i,backendName:"cpu",kernelFunc:k_};function Iq(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,dy:s}=t,{strides:i,pad:o,dataFormat:l,dimRoundingMode:u,filterShape:p}=a;xe([r,s],"conv2dBackpropFilter");let d=C.convertConv2DDataFormat(l),c=C.computeConv2DInfo(r.shape,p,i,1,o,u,!1,d),{strideHeight:h,strideWidth:m,filterHeight:f,filterWidth:g}=c,y=c.dataFormat==="channelsLast",b=new qt(c.filterShape,"float32"),x=c.padInfo.left,v=c.padInfo.top,k=n.data.get(r.dataId).values,T=n.data.get(s.dataId).values,_=new qt(r.shape,r.dtype,k),E=new qt(s.shape,s.dtype,T);for(let A=0;A<f;++A){let M=Math.max(0,Math.ceil((v-A)/h)),$=Math.min(c.outHeight,(c.inHeight+v-A)/h);for(let S=0;S<g;++S){let P=Math.max(0,Math.ceil((x-S)/m)),V=Math.min(c.outWidth,(c.inWidth+x-S)/m);for(let j=0;j<c.inChannels;++j)for(let q=0;q<c.outChannels;++q){let K=0;for(let Z=0;Z<c.batchSize;++Z)for(let ee=M;ee<$;++ee){let re=A+ee*h-v;for(let Y=P;Y<V;++Y){let ie=S+Y*m-x;y?K+=_.get(Z,re,ie,j)*E.get(Z,ee,Y,q):K+=_.get(Z,j,re,ie)*E.get(Z,q,ee,Y)}}b.set(K,A,S,j,q)}}}return n.makeTensorInfo(b.shape,b.dtype,b.values)}var Sq={kernelName:_m,backendName:"cpu",kernelFunc:Iq};function Nq(e){let{inputs:t,backend:n,attrs:a}=e,{dy:r,filter:s}=t,{inputShape:i,strides:o,pad:l,dataFormat:u,dimRoundingMode:p}=a;xe([r,s],"conv2dBackpropInput");let d=w.computeStrides(s.shape),c=w.computeStrides(r.shape),h=C.convertConv2DDataFormat(u),m=C.computeConv2DInfo(i,s.shape,o,1,l,p,!1,h),f=new qt(m.inShape,"float32"),g=f.values,y=n.data.get(r.dataId).values,b=n.data.get(s.dataId).values,[x,v,k]=d,{batchSize:T,filterHeight:_,filterWidth:E,inChannels:A,inHeight:M,inWidth:$,outChannels:S,outHeight:P,outWidth:V,strideHeight:j,strideWidth:q}=m;h=m.dataFormat;let K=_-1-m.padInfo.top,Z=E-1-m.padInfo.left,ee=h==="channelsLast",re=f.strides[0],Y=ee?f.strides[1]:f.strides[2],ie=ee?f.strides[2]:1,ae=ee?1:f.strides[1],le=c[0],ue=ee?c[1]:c[2],ke=ee?c[2]:1,ye=ee?1:c[1];for(let Ie=0;Ie<T;++Ie)for(let Ee=0;Ee<A;++Ee)for(let $e=0;$e<M;++$e){let Be=$e-K,je=Math.max(0,Math.ceil(Be/j)),st=Math.min(P,(_+Be)/j);for(let nt=0;nt<$;++nt){let at=nt-Z,Te=Math.max(0,Math.ceil(at/q)),gt=Math.min(V,(E+at)/q),ct=0;for(let Qt=je;Qt<st;++Qt){let Dn=Qt*j-Be;for(let Gt=Te;Gt<gt;++Gt){let Jt=Gt*q-at,Da=le*Ie+ue*Qt+ke*Gt,Rn=x*(_-1-Dn)+v*(E-1-Jt)+k*Ee;for(let Ht=0;Ht<S;++Ht){let ia=y[Da+ye*Ht],oa=b[Rn+Ht];ct+=ia*oa}}}let bn=re*Ie+Y*$e+ie*nt+ae*Ee;g[bn]=ct}}return n.makeTensorInfo(f.shape,f.dtype,f.values)}var Tq={kernelName:Ci,backendName:"cpu",kernelFunc:Nq};function _q(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,filter:s}=t,{strides:i,pad:o,dilations:l}=a;xe([r,s],"conv3d");let u=C.computeConv3DInfo(r.shape,s.shape,i,l,o),{filterDepth:p,filterHeight:d,filterWidth:c,dilationDepth:h,dilationHeight:m,dilationWidth:f,padInfo:g}=u,y=g.front,b=g.left,x=g.top,v=new qt(u.outShape,r.dtype),k=n.data.get(r.dataId).values,T=n.data.get(s.dataId).values,_=v.values,E=w.computeStrides(r.shape),A=w.computeStrides(s.shape);for(let M=0;M<u.batchSize;++M){let $=M*E[0],S=M*v.strides[0];for(let P=0;P<u.outDepth;++P){let V=S+P*v.strides[1],j=P*u.strideDepth-y;for(let q=0;q<p;++q){let K=j+q*h;if(K<0||K>=u.inDepth)continue;let Z=q*A[0],ee=$+K*E[1];for(let re=0;re<u.outHeight;++re){let Y=V+re*v.strides[2],ie=re*u.strideHeight-x;for(let ae=0;ae<d;++ae){let le=ie+ae*m;if(le<0||le>=u.inHeight)continue;let ue=Z+ae*A[1],ke=ee+le*E[2];for(let ye=0;ye<u.outWidth;++ye){let Ie=Y+ye*u.outChannels,Ee=ye*u.strideWidth-b;for(let $e=0;$e<c;++$e){let Be=Ee+$e*f;if(Be<0||Be>=u.inWidth)continue;let je=ue+$e*A[2],st=ke+Be*u.inChannels,nt=je;for(let at=0;at<u.inChannels;++at){let Te=k[st+at];for(let gt=0;gt<u.outChannels;++gt)_[Ie+gt]+=Te*T[nt+gt];nt+=u.outChannels}}}}}}}}return n.makeTensorInfo(v.shape,v.dtype,v.values)}var Cq={kernelName:vc,backendName:"cpu",kernelFunc:_q};function Eq(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,dy:s}=t,{strides:i,pad:o,filterShape:l}=a;xe([r,s],"conv3dBackpropFilterV2");let u=w.computeStrides(r.shape),p=w.computeStrides(s.shape),d=C.computeConv3DInfo(r.shape,l,i,1,o),c=d.strideDepth,h=d.strideHeight,m=d.strideWidth,f=d.filterDepth,g=d.filterHeight,y=d.filterWidth,b=new qt(d.filterShape,"float32"),x=b.values,[v,k,T,_]=b.strides,E=n.data.get(s.dataId).values,[A,M,$,S]=p,P=n.data.get(r.dataId).values,[V,j,q,K]=u,Z=d.padInfo.front,ee=d.padInfo.left,re=d.padInfo.top;for(let Y=0;Y<f;++Y){let ie=Math.max(0,Math.ceil((Z-Y)/c)),ae=Math.min(d.outDepth,(d.inDepth+Z-Y)/c),le=Y*v;for(let ue=0;ue<g;++ue){let ke=Math.max(0,Math.ceil((re-ue)/h)),ye=Math.min(d.outHeight,(d.inHeight+re-ue)/h),Ie=ue*k+le;for(let Ee=0;Ee<y;++Ee){let $e=Math.max(0,Math.ceil((ee-Ee)/m)),Be=Math.min(d.outWidth,(d.inWidth+ee-Ee)/m),je=Ee*T+Ie;for(let st=0;st<d.inChannels;++st){let nt=st*_+je;for(let at=0;at<d.outChannels;++at){let Te=0;for(let gt=0;gt<d.batchSize;++gt){let ct=gt*V,bn=gt*A;for(let Qt=ie;Qt<ae;++Qt){let Dn=(Y+Qt*c-Z)*j+ct,Gt=Qt*M+bn;for(let Jt=ke;Jt<ye;++Jt){let Da=(ue+Jt*h-re)*q+Dn,Rn=Jt*$+Gt;for(let Ht=$e;Ht<Be;++Ht){let ia=(Ee+Ht*m-ee)*K+Da,oa=Ht*S+Rn;Te+=P[ia+st]*E[oa+at]}}}}x[nt+at]=Te}}}}}return n.makeTensorInfo(b.shape,b.dtype,b.values)}var Fq={kernelName:Cm,backendName:"cpu",kernelFunc:Eq};function Aq(e){let{inputs:t,backend:n,attrs:a}=e,{dy:r,filter:s}=t,{pad:i,strides:o,inputShape:l}=a;xe([r],"conv3dBackpropInputV2");let u=w.computeStrides(r.shape),p=w.computeStrides(s.shape),d=C.computeConv3DInfo(l,s.shape,o,1,i),c=new qt(d.inShape,"float32"),h=c.values,[m,f,g,y]=c.strides,b=n.data.get(r.dataId).values,[x,v,k,T]=u,_=n.data.get(s.dataId).values,[E,A,M,$]=p,{batchSize:S,filterDepth:P,filterHeight:V,filterWidth:j,inChannels:q,inDepth:K,inHeight:Z,inWidth:ee,outChannels:re,outDepth:Y,outHeight:ie,outWidth:ae,strideDepth:le,strideHeight:ue,strideWidth:ke}=d,ye=P-1-d.padInfo.front,Ie=V-1-d.padInfo.top,Ee=j-1-d.padInfo.left;for(let $e=0;$e<S;++$e)for(let Be=0;Be<q;++Be)for(let je=0;je<K;++je){let st=je-ye,nt=Math.max(0,Math.ceil(st/le)),at=Math.min(Y,(P+st)/le);for(let Te=0;Te<Z;++Te){let gt=Te-Ie,ct=Math.max(0,Math.ceil(gt/ue)),bn=Math.min(ie,(V+gt)/ue);for(let Qt=0;Qt<ee;++Qt){let Dn=Qt-Ee,Gt=Math.max(0,Math.ceil(Dn/ke)),Jt=Math.min(ae,(j+Dn)/ke),Da=0;for(let Rn=nt;Rn<at;++Rn){let Ht=Rn*le-st;for(let ia=ct;ia<bn;++ia){let oa=ia*ue-gt;for(let qr=Gt;qr<Jt;++qr){let Ms=qr*ke-Dn,Pd=x*$e+v*Rn+k*ia+T*qr,Kr=E*(P-1-Ht)+A*(V-1-oa)+M*(j-1-Ms)+$*Be;for(let kr=0;kr<re;++kr){let kp=b[Pd+kr],Yo=_[Kr+kr];Da+=kp*Yo}}}}h[m*$e+f*je+g*Te+y*Qt+Be]=Da}}}return n.makeTensorInfo(c.shape,c.dtype,c.values)}var $q={kernelName:Em,backendName:"cpu",kernelFunc:Aq},Dq=ot(Ei,e=>Math.cos(e)),Rq={kernelName:Ei,backendName:"cpu",kernelFunc:Dq},Mq=ot(Fi,e=>Math.cosh(e)),Pq={kernelName:Fi,backendName:"cpu",kernelFunc:Mq};function Oq(e){let{inputs:t,backend:n,attrs:a}=e,{image:r,boxes:s,boxInd:i}=t,{cropSize:o,method:l,extrapolationValue:u}=a,[p,d,c,h]=r.shape,m=s.shape[0],[f,g]=o,y=Ve([m,f,g,h],"float32"),b=n.data.get(s.dataId).values,x=n.data.get(i.dataId).values,v=n.data.get(r.dataId).values,k=w.computeStrides(r.shape),T=w.computeStrides(y.shape);for(let _=0;_<m;_++){let E=_*4,A=b[E],M=b[E+1],$=b[E+2],S=b[E+3],P=x[_];if(P>=p)continue;let V=f>1?($-A)*(d-1)/(f-1):0,j=g>1?(S-M)*(c-1)/(g-1):0;for(let q=0;q<f;q++){let K=f>1?A*(d-1)+q*V:.5*(A+$)*(d-1);if(K<0||K>d-1){for(let Z=0;Z<g;Z++)for(let ee=0;ee<h;ee++){let re=ee+Z*T[2]+q*T[1]+_*T[0];y.values[re]=u}continue}if(l==="bilinear"){let Z=Math.floor(K),ee=Math.ceil(K),re=K-Z;for(let Y=0;Y<g;Y++){let ie=g>1?M*(c-1)+Y*j:.5*(M+S)*(c-1);if(ie<0||ie>c-1){for(let ke=0;ke<h;ke++){let ye=ke+Y*T[2]+q*T[1]+_*T[0];y.values[ye]=u}continue}let ae=Math.floor(ie),le=Math.ceil(ie),ue=ie-ae;for(let ke=0;ke<h;ke++){let ye=ke+ae*k[2]+Z*k[1]+P*k[0],Ie=v[ye];ye=ke+le*k[2]+Z*k[1]+P*k[0];let Ee=v[ye];ye=ke+ae*k[2]+ee*k[1]+P*k[0];let $e=v[ye];ye=ke+le*k[2]+ee*k[1]+P*k[0];let Be=v[ye],je=Ie+(Ee-Ie)*ue,st=$e+(Be-$e)*ue;ye=ke+Y*T[2]+q*T[1]+_*T[0],y.values[ye]=je+(st-je)*re}}}else for(let Z=0;Z<g;++Z){let ee=g>1?M*(c-1)+Z*j:.5*(M+S)*(c-1);if(ee<0||ee>c-1){for(let ie=0;ie<h;ie++){let ae=ie+Z*T[2]+q*T[1]+_*T[0];y.values[ae]=u}continue}let re=Math.round(ee),Y=Math.round(K);for(let ie=0;ie<h;ie++){let ae=ie+re*k[2]+Y*k[1]+P*k[0],le=ie+Z*T[2]+q*T[1]+_*T[0];y.values[le]=v[ae]}}}}return n.makeTensorInfo(y.shape,y.dtype,y.values)}var Lq={kernelName:Ql,backendName:"cpu",kernelFunc:Oq};function zq(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{axis:s,exclusive:i,reverse:o}=a;xe(r,"cumprod");let l=C.getAxesPermutation([s],r.shape.length),u=r;l!=null&&(u=Vn({inputs:{x:r},backend:n,attrs:{perm:l}}));let p=C.getInnerMostAxes(1,r.shape.length)[0];if(p!==u.shape.length-1)throw new Error(`backend.cumprod in CPU expects an inner-most axis=${u.shape.length-1} but got axis=${p}`);let d=fa(u.dtype,"int32"),c=w.makeOnesTypedArray(w.sizeFromShape(u.shape),d),h=n.data.get(u.dataId).values,m=u.shape[u.shape.length-1],f=o?(y,b)=>y+m-b-1:(y,b)=>y+b;for(let y=0;y<h.length;y+=m)for(let b=0;b<m;b++){let x=f(y,b);if(b===0)c[x]=i?1:h[x];else{let v=f(y,b-1);c[x]=i?h[v]*c[v]:h[x]*c[v]}}let g=n.makeTensorInfo(u.shape,d,c);if(l!=null){let y=C.getUndoAxesPermutation(l),b=Vn({inputs:{x:g},backend:n,attrs:{perm:y}});return n.disposeIntermediateTensorInfo(g),n.disposeIntermediateTensorInfo(u),b}return g}var Bq={kernelName:Yl,backendName:"cpu",kernelFunc:zq};function Wq(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{axis:s,exclusive:i,reverse:o}=a;xe(r,"cumsum");let l=C.getAxesPermutation([s],r.shape.length),u=r;l!=null&&(u=Vn({inputs:{x:r},backend:n,attrs:{perm:l}}));let p=C.getInnerMostAxes(1,r.shape.length)[0];if(p!==u.shape.length-1)throw new Error(`backend.cumsum in CPU expects an inner-most axis=${u.shape.length-1} but got axis=${p}`);let d=fa(u.dtype,"int32"),c=w.makeZerosTypedArray(w.sizeFromShape(u.shape),d),h=n.data.get(u.dataId).values,m=u.shape[u.shape.length-1],f=o?(y,b)=>y+m-b-1:(y,b)=>y+b;for(let y=0;y<h.length;y+=m)for(let b=0;b<m;b++){let x=f(y,b);if(b===0)c[x]=i?0:h[x];else{let v=f(y,b-1);c[x]=i?h[v]+c[v]:h[x]+c[v]}}let g=n.makeTensorInfo(u.shape,d,c);if(l!=null){let y=C.getUndoAxesPermutation(l),b=Vn({inputs:{x:g},backend:n,attrs:{perm:y}});return n.disposeIntermediateTensorInfo(g),n.disposeIntermediateTensorInfo(u),b}return g}var Vq={kernelName:Ai,backendName:"cpu",kernelFunc:Wq};function Uq(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,weights:s}=t,{size:i,binaryOutput:o}=a;if(r.shape.length===1){let l=n.data.get(r.dataId).values,u=n.data.get(s.dataId).values,p=w0(l,u,s.dtype,s.shape,i);return n.makeTensorInfo([i],s.dtype,p)}else if(r.shape.length===2){let l=n.bufferSync(r),u=n.bufferSync(s),p=$T(l,u,i,o);return n.makeTensorInfo(p.shape,s.dtype,p.values)}throw new Error(`Error in denseBincount: input must be at most rank 2, but got rank${r.shape.length}.`)}var Gq={kernelName:Fm,backendName:"cpu",kernelFunc:Uq};function Hq(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{blockSize:s,dataFormat:i}=a;w.assert(i==="NHWC",()=>`Only NHWC dataFormat supported on CPU for depthToSpace. Got ${i}`);let o=r.shape[0],l=r.shape[1],u=r.shape[2],p=r.shape[3],d=l*s,c=u*s,h=p/(s*s),m=n.data.get(r.dataId).values,f=new Float32Array(o*d*c*h),g=0;for(let y=0;y<o;++y)for(let b=0;b<d;++b){let x=Math.floor(b/s),v=b%s;for(let k=0;k<c;++k){let T=Math.floor(k/s),_=k%s,E=(v*s+_)*h;for(let A=0;A<h;++A){let M=A+E+p*(T+u*(x+l*y));f[g++]=m[M]}}}return n.makeTensorInfo([o,d,c,h],r.dtype,f)}var jq={kernelName:Jl,backendName:"cpu",kernelFunc:Hq};function I_(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,filter:s}=t,{strides:i,pad:o,dilations:l,dimRoundingMode:u}=a;xe([r,s],"depthwiseConv2DNative");let p=w.computeStrides(r.shape),d=w.computeStrides(s.shape),c=l;c==null&&(c=[1,1]),w.assert(C.eitherStridesOrDilationsAreOne(i,c),()=>`Error in depthwiseConv2d: Either strides or dilations must be 1. Got strides ${i} and dilations '${c}'`);let h=C.computeConv2DInfo(r.shape,s.shape,i,c,o,u,!0),{filterHeight:m,filterWidth:f,dilationHeight:g,dilationWidth:y,padInfo:b}=h,x=b.left,v=b.top,k=h.outChannels/h.inChannels,T=new qt(h.outShape,r.dtype),_=n.data.get(r.dataId).values,E=n.data.get(s.dataId).values,A=T.values;for(let M=0;M<h.batchSize;++M){let $=M*p[0],S=M*T.strides[0];for(let P=0;P<h.outHeight;++P){let V=S+P*T.strides[1],j=P*h.strideHeight-v;for(let q=0;q<m;++q){let K=j+q*g;if(K<0||K>=h.inHeight)continue;let Z=q*d[0],ee=$+K*p[1];for(let re=0;re<h.outWidth;++re){let Y=V+re*T.strides[2],ie=re*h.strideWidth-x;for(let ae=0;ae<f;++ae){let le=ie+ae*y;if(le<0||le>=h.inWidth)continue;let ue=Z+ae*d[1],ke=ee+le*h.inChannels,ye=Y,Ie=ue;for(let Ee=0;Ee<h.inChannels;++Ee){let $e=_[ke+Ee];for(let Be=0;Be<k;++Be)A[ye+Be]+=$e*E[Ie+Be];ye+=k,Ie+=k}}}}}}return n.makeTensorInfo(T.shape,T.dtype,T.values)}var qq={kernelName:$i,backendName:"cpu",kernelFunc:I_};function Kq(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,dy:s}=t,{strides:i,dilations:o,pad:l,dimRoundingMode:u,filterShape:p}=a;xe([r,s],"depthwiseConv2dNativeBackpropFilter");let d=C.computeConv2DInfo(r.shape,p,i,o,l,u,!0),{strideHeight:c,strideWidth:h,filterHeight:m,filterWidth:f}=d,g=new qt(d.filterShape,"float32"),y=d.padInfo.left,b=d.padInfo.top,x=d.outChannels/d.inChannels,v=n.data.get(r.dataId).values,k=new qt(r.shape,r.dtype,v),T=n.data.get(s.dataId).values,_=new qt(s.shape,s.dtype,T);for(let E=0;E<m;++E){let A=Math.max(0,Math.ceil((b-E)/c)),M=Math.min(d.outHeight,(d.inHeight+b-E)/c);for(let $=0;$<f;++$){let S=Math.max(0,Math.ceil((y-$)/h)),P=Math.min(d.outWidth,(d.inWidth+y-$)/h);for(let V=0;V<d.outChannels;++V){let j=Math.trunc(V/x),q=V%x,K=0;for(let Z=0;Z<d.batchSize;++Z)for(let ee=A;ee<M;++ee){let re=E+ee*c-b;for(let Y=S;Y<P;++Y){let ie=$+Y*h-y;K+=k.get(Z,re,ie,j)*_.get(Z,ee,Y,V)}}g.set(K,E,$,j,q)}}}return n.makeTensorInfo(g.shape,g.dtype,g.values)}var Xq={kernelName:Am,backendName:"cpu",kernelFunc:Kq};function Yq(e){let{inputs:t,backend:n,attrs:a}=e,{dy:r,filter:s}=t,{strides:i,dilations:o,pad:l,dimRoundingMode:u,inputShape:p}=a;xe([r,s],"depthwiseConv2DNativeBackpropInput");let d=w.computeStrides(r.shape),c=w.computeStrides(s.shape),h=C.computeConv2DInfo(p,s.shape,i,o,l,u,!0),m=new qt(h.inShape,"float32"),f=m.values,[g,y,b]=m.strides,x=n.data.get(r.dataId).values,[v,k,T]=d,_=n.data.get(s.dataId).values,[E,A,M]=c,{batchSize:$,filterHeight:S,filterWidth:P,inChannels:V,inHeight:j,inWidth:q,outChannels:K,outHeight:Z,outWidth:ee,strideHeight:re,strideWidth:Y}=h,ie=S-1-h.padInfo.top,ae=P-1-h.padInfo.left,le=K/V;for(let ue=0;ue<$;++ue)for(let ke=0;ke<V;++ke)for(let ye=0;ye<j;++ye){let Ie=ye-ie,Ee=Math.max(0,Math.ceil(Ie/re)),$e=Math.min(Z,(S+Ie)/re);for(let Be=0;Be<q;++Be){let je=Be-ae,st=Math.max(0,Math.ceil(je/Y)),nt=Math.min(ee,(P+je)/Y),at=0;for(let Te=Ee;Te<$e;++Te){let gt=Te*re-Ie;for(let ct=st;ct<nt;++ct){let bn=ct*Y-je,Qt=v*ue+k*Te+T*ct,Dn=E*(S-1-gt)+A*(P-1-bn)+M*ke;for(let Gt=0;Gt<le;++Gt){let Jt=ke*le+Gt,Da=x[Qt+Jt],Rn=_[Dn+Gt];at+=Da*Rn}}}f[g*ue+y*ye+b*Be+ke]=at}}return n.makeTensorInfo(m.shape,m.dtype,m.values)}var Qq={kernelName:$m,backendName:"cpu",kernelFunc:Yq};function Jq(e){let{inputs:t,backend:n}=e,{x:a}=t,r=w.sizeFromShape(a.shape),s=n.data.get(a.dataId).values,i=Ve([r,r],a.dtype),o=i.values;for(let u=0;u<s.length;u++)o[u*r+u]=s[u];let l=[...a.shape,...a.shape];return n.makeTensorInfo(l,i.dtype,i.values)}var Zq={kernelName:Dm,backendName:"cpu",kernelFunc:Jq},e8={kernelName:wc,backendName:"cpu",kernelFunc:({inputs:e,backend:t,attrs:n})=>{let{x:a,filter:r}=e,{strides:s,pad:i,dilations:o}=n,l=t,u=l.data.get(a.dataId).values,p=a.shape.length,d=l.data.get(r.dataId).values,c=r.shape.length,{batchSize:h,inHeight:m,inWidth:f,inChannels:g,outHeight:y,outWidth:b,padInfo:x,strideHeight:v,strideWidth:k,filterHeight:T,filterWidth:_,dilationHeight:E,dilationWidth:A,outShape:M}=C.computeDilation2DInfo(a.shape,r.shape,s,i,"NHWC",o),$=w.sizeFromShape(M),S=M.length,P=w.getArrayFromDType(a.dtype,$);for(let V=0;V<h;++V)for(let j=0;j<y;++j){let q=j*v-x.top;for(let K=0;K<b;++K){let Z=K*k-x.left;for(let ee=0;ee<g;++ee){let re=Number.MIN_SAFE_INTEGER;for(let ie=0;ie<T;++ie){let ae=q+ie*E;if(ae>=0&&ae<m)for(let le=0;le<_;++le){let ue=Z+le*A;if(ue>=0&&ue<f){let ke=w.locToIndex([V,ae,ue,ee],p,w.computeStrides(a.shape)),ye=w.locToIndex([ie,le,ee],c,w.computeStrides(r.shape)),Ie=u[ke]+d[ye];Ie>re&&(re=Ie)}}}let Y=w.locToIndex([V,j,K,ee],S,w.computeStrides(M));P[Y]=re}}}return{dataId:l.write(w.toTypedArray(P,a.dtype),M,a.dtype),shape:M,dtype:a.dtype}}},t8={kernelName:Hh,backendName:"cpu",kernelFunc:({inputs:e,backend:t,attrs:n})=>{let{x:a,filter:r,dy:s}=e,{strides:i,pad:o,dilations:l}=n,u=t,p=w.toNestedArray(a.shape,u.data.get(a.dataId).values),d=w.toNestedArray(r.shape,u.data.get(r.dataId).values),{batchSize:c,inHeight:h,inWidth:m,inChannels:f,outHeight:g,outWidth:y,padInfo:b,strideHeight:x,strideWidth:v,filterHeight:k,filterWidth:T,dilationHeight:_,dilationWidth:E,outShape:A}=C.computeDilation2DInfo(a.shape,r.shape,i,o,"NHWC",l);w.assert(s.rank===A.length,()=>`Error in ${Hh}, dy must have the same rank as output ${A.length}, but got ${s.rank}`);let M=w.toNestedArray(A,u.data.get(s.dataId).values),$=w.makeZerosNestedTypedArray(r.shape,r.dtype);for(let S=0;S<c;++S)for(let P=0;P<g;++P){let V=P*x-b.top;for(let j=0;j<y;++j){let q=j*v-b.left;for(let K=0;K<f;++K){let Z=Number.MIN_SAFE_INTEGER,ee=0,re=0;for(let Y=0;Y<k;++Y){let ie=V+Y*_;if(ie>=0&&ie<h)for(let ae=0;ae<T;++ae){let le=q+ae*E;if(le>=0&&le<m){let ue=p[S][ie][le][K]+d[Y][ae][K];ue>Z&&(Z=ue,ee=Y,re=ae)}}}$[ee][re][K]+=M[S][P][j][K]}}}return{dataId:u.write(w.toTypedArray($,a.dtype),r.shape,r.dtype),shape:r.shape,dtype:r.dtype}}},n8={kernelName:Gh,backendName:"cpu",kernelFunc:({inputs:e,backend:t,attrs:n})=>{let{x:a,filter:r,dy:s}=e,{strides:i,pad:o,dilations:l}=n,u=t,p=w.toNestedArray(a.shape,u.data.get(a.dataId).values),d=w.toNestedArray(r.shape,u.data.get(r.dataId).values),{batchSize:c,inHeight:h,inWidth:m,inChannels:f,outHeight:g,outWidth:y,padInfo:b,strideHeight:x,strideWidth:v,filterHeight:k,filterWidth:T,dilationHeight:_,dilationWidth:E,outShape:A}=C.computeDilation2DInfo(a.shape,r.shape,i,o,"NHWC",l);w.assert(s.rank===A.length,()=>`Error in ${Gh}, dy must have the same rank as output ${A.length}, but got ${s.rank}`);let M=w.toNestedArray(A,u.data.get(s.dataId).values),$=w.makeZerosNestedTypedArray(a.shape,a.dtype);for(let S=0;S<c;++S)for(let P=0;P<g;++P){let V=P*x-b.top;for(let j=0;j<y;++j){let q=j*v-b.left;for(let K=0;K<f;++K){let Z=Number.MIN_SAFE_INTEGER,ee=V<0?0:V,re=q<0?0:q;for(let Y=0;Y<k;++Y){let ie=V+Y*_;if(ie>=0&&ie<h)for(let ae=0;ae<T;++ae){let le=q+ae*E;if(le>=0&&le<m){let ue=p[S][ie][le][K]+d[Y][ae][K];ue>Z&&(Z=ue,ee=ie,re=le)}}}$[S][ee][re][K]+=M[S][P][j][K]}}}return{dataId:u.write(w.toTypedArray($,a.dtype),a.shape,a.dtype),shape:a.shape,dtype:a.dtype}}};function ud(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{axis:s,keepDims:i}=a;xe(r,"sum");let o;r.dtype==="bool"?o=fs({inputs:{x:r},backend:n,attrs:{dtype:"int32"}}):o=hr({inputs:{x:r},backend:n});let l=o.shape.length,u=w.parseAxisParam(s,o.shape),p=C.getAxesPermutation(u,l),d=u,c=o;p!=null&&(c=Vn({inputs:{x:o},backend:n,attrs:{perm:p}}),d=C.getInnerMostAxes(d.length,l)),C.assertAxesAreInnerMostDims("sum",d,c.shape.length);let[h,m]=C.computeOutAndReduceShapes(c.shape,d),f=C.upcastType(c.dtype,"int32"),g=cm(n,h,f),y=w.sizeFromShape(m),b=n.data.get(g.dataId).values,x=n.data.get(c.dataId).values;for(let v=0;v<b.length;++v){let k=v*y,T=0;for(let _=0;_<y;++_)T+=x[k+_];b[v]=T}if(i){let v=C.expandShapeToKeepDim(g.shape,u),k=g;g=xt({inputs:{x:g},backend:n,attrs:{shape:v}}),n.disposeIntermediateTensorInfo(k)}return n.disposeIntermediateTensorInfo(o),p!=null&&n.disposeIntermediateTensorInfo(c),g}var a8={kernelName:co,backendName:"cpu",kernelFunc:ud};function r8(e){let{inputs:t,backend:n,attrs:a}=e,{equation:r}=a,s=t,{allDims:i,summedDims:o,idDims:l}=C.decodeEinsumEquation(r,s.length);C.checkEinsumDimSizes(i.length,l,s);let{path:u,steps:p}=C.getEinsumComputePath(o,l),d=p.length,c=null,h=i.length,m=[];for(let f=0;f<d;++f){for(let g of p[f]){let{permutationIndices:y,expandDims:b}=C.getEinsumPermutation(h,l[g]),x;C.isIdentityPermutation(y)?x=s[g]:(x=Vn({inputs:{x:s[g]},backend:n,attrs:{perm:y}}),m.push(x));let v=x.shape.slice();for(let k=0;k<b.length;++k)v.splice(b[k],0,1);w.arraysEqual(x.shape,v)||(x=xt({inputs:{x},backend:n,attrs:{shape:v}}),m.push(x)),c===null?c=x:(c=ag({inputs:{a:x,b:c},backend:n}),m.push(c))}f<d-1&&(u[f]>=0&&(c=ud({inputs:{x:c},backend:n,attrs:{axis:u[f]-(i.length-h),keepDims:!1}}),m.push(c)),h--)}for(let f of m)f!==c&&n.disposeIntermediateTensorInfo(f);return c}var s8={kernelName:Rm,backendName:"cpu",kernelFunc:r8};function i8(e){let{inputs:t,backend:n}=e,{dy:a,y:r}=t;xe([a,r],"eluGrad");let s=new Float32Array(w.sizeFromShape(r.shape)),i=n.data.get(r.dataId).values,o=n.data.get(a.dataId).values;for(let l=0;l<i.length;++l){let u=i[l];u>=1?s[l]=o[l]:s[l]=o[l]*(u+1)}return n.makeTensorInfo(r.shape,"float32",s)}var o8={kernelName:Mm,backendName:"cpu",kernelFunc:i8},l8=C.ERF_P,u8=C.ERF_A1,p8=C.ERF_A2,c8=C.ERF_A3,d8=C.ERF_A4,h8=C.ERF_A5,m8=ot(Zl,e=>{let t=Math.sign(e),n=Math.abs(e),a=1/(1+l8*n);return t*(1-((((h8*a+d8)*a+c8)*a+p8)*a+u8)*a*Math.exp(-n*n))}),f8={kernelName:Zl,backendName:"cpu",kernelFunc:m8};function mm(e){let{inputs:t,backend:n,attrs:a}=e,{input:r}=t,{dim:s}=a,i=r.shape.length,o=r.shape.slice(),l=s;return s<0&&(w.assert(-(i+1)<=s,()=>`Axis must be in the interval [${-(i+1)}, ${i}]`),l=i+s+1),o.splice(l,0,1),xt({inputs:{x:r},backend:n,attrs:{shape:o}})}var g8={kernelName:tu,backendName:"cpu",kernelFunc:mm},y8=Ut((e,t)=>e/t),E0=sn(Di,y8),fx={kernelName:Di,backendName:"cpu",kernelFunc:E0};function S_(e,t,n){let a=e.shape,r=a[0],s=a[1],i=n.data.get(e.dataId),o=i.complexTensorInfos.real,l=i.complexTensorInfos.imag,u=[r,s],p=w.sizeFromShape(u),d=w.getTypedArrayFromDType("float32",p),c=w.getTypedArrayFromDType("float32",p);for(let g=0;g<r;g++){let y=fi({inputs:{x:o},backend:n,attrs:{begin:[g,0],size:[1,s]}}),b=fi({inputs:{x:l},backend:n,attrs:{begin:[g,0],size:[1,s]}}),x=Yn({inputs:{real:y,imag:b},backend:n}),{real:v,imag:k}=b8(x,t,n),T=C.mergeRealAndImagArrays(v,k);for(let _=0;_<s;_++){let E=C.getComplexWithIndex(T,_);d[g*s+_]=E.real,c[g*s+_]=E.imag}n.disposeIntermediateTensorInfo(y),n.disposeIntermediateTensorInfo(b),n.disposeIntermediateTensorInfo(x)}let h=n.makeTensorInfo(u,"float32",d),m=n.makeTensorInfo(u,"float32",c),f=Yn({inputs:{real:h,imag:m},backend:n});return n.disposeIntermediateTensorInfo(h),n.disposeIntermediateTensorInfo(m),f}function b8(e,t,n){let a=w.sizeFromShape(e.shape),r=n.data.get(e.dataId),s=n.data.get(r.complexTensorInfos.real.dataId).values,i=n.data.get(r.complexTensorInfos.imag.dataId).values;if(x8(a)){let o=gx(s,i,a,t,n),l=[e.shape[0],e.shape[1]];if(t){let u=n.makeTensorInfo(l,"float32",o.real),p=n.makeTensorInfo(l,"float32",o.imag),d=n.makeTensorInfo([],"float32",w.createScalarValue(a,"float32")),c=hr({inputs:{x:d},backend:n}),h=fx.kernelFunc({inputs:{a:u,b:d},backend:n}),m=fx.kernelFunc({inputs:{a:p,b:c},backend:n}),f=n.data.get(h.dataId).values,g=n.data.get(m.dataId).values;return n.disposeIntermediateTensorInfo(u),n.disposeIntermediateTensorInfo(p),n.disposeIntermediateTensorInfo(d),n.disposeIntermediateTensorInfo(c),n.disposeIntermediateTensorInfo(h),n.disposeIntermediateTensorInfo(m),{real:f,imag:g}}return o}else{let o=C.mergeRealAndImagArrays(s,i),l=v8(o,a,t);return C.splitRealAndImagArrays(l)}}function x8(e){return(e&e-1)===0}function gx(e,t,n,a,r){if(n===1)return{real:e,imag:t};let s=C.mergeRealAndImagArrays(e,t),i=n/2,o=C.complexWithEvenIndex(s),l=o.real,u=o.imag,p=[l.length],d=r.makeTensorInfo(p,"float32",l),c=r.makeTensorInfo(p,"float32",u),h=Yn({inputs:{real:d,imag:c},backend:r}),m=C.complexWithOddIndex(s),f=m.real,g=m.imag,y=[f.length],b=r.makeTensorInfo(y,"float32",f),x=r.makeTensorInfo(y,"float32",g),v=Yn({inputs:{real:b,imag:x},backend:r}),k=gx(l,u,i,a,r),T=k.real,_=k.imag,E=[T.length],A=r.makeTensorInfo(E,"float32",T),M=r.makeTensorInfo(E,"float32",_),$=Yn({inputs:{real:A,imag:M},backend:r}),S=gx(f,g,i,a,r),P=S.real,V=S.imag,j=[P.length],q=r.makeTensorInfo(j,"float32",P),K=r.makeTensorInfo(j,"float32",V),Z=Yn({inputs:{real:q,imag:K},backend:r}),ee=C.exponents(n,a),re=[ee.real.length],Y=r.makeTensorInfo(re,"float32",ee.real),ie=r.makeTensorInfo(re,"float32",ee.imag),ae=Yn({inputs:{real:Y,imag:ie},backend:r}),le=ag({inputs:{a:ae,b:Z},backend:r}),ue=Dl({inputs:{a:$,b:le},backend:r}),ke=_0({inputs:{a:$,b:le},backend:r}),ye=mi({inputs:{input:ue},backend:r}),Ie=mi({inputs:{input:ke},backend:r}),Ee=Rl({inputs:{input:ue},backend:r}),$e=Rl({inputs:{input:ke},backend:r}),Be=Ml({inputs:[ye,Ie],backend:r,attrs:{axis:0}}),je=Ml({inputs:[Ee,$e],backend:r,attrs:{axis:0}}),st=r.data.get(Be.dataId).values,nt=r.data.get(je.dataId).values;return r.disposeIntermediateTensorInfo(d),r.disposeIntermediateTensorInfo(c),r.disposeIntermediateTensorInfo(h),r.disposeIntermediateTensorInfo(b),r.disposeIntermediateTensorInfo(x),r.disposeIntermediateTensorInfo(v),r.disposeIntermediateTensorInfo(A),r.disposeIntermediateTensorInfo(M),r.disposeIntermediateTensorInfo($),r.disposeIntermediateTensorInfo(q),r.disposeIntermediateTensorInfo(K),r.disposeIntermediateTensorInfo(Z),r.disposeIntermediateTensorInfo(Y),r.disposeIntermediateTensorInfo(ie),r.disposeIntermediateTensorInfo(ae),r.disposeIntermediateTensorInfo(le),r.disposeIntermediateTensorInfo(ue),r.disposeIntermediateTensorInfo(ke),r.disposeIntermediateTensorInfo(ye),r.disposeIntermediateTensorInfo(Ee),r.disposeIntermediateTensorInfo(Ie),r.disposeIntermediateTensorInfo($e),r.disposeIntermediateTensorInfo(Be),r.disposeIntermediateTensorInfo(je),{real:st,imag:nt}}function v8(e,t,n){let a=new Float32Array(t*2);for(let r=0;r<t;r++){let s=0,i=0;for(let o=0;o<t;o++){let l=C.exponent(r*o,t,n),u=C.getComplexWithIndex(e,o);s+=u.real*l.real-u.imag*l.imag,i+=u.real*l.imag+u.imag*l.real}n&&(s/=t,i/=t),C.assignToTypedArray(a,s,i,r)}return a}function w8(e){let{inputs:t,backend:n}=e,{input:a}=t,r=w.sizeFromShape(a.shape),s=a.shape[a.shape.length-1],i=r/s,o=xt({inputs:{x:a},backend:n,attrs:{shape:[i,s]}}),l=S_(o,!1,n),u=xt({inputs:{x:l},backend:n,attrs:{shape:a.shape}});return n.disposeIntermediateTensorInfo(o),n.disposeIntermediateTensorInfo(l),u}var k8={kernelName:Pm,backendName:"cpu",kernelFunc:w8};function F0(e){let{backend:t,attrs:n}=e,{shape:a,value:r,dtype:s}=n,i=s||w.inferDtype(r),o=w.getArrayFromDType(i,w.sizeFromShape(a));return S8(o,r,i),t.makeTensorInfo(a,i,o)}var I8={kernelName:kc,backendName:"cpu",kernelFunc:F0};function S8(e,t,n){e.fill(t)}var N8={kernelName:au,backendName:"cpu",kernelFunc:({inputs:e,attrs:t,backend:n})=>{let{image:a}=e,r=n,s=w.getTypedArrayFromDType(a.dtype,w.sizeFromShape(a.shape)),[i,o,l,u]=a.shape,p=r.data.get(a.dataId).values;for(let d=0;d<i;d++){let c=d*l*o*u;for(let h=0;h<o;h++){let m=h*(l*u);for(let f=0;f<l;f++){let g=f*u;for(let y=0;y<u;y++){let b=Math.round(l-f-1),x=c+m+g+y,v=p[x];if(b>=0&&b<l){let k=b*u,T=c+m+k+y;v=p[T]}s[x]=v}}}}return{dataId:r.write(s,a.shape,a.dtype),shape:a.shape,dtype:a.dtype}}},T8=Ut((e,t)=>Math.floor(e/t)),_8=sn(Oi,T8,null,"int32"),C8={kernelName:Oi,backendName:"cpu",kernelFunc:_8};function E8(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,filter:s,bias:i,preluActivationWeights:o}=t,{strides:l,pad:u,dataFormat:p,dilations:d,dimRoundingMode:c,activation:h,leakyreluAlpha:m}=a,f=k_({inputs:{x:r,filter:s},backend:n,attrs:{strides:l,pad:u,dataFormat:p,dilations:d,dimRoundingMode:c}});if(i){let g=f;if(p==="NCHW"&&i.shape.length===1&&i.shape[0]!==1){let y=xt({inputs:{x:i},backend:n,attrs:{shape:[i.shape[0],1,1]}});f=Dl({inputs:{a:f,b:y},backend:n}),n.disposeIntermediateTensorInfo(y)}else f=Dl({inputs:{a:f,b:i},backend:n});n.disposeIntermediateTensorInfo(g)}if(h){let g=f;if(p==="NCHW"&&h==="prelu"&&o.shape.length===1&&o.shape[0]!==1){let y=xt({inputs:{x:o},backend:n,attrs:{shape:[o.shape[0],1,1]}});f=hm(n,f,h,y,m),n.disposeIntermediateTensorInfo(y)}else f=hm(n,f,h,o,m);n.disposeIntermediateTensorInfo(g)}return f}var F8={kernelName:ri,backendName:"cpu",kernelFunc:E8};function A8(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,filter:s,bias:i,preluActivationWeights:o}=t,{strides:l,pad:u,dataFormat:p,dilations:d,dimRoundingMode:c,activation:h,leakyreluAlpha:m}=a,f=I_({inputs:{x:r,filter:s},backend:n,attrs:{strides:l,pad:u,dataFormat:p,dilations:d,dimRoundingMode:c}});if(i){let g=f;f=Dl({inputs:{a:f,b:i},backend:n}),n.disposeIntermediateTensorInfo(g)}if(h){let g=f;f=hm(n,f,h,o,m),n.disposeIntermediateTensorInfo(g)}return f}var $8={kernelName:si,backendName:"cpu",kernelFunc:A8};function D8(e){let{inputs:t,backend:n}=e,{params:a,indices:r}=t,s=w.sizeFromShape(a.shape),i=r.shape,o=i[i.length-1],[l,u,p,d]=C.prepareAndValidate(a,r);if(u===0)return n.makeTensorInfo(l,a.dtype,[]);let c=n.data.get(r.dataId).values,h=n.bufferSync(a),m=BT(c,h,a.dtype,u,o,p,d,a.shape,s);return n.makeTensorInfo(l,a.dtype,m.values)}var R8={kernelName:su,backendName:"cpu",kernelFunc:D8};function M8(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,indices:s}=t,{axis:i,batchDims:o}=a;xe([r,s],"gatherV2");let l=w.parseAxisParam(i,r.shape)[0],u=n.data.get(s.dataId).values,p=r.shape[l];for(let v=0;v<u.length;++v){let k=u[v];w.assert(k<=p-1&&k>=0,()=>`GatherV2: the index value ${k} is not in [0, ${p-1}]`)}let d=o;o==null&&(d=0);let c=w.sizeFromShape(s.shape),h=C.segment_util.collectGatherOpShapeInfo(r,s,l,d),m=xt({inputs:{x:r},backend:n,attrs:{shape:[h.batchSize,h.outerSize,h.dimSize,h.sliceSize]}}),f=xt({inputs:{x:s},backend:n,attrs:{shape:[h.batchSize,c/h.batchSize]}}),g=[h.batchSize,h.outerSize,c/h.batchSize,h.sliceSize],y=n.bufferSync(f),b=n.bufferSync(m),x=WT(b,y,g);return n.disposeIntermediateTensorInfo(m),n.disposeIntermediateTensorInfo(f),n.makeTensorInfo(h.outputShape,x.dtype,x.values)}var P8={kernelName:ru,backendName:"cpu",kernelFunc:M8};function O8(e){let{inputs:t,backend:n}=e,{input:a}=t,r=w.sizeFromShape(a.shape),s=a.shape[a.shape.length-1],i=r/s,o=xt({inputs:{x:a},backend:n,attrs:{shape:[i,s]}}),l=S_(o,!0,n),u=xt({inputs:{x:l},backend:n,attrs:{shape:a.shape}});return n.disposeIntermediateTensorInfo(o),n.disposeIntermediateTensorInfo(l),u}var L8={kernelName:Om,backendName:"cpu",kernelFunc:O8},z8=ot(ou,e=>Number.isFinite(e)?1:0,"bool"),B8={kernelName:ou,backendName:"cpu",kernelFunc:z8},W8=ot(lu,e=>Math.abs(e)===1/0?1:0,"bool"),V8={kernelName:lu,backendName:"cpu",kernelFunc:W8},U8=ot(uu,e=>Number.isNaN(e)?1:0,"bool"),G8={kernelName:uu,backendName:"cpu",kernelFunc:U8};function H8(e){let{backend:t,attrs:n}=e,{start:a,stop:r,num:s}=n,i=jT(a,r,s);return t.makeTensorInfo([i.length],"float32",i)}var j8={kernelName:zm,backendName:"cpu",kernelFunc:H8},q8=ot(du,e=>Math.log1p(e)),K8={kernelName:du,backendName:"cpu",kernelFunc:q8},X8=Ut((e,t)=>e&&t),Y8=sn(hu,X8,null,"bool"),Q8={kernelName:hu,backendName:"cpu",kernelFunc:Y8},J8=ot(Ic,e=>e?0:1,"bool"),Z8={kernelName:Ic,backendName:"cpu",kernelFunc:J8},eK=Ut((e,t)=>e||t),tK=sn(Sc,eK,null,"bool"),nK={kernelName:Sc,backendName:"cpu",kernelFunc:tK};function aK(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{depthRadius:s,bias:i,alpha:o,beta:l}=a;xe(r,"LRN");let u=r.shape[3],p=u-1,d=n.data.get(r.dataId).values,c=w.sizeFromShape(r.shape),h=new Float32Array(c);function m(f){let g=f%u,y=f-g+Math.max(0,g-s),b=f-g+Math.min(g+s,p),x=0;for(;y<=b;y++){let v=d[y];x+=v*v}return x}for(let f=0;f<c;f++){let g=m(f),y=d[f]*Math.pow(i+o*g,-l);h[f]=y}return n.makeTensorInfo(r.shape,r.dtype,h)}var rK={kernelName:Nc,backendName:"cpu",kernelFunc:aK};function sK(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,y:s,dy:i}=t,{depthRadius:o,bias:l,alpha:u,beta:p}=a;xe(i,"LRNGrad");let d=w.sizeFromShape(i.shape),c=i.shape[3],h=n.data.get(i.dataId).values,m=n.data.get(r.dataId).values,f=n.data.get(s.dataId).values,g=new Float32Array(d),y=d;for(let b=0;b<y;b++){let x=b%c,v=b-x+Math.max(0,x-o),k=b-x+Math.min(c,x+o+1),T=0;for(let _=v;_<k;_++)T+=Math.pow(m[_],2);T=u*T+l;for(let _=v;_<k;_++){let E=-2*u*p*m[_]*f[b]/T;b===_&&(E+=Math.pow(T,-p)),E*=h[b],g[_]+=E}}return n.makeTensorInfo(i.shape,r.dtype,g)}var iK={kernelName:Bm,backendName:"cpu",kernelFunc:sK};function N_(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{reductionIndices:s,keepDims:i}=a,o=n,l=r.shape,u=l.length,p=w.parseAxisParam(s,l),d=p,c=C.getAxesPermutation(d,u),h=o.data.get(r.dataId).values;if(c!=null){let v=new Array(u);for(let k=0;k<v.length;k++)v[k]=l[c[k]];h=S0(h,l,r.dtype,c,v),d=C.getInnerMostAxes(d.length,u),l=v}xe(r,"max"),C.assertAxesAreInnerMostDims("max",d,u);let[m,f]=C.computeOutAndReduceShapes(l,d),g=w.sizeFromShape(f),y=KT(h,g,m,r.dtype),b=o.write(y,m,r.dtype),x=m;return i&&(x=C.expandShapeToKeepDim(m,p)),{dataId:b,shape:x,dtype:r.dtype}}var oK={kernelName:Ui,backendName:"cpu",kernelFunc:N_};function lK(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t;xe(r,"maxPool");let{filterSize:s,strides:i,pad:o,dimRoundingMode:l}=a,u=1;w.assert(C.eitherStridesOrDilationsAreOne(i,u),()=>`Error in maxPool: Either strides or dilations must be 1. Got strides ${i} and dilations '${u}'`);let p=C.computePool2DInfo(r.shape,s,i,u,o,l),d;if(p.filterWidth===1&&p.filterHeight===1&&w.arraysEqual(p.inShape,p.outShape))d=hr({inputs:{x:r},backend:n});else{let c=n.data.get(r.dataId).values,h=w.computeStrides(r.shape),m=C0(c,r.shape,r.dtype,h,p,"max");d=n.makeTensorInfo(p.outShape,r.dtype,m.values)}return d}var uK={kernelName:Hi,backendName:"cpu",kernelFunc:lK};function pK(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{filterSize:s,strides:i,pad:o,dimRoundingMode:l,dataFormat:u}=a;xe(r,"maxPool3d");let p=C.computePool3DInfo(r.shape,s,i,1,o,l,u),d=n.data.get(r.dataId).values,c=w_(d,r.shape,r.dtype,w.computeStrides(r.shape),p,"max");return n.makeTensorInfo(c.shape,"float32",c.values)}var cK={kernelName:Tc,backendName:"cpu",kernelFunc:pK};function dK(e){let{inputs:t,backend:n,attrs:a}=e,{dy:r,input:s}=t,{filterSize:i,strides:o,pad:l,dimRoundingMode:u}=a;xe([r,s],"maxPool3DGrad");let p=C.computePool3DInfo(s.shape,i,o,1,l,u),d=n.bufferSync(s),c=Z5(d,p),h=p.strideDepth,m=p.strideHeight,f=p.strideWidth,g=p.dilationDepth,y=p.dilationHeight,b=p.dilationWidth,x=p.effectiveFilterDepth,v=p.effectiveFilterHeight,k=p.effectiveFilterWidth,T=x-1-p.padInfo.front,_=k-1-p.padInfo.left,E=v-1-p.padInfo.top,A=Ve(s.shape,"float32"),M=n.bufferSync(r);for(let $=0;$<p.batchSize;++$)for(let S=0;S<p.inChannels;++S)for(let P=0;P<p.inDepth;++P)for(let V=0;V<p.inHeight;++V)for(let j=0;j<p.inWidth;++j){let q=P-T,K=V-E,Z=j-_,ee=0;for(let re=0;re<x;re+=g){let Y=(q+re)/h;if(!(Y<0||Y>=p.outDepth||Math.floor(Y)!==Y))for(let ie=0;ie<v;ie+=y){let ae=(K+ie)/m;if(!(ae<0||ae>=p.outHeight||Math.floor(ae)!==ae))for(let le=0;le<k;le+=b){let ue=(Z+le)/f;if(ue<0||ue>=p.outWidth||Math.floor(ue)!==ue)continue;let ke=x*v*k-1-c.get($,Y,ae,ue,S),ye=re*v*k+ie*k+le,Ie=ke===ye?1:0;Ie!==0&&(ee+=M.get($,Y,ae,ue,S)*Ie)}}}A.set(ee,$,P,V,j,S)}return n.makeTensorInfo(A.shape,A.dtype,A.values)}var hK={kernelName:Vm,backendName:"cpu",kernelFunc:dK};function mK(e){let{inputs:t,backend:n,attrs:a}=e,{dy:r,input:s,output:i}=t,o=s;xe([s,i],"maxPoolGrad");let{filterSize:l,strides:u,pad:p,dimRoundingMode:d}=a,c=C.computePool2DInfo(o.shape,l,u,1,p,d),h=n.data.get(o.dataId).values,m=Ve(c.outShape,o.dtype,v_(h,o.shape,o.dtype,c).values),f=c.strideHeight,g=c.strideWidth,y=c.dilationHeight,b=c.dilationWidth,x=c.effectiveFilterHeight,v=c.effectiveFilterWidth,k=v-1-c.padInfo.left,T=x-1-c.padInfo.top,_=Ve(o.shape,"float32"),E=n.data.get(r.dataId).values,A=Ve(r.shape,"float32",E);for(let M=0;M<c.batchSize;++M)for(let $=0;$<c.inChannels;++$)for(let S=0;S<c.inHeight;++S)for(let P=0;P<c.inWidth;++P){let V=S-T,j=P-k,q=0;for(let K=0;K<x;K+=y){let Z=(V+K)/f;if(!(Z<0||Z>=c.outHeight||Math.floor(Z)!==Z))for(let ee=0;ee<v;ee+=b){let re=(j+ee)/g;if(re<0||re>=c.outWidth||Math.floor(re)!==re)continue;let Y=x*v-1-m.get(M,Z,re,$),ie=K*v+ee,ae=Y===ie?1:0;ae!==0&&(q+=A.get(M,Z,re,$)*ae)}}_.set(q,M,S,P,$)}return n.makeTensorInfo(_.shape,_.dtype,_.values)}var fK={kernelName:Wm,backendName:"cpu",kernelFunc:mK};function gK(e,t,n,a,r){let s=w.computeStrides(t),i=C0(e,t,n,s,r,"max"),o=v_(e,t,n,r,!0,a);return[i.values,o.values]}var yK={kernelName:Um,backendName:"cpu",kernelFunc:({inputs:e,attrs:t,backend:n})=>{let{x:a}=e,{filterSize:r,strides:s,pad:i,includeBatchInIndex:o}=t,l=n;xe(a,"MaxPoolWithArgmax");let u=l.data.get(a.dataId).values,p=C.computePool2DInfo(a.shape,r,s,[1,1],i),[d,c]=gK(u,a.shape,a.dtype,o,p),h=l.write(d,p.outShape,a.dtype),m=l.write(c,p.outShape,a.dtype);return[{dataId:h,shape:p.outShape,dtype:a.dtype},{dataId:m,shape:p.outShape,dtype:"int32"}]}};function bK(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{axis:s,keepDims:i}=a,o=w.parseAxisParam(s,r.shape),l=C.computeOutAndReduceShapes(r.shape,o)[1],u=w.sizeFromShape(l),p=[],d=n.makeTensorInfo([],"float32",new Float32Array([u]));p.push(d);let c=fs({inputs:{x:r},backend:n,attrs:{dtype:"float32"}});p.push(c);let h=E0({inputs:{a:c,b:d},backend:n});p.push(h);let m=ud({inputs:{x:h},backend:n,attrs:{axis:s,keepDims:i}});return p.forEach(f=>n.disposeIntermediateTensorInfo(f)),m}var xK={kernelName:ji,backendName:"cpu",kernelFunc:bK};function vK(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{axis:s,keepDims:i}=a;xe(r,"min");let o=w.parseAxisParam(s,r.shape),l=o,u=C.getAxesPermutation(l,r.shape.length),p=r;u!=null&&(p=Vn({inputs:{x:r},backend:n,attrs:{perm:u}}),l=C.getInnerMostAxes(l.length,r.shape.length)),C.assertAxesAreInnerMostDims("min",l,p.shape.length);let[d,c]=C.computeOutAndReduceShapes(p.shape,l),h=w.sizeFromShape(c),m=w.makeZerosTypedArray(w.sizeFromShape(d),p.dtype),f=n.data.get(p.dataId).values;for(let y=0;y<m.length;++y){let b=y*h,x=f[b];for(let v=0;v<h;++v){let k=f[b+v];(Number.isNaN(k)||k<x)&&(x=k)}m[y]=x}u!=null&&n.disposeIntermediateTensorInfo(p);let g=n.makeTensorInfo(d,p.dtype,m);if(i){let y=C.expandShapeToKeepDim(d,o),b=xt({inputs:{x:g},backend:n,attrs:{shape:y}});return n.disposeIntermediateTensorInfo(g),b}return g}var wK={kernelName:qi,backendName:"cpu",kernelFunc:vK};function kK(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{paddings:s,mode:i}=a;xe(r,"mirrorPad");let o=s.map((b,x)=>b[0]+r.shape[x]+b[1]),l=s.map(b=>b[0]),u=s.map((b,x)=>b[0]+r.shape[x]),p=i==="reflect"?0:1,d=n.data.get(r.dataId).values,c=r.shape.length,h=w.computeStrides(r.shape),m=w.sizeFromShape(o),f=o.length,g=w.computeStrides(o),y=w.getTypedArrayFromDType(r.dtype,m);for(let b=0;b<m;b++){let x=w.indexToLoc(b,f,g);for(let k=0;k<f;k++)x[k]<l[k]?x[k]=l[k]*2-x[k]-p:x[k]>=u[k]&&(x[k]=(u[k]-1)*2-x[k]+p);x=x.map((k,T)=>k-l[T]);let v=w.locToIndex(x,c,h);y[b]=d[v]}return{dataId:n.write(y,o,r.dtype),shape:o,dtype:r.dtype}}var IK={kernelName:Xi,backendName:"cpu",kernelFunc:kK},SK=Ut((e,t)=>{let n=e%t;return e<0&&t<0||e>=0&&t>=0?n:(n+t)%t}),NK=sn(mu,SK),TK={kernelName:mu,backendName:"cpu",kernelFunc:NK},_K=xi(bI());function T_(e){let{inputs:t,backend:n,attrs:a}=e,{logits:r}=t,{dim:s}=a,i=r.shape.length,o=s;if(o===-1&&(o=i-1),o!==i-1)throw Error(`Softmax along a non-last dimension is not yet supported. Logits was rank ${i} and dim was ${o}`);let l=w.parseAxisParam([o],r.shape),u=N_({inputs:{x:r},backend:n,attrs:{reductionIndices:l,keepDims:!1}}),p=C.expandShapeToKeepDim(u.shape,l),d=xt({inputs:{x:u},backend:n,attrs:{shape:p}}),c=_0({inputs:{a:r,b:d},backend:n}),h=OT({inputs:{x:c},backend:n}),m=ud({inputs:{x:h},backend:n,attrs:{axis:l,keepDims:!1}}),f=xt({inputs:{x:m},backend:n,attrs:{shape:p}}),g=E0({inputs:{a:h,b:f},backend:n});return n.disposeIntermediateTensorInfo(u),n.disposeIntermediateTensorInfo(d),n.disposeIntermediateTensorInfo(c),n.disposeIntermediateTensorInfo(h),n.disposeIntermediateTensorInfo(m),n.disposeIntermediateTensorInfo(f),g}var CK={kernelName:ho,backendName:"cpu",kernelFunc:T_};function EK(e){let{inputs:t,backend:n,attrs:a}=e,{logits:r}=t,{numSamples:s,seed:i,normalized:o}=a;xe(r,"multinomial");let l=o?r:T_({inputs:{logits:r},backend:n,attrs:{dim:-1}}),u=l.shape[0],p=l.shape[1],d=n.data.get(l.dataId).values,c=[u,s],h=w.makeZerosTypedArray(w.sizeFromShape(c),"int32");for(let m=0;m<u;++m){let f=m*p,g=new Float32Array(p-1);g[0]=d[f];for(let x=1;x<g.length;++x)g[x]=g[x-1]+d[f+x];let y=_K.alea(i.toString()),b=m*s;for(let x=0;x<s;++x){let v=y();h[b+x]=g.length;for(let k=0;k<g.length;k++)if(v<g[k]){h[b+x]=k;break}}}return o||n.disposeIntermediateTensorInfo(l),n.makeTensorInfo(c,"int32",h)}var FK={kernelName:Gm,backendName:"cpu",kernelFunc:EK},AK=gr.nonMaxSuppressionV3Impl;function $K(e){let{inputs:t,backend:n,attrs:a}=e,{boxes:r,scores:s}=t,{maxOutputSize:i,iouThreshold:o,scoreThreshold:l}=a;xe(r,"NonMaxSuppression");let u=n.data.get(r.dataId).values,p=n.data.get(s.dataId).values,{selectedIndices:d}=AK(u,p,i,o,l);return n.makeTensorInfo([d.length],"int32",new Int32Array(d))}var DK={kernelName:yu,backendName:"cpu",kernelFunc:$K},RK=gr.nonMaxSuppressionV4Impl;function MK(e){let{inputs:t,backend:n,attrs:a}=e,{boxes:r,scores:s}=t,{maxOutputSize:i,iouThreshold:o,scoreThreshold:l,padToMaxOutputSize:u}=a;xe(r,"NonMaxSuppressionPadded");let p=n.data.get(r.dataId).values,d=n.data.get(s.dataId).values,{selectedIndices:c,validOutputs:h}=RK(p,d,i,o,l,u);return[n.makeTensorInfo([c.length],"int32",new Int32Array(c)),n.makeTensorInfo([],"int32",new Int32Array([h]))]}var PK={kernelName:bu,backendName:"cpu",kernelFunc:MK},OK=gr.nonMaxSuppressionV5Impl;function LK(e){let{inputs:t,backend:n,attrs:a}=e,{boxes:r,scores:s}=t,{maxOutputSize:i,iouThreshold:o,scoreThreshold:l,softNmsSigma:u}=a;xe(r,"NonMaxSuppressionWithScore");let p=n.data.get(r.dataId).values,d=n.data.get(s.dataId).values,c=i,h=o,m=l,f=u,{selectedIndices:g,selectedScores:y}=OK(p,d,c,h,m,f);return[n.makeTensorInfo([g.length],"int32",new Int32Array(g)),n.makeTensorInfo([y.length],"float32",new Float32Array(y))]}var zK={kernelName:xu,backendName:"cpu",kernelFunc:LK};function BK(e){let{inputs:t,backend:n,attrs:a}=e,{indices:r}=t,{depth:s,onValue:i,offValue:o}=a;xe(r,"oneHot");let l=w.sizeFromShape(r.shape),u=new Float32Array(l*s);u.fill(o);let p=n.data.get(r.dataId).values;for(let d=0;d<l;++d)p[d]>=0&&p[d]<s&&(u[d*s+p[d]]=i);return n.makeTensorInfo([...r.shape,s],"int32",u)}var WK={kernelName:Qi,backendName:"cpu",kernelFunc:BK};function fm(e){let{inputs:t,backend:n}=e,{x:a}=t;if(a.dtype==="string")throw new Error("zerosLike is not supported for string tensors");if(a.dtype==="complex64"){let r=mi({inputs:{input:a},backend:n}),s=fm({inputs:{x:r},backend:n}),i=Rl({inputs:{input:a},backend:n}),o=fm({inputs:{x:i},backend:n}),l=Yn({inputs:{real:s,imag:o},backend:n});return n.disposeIntermediateTensorInfo(r),n.disposeIntermediateTensorInfo(s),n.disposeIntermediateTensorInfo(i),n.disposeIntermediateTensorInfo(o),l}else return F0({backend:n,attrs:{shape:a.shape,value:0,dtype:a.dtype}})}var VK={kernelName:Lu,backendName:"cpu",kernelFunc:fm};function __(e){let{inputs:t,backend:n}=e,{x:a}=t;if(a.dtype==="string")throw new Error("onesLike is not supported for string tensors");if(a.dtype==="complex64"){let r=mi({inputs:{input:a},backend:n}),s=__({inputs:{x:r},backend:n}),i=Rl({inputs:{input:a},backend:n}),o=fm({inputs:{x:i},backend:n}),l=Yn({inputs:{real:s,imag:o},backend:n});return n.disposeIntermediateTensorInfo(r),n.disposeIntermediateTensorInfo(s),n.disposeIntermediateTensorInfo(i),n.disposeIntermediateTensorInfo(o),l}else return F0({backend:n,attrs:{shape:a.shape,value:1,dtype:a.dtype}})}var UK={kernelName:vu,backendName:"cpu",kernelFunc:__};function C_(e){let{inputs:t,backend:n,attrs:a}=e,{axis:r}=a;if(t.length===1)return mm({inputs:{input:t[0]},backend:n,attrs:{dim:r}});let s=t[0].shape,i=t[0].dtype;t.forEach(p=>{w.assertShapesMatch(s,p.shape,"All tensors passed to stack must have matching shapes"),w.assert(i===p.dtype,()=>"All tensors passed to stack must have matching dtypes")});let o=[],l=t.map(p=>{let d=mm({inputs:{input:p},backend:n,attrs:{dim:r}});return o.push(d),d}),u=Ml({inputs:l,backend:n,attrs:{axis:r}});return o.forEach(p=>n.disposeIntermediateTensorInfo(p)),u}var GK={kernelName:wu,backendName:"cpu",kernelFunc:C_};function HK(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{paddings:s,constantValue:i}=a;xe(r,"pad");let o=s.map((y,b)=>y[0]+r.shape[b]+y[1]),l=s.map(y=>y[0]),u=n.data.get(r.dataId).values,p=w.sizeFromShape(r.shape),d=r.shape.length,c=w.computeStrides(r.shape),h=w.sizeFromShape(o),m=o.length,f=w.computeStrides(o),g=w.getTypedArrayFromDType(r.dtype,h);i!==0&&g.fill(i);for(let y=0;y<p;y++){let b=w.indexToLoc(y,d,c).map((v,k)=>v+l[k]),x=w.locToIndex(b,m,f);g[x]=u[y]}return{dataId:n.write(g,o,r.dtype),shape:o,dtype:r.dtype}}var E_={kernelName:Ji,backendName:"cpu",kernelFunc:HK},jK=Ut((e,t)=>Math.pow(e,t)),qK=sn(Zi,jK),KK={kernelName:Zi,backendName:"cpu",kernelFunc:qK};function XK(e){let{backend:t,attrs:n}=e,{start:a,stop:r,dtype:s,step:i}=n,o=N0(a,r,i,s);return t.makeTensorInfo([o.length],s,o)}var YK={kernelName:_c,backendName:"cpu",kernelFunc:XK},QK=ot(ku,e=>1/e),JK={kernelName:ku,backendName:"cpu",kernelFunc:QK};function ZK(e){let{inputs:t,backend:n,attrs:a}=e,{images:r}=t,{alignCorners:s,halfPixelCenters:i,size:o}=a;xe(r,"resizeBilinear");let l=w.computeStrides(r.shape),[u,p]=o,[d,c,h,m]=r.shape,f=n.data.get(r.dataId).values,g=new Float32Array(w.sizeFromShape([d,u,p,m])),y=[s&&u>1?c-1:c,s&&p>1?h-1:h],b=[s&&u>1?u-1:u,s&&p>1?p-1:p],x=0,v=y[0]/b[0],k=y[1]/b[1];for(let T=0;T<d;T++)for(let _=0;_<u;_++){let E;i?E=v*(_+.5)-.5:E=v*_;let A=Math.max(0,Math.floor(E)),M=E-A,$=Math.min(c-1,Math.ceil(E)),S=T*l[0]+A*l[1],P=T*l[0]+$*l[1];for(let V=0;V<p;V++){let j;i?j=k*(V+.5)-.5:j=k*V;let q=Math.max(0,Math.floor(j)),K=j-q,Z=Math.min(h-1,Math.ceil(j)),ee=S+q*l[2],re=P+q*l[2],Y=S+Z*l[2],ie=P+Z*l[2];for(let ae=0;ae<m;ae++){let le=f[ee+ae],ue=f[re+ae],ke=f[Y+ae],ye=f[ie+ae],Ie=le+(ke-le)*K,Ee=ue+(ye-ue)*K,$e=Ie+(Ee-Ie)*M;g[x++]=$e}}}return n.makeTensorInfo([d,u,p,m],"float32",g)}var eX={kernelName:ao,backendName:"cpu",kernelFunc:ZK};function tX(e){let{inputs:t,backend:n,attrs:a}=e,{images:r,dy:s}=t,{alignCorners:i}=a;xe([s,r],"resizeBilinearGrad");let o=w.computeStrides(r.shape),[l,u,p,d]=r.shape,[,c,h]=s.shape,m=new Float32Array(l*u*p*d),f=[i&&c>1?u-1:u,i&&h>1?p-1:p],g=[i&&c>1?c-1:c,i&&h>1?h-1:h],y=f[0]/g[0],b=f[1]/g[1],x=n.data.get(s.dataId).values,v=0;for(let k=0;k<l;k++){let T=k*o[0];for(let _=0;_<c;_++){let E=_*y,A=Math.floor(E),M=Math.min(Math.ceil(E),u-1),$=T+A*o[1],S=T+M*o[1],P=E-A,V=1-P;for(let j=0;j<h;j++){let q=j*b,K=Math.floor(q),Z=Math.min(Math.ceil(q),p-1),ee=q-K,re=1-ee,Y=$+K*o[2],ie=$+Z*o[2],ae=S+K*o[2],le=S+Z*o[2],ue=V*re,ke=V*ee,ye=P*re,Ie=P*ee;for(let Ee=0;Ee<d;Ee++){let $e=x[v++];m[Y+Ee]+=$e*ue,m[ie+Ee]+=$e*ke,m[ae+Ee]+=$e*ye,m[le+Ee]+=$e*Ie}}}}return n.makeTensorInfo([l,p,u,d],"float32",m)}var nX={kernelName:qm,backendName:"cpu",kernelFunc:tX};function aX(e){let{inputs:t,backend:n,attrs:a}=e,{images:r}=t,{alignCorners:s,halfPixelCenters:i,size:o}=a;xe(r,"resizeNearestNeighbor");let l=w.computeStrides(r.shape),[u,p]=o,[d,c,h,m]=r.shape,f=n.data.get(r.dataId).values,g=new Float32Array(d*u*p*m),y=[s&&u>1?c-1:c,s&&p>1?h-1:h],b=[s&&u>1?u-1:u,s&&p>1?p-1:p],x=y[0]/b[0],v=y[1]/b[1],k=0;for(let T=0;T<d;T++){let _=T*l[0];for(let E=0;E<u;E++){let A=i?x*(E+.5):x*E,M=Math.min(c-1,s?Math.round(A):Math.floor(A));i&&(M=Math.max(0,M));let $=_+M*l[1];for(let S=0;S<p;S++){let P=i?v*(S+.5):v*S,V=Math.min(h-1,s?Math.round(P):Math.floor(P));i&&(V=Math.max(0,V));let j=$+V*l[2];for(let q=0;q<m;q++){let K=f[j+q];g[k++]=K}}}}return n.makeTensorInfo([d,u,p,m],r.dtype,g)}var rX={kernelName:Cc,backendName:"cpu",kernelFunc:aX};function sX(e){let{inputs:t,backend:n,attrs:a}=e,{images:r,dy:s}=t,{alignCorners:i}=a;xe([s,r],"resizeNearestNeighborGrad");let o=w.computeStrides(r.shape),l=w.computeStrides(s.shape),[u,p,d,c]=r.shape,[,h,m]=s.shape,f=new Float32Array(u*p*d*c),g=n.data.get(s.dataId).values,y=[i&&h>1?p-1:p,i&&m>1?d-1:d],b=[i&&h>1?h-1:h,i&&m>1?m-1:m],x=y[0]/b[0],v=y[1]/b[1],k=1/x,T=1/v,_=Math.ceil(k)*2+2,E=Math.ceil(T)*2+2;for(let A=0;A<u;A++){let M=A*o[0];for(let $=0;$<p;$++){let S=M+$*o[1],P=Math.floor($*k),V=Math.floor(P-_/2);for(let j=0;j<d;j++){let q=S+j*o[2],K=Math.floor(j*T),Z=Math.floor(K-E/2);for(let ee=0;ee<c;ee++){let re=0;for(let Y=0;Y<_;Y++){let ie=Y+V;if(ie<0||ie>=h)continue;let ae=M+ie*l[1],le=ie*x,ue=Math.min(p-1,i?Math.round(le):Math.floor(le));if($===ue)for(let ke=0;ke<E;ke++){let ye=ke+Z;if(ye<0||ye>=m)continue;let Ie=ae+ye*l[2],Ee=ye*v,$e=Math.min(d-1,i?Math.round(Ee):Math.floor(Ee));j===$e&&(re+=g[Ie+ee])}}f[q+ee]=re}}}}return n.makeTensorInfo(r.shape,r.dtype,f)}var iX={kernelName:jm,backendName:"cpu",kernelFunc:sX};function oX(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{dims:s}=a;xe(r,"reverse");let i=r.shape.length,o=w.parseAxisParam(s,r.shape);if(i===0)return hr({inputs:{x:r},backend:n});let l=new qt(r.shape,r.dtype),u=n.bufferSync(r);for(let p=0;p<l.size;p++){let d=l.indexToLoc(p),c=d.slice();o.forEach(h=>c[h]=r.shape[h]-1-c[h]),l.set(u.get(...c),...d)}return n.makeTensorInfo(l.shape,l.dtype,l.values)}var lX={kernelName:so,backendName:"cpu",kernelFunc:oX},uX={kernelName:zu,backendName:"cpu",kernelFunc:({inputs:e,attrs:t,backend:n})=>{let{image:a}=e,{radians:r,fillValue:s,center:i}=t,o=n,l=w.getTypedArrayFromDType(a.dtype,w.sizeFromShape(a.shape)),[u,p,d,c]=a.shape,[h,m]=C.getImageCenter(i,p,d),f=255,g=Math.sin(r),y=Math.cos(r),b=o.data.get(a.dataId).values;for(let x=0;x<u;x++){let v=x*d*p*c;for(let k=0;k<p;k++){let T=k*(d*c);for(let _=0;_<d;_++){let E=_*c;for(let A=0;A<c;A++){let M=[u,k,_,A],$=M[2],S=M[1],P=($-h)*y-(S-m)*g,V=($-h)*g+(S-m)*y;P=Math.round(P+h),V=Math.round(V+m);let j=s;if(typeof s!="number"&&(A===3?j=f:j=s[A]),P>=0&&P<d&&V>=0&&V<p){let K=V*(d*c),Z=P*c,ee=v+K+Z+A;j=b[ee]}let q=v+T+E+A;l[q]=j}}}}return{dataId:o.write(l,a.shape,a.dtype),shape:a.shape,dtype:a.dtype}}},pX=ot(io,e=>{let t=Math.floor(e);return e-t<.5?Math.floor(e):e-t>.5?Math.ceil(e):t%2===0?t:t+1}),cX={kernelName:io,backendName:"cpu",kernelFunc:pX};function dX(e){let{inputs:t,backend:n,attrs:a}=e,{indices:r,updates:s}=t,{shape:i}=a,{sliceRank:o,numUpdates:l,sliceSize:u,strides:p,outputSize:d}=C.calculateShapes(s,r,i),c=!0,h=n.bufferSync(r),m=n.bufferSync(s),f=fl(h,m,i,d,u,l,o,p,0,c);return n.makeTensorInfo(i,f.dtype,f.values)}var hX={kernelName:Su,backendName:"cpu",kernelFunc:dX};function mX(e,t){let n=0,a=e.length,r=0;for(;n<a;)r=Math.floor((n+a)/2),e[r]<t?n=r+1:a=r;return a}function fX(e,t){let n=0,a=e.length,r=0;for(;n<a;)r=Math.floor((n+a)/2),e[r]<=t?n=r+1:a=r;return a}function gX(e,t,n,a,r,s){let i=w.getArrayFromDType("int32",n*r);for(let o=0;o<n;++o){let l=e.slice(o*a,(o+1)*a),u=o*r;for(let p=0;p<r;++p)i[u+p]=s==="left"?mX(l,t[p+u]):fX(l,t[p+u])}return i}function yX(e){let{inputs:t,backend:n,attrs:a}=e,{sortedSequence:r,values:s}=t,{side:i}=a,o=n.data.get(r.dataId).values,l=n.data.get(s.dataId).values,u=gX(o,l,r.shape[0],r.shape[1],s.shape[1],i);return n.makeTensorInfo(s.shape,"int32",u)}var bX={kernelName:Km,backendName:"cpu",kernelFunc:yX};function xX(e){let{inputs:t,backend:n}=e,{condition:a,t:r,e:s}=t;xe([a,r,s],"select");let i=a.shape.length,o=n.data.get(a.dataId).values,l=n.data.get(r.dataId).values,u=n.data.get(s.dataId).values,p=fa(r.dtype,s.dtype),d=w.makeZerosTypedArray(w.sizeFromShape(r.shape),p),c=0,h=i===0||i>1||r.shape.length===1?1:w.sizeFromShape(r.shape.slice(1));for(let m=0;m<o.length;m++)for(let f=0;f<h;f++)o[m]===1?d[c++]=l[m]:d[c++]=u[m];return n.makeTensorInfo(r.shape,p,d)}var vX={kernelName:Nu,backendName:"cpu",kernelFunc:xX},wX=C.SELU_SCALEALPHA,kX=C.SELU_SCALE,IX=ot(Tu,e=>e>=0?kX*e:wX*(Math.exp(e)-1)),SX={kernelName:Tu,backendName:"cpu",kernelFunc:IX},NX=ot(Eu,e=>e<0?-1:e>0?1:0),TX={kernelName:Eu,backendName:"cpu",kernelFunc:NX},_X=ot(lo,e=>Math.sin(e)),CX={kernelName:lo,backendName:"cpu",kernelFunc:_X},EX=ot(Cu,e=>Math.sinh(e)),FX={kernelName:Cu,backendName:"cpu",kernelFunc:EX},AX=11920928955078125e-23,Gk=Math.log(AX)+2,$X=ot(Fu,e=>{let t=e>-Gk,n=e<Gk,a=Math.exp(e),r;return n?r=a:t?r=e:r=Math.log(1+a),r}),DX={kernelName:Fu,backendName:"cpu",kernelFunc:$X};function RX(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{blockShape:s,paddings:i}=a;xe([r],"spaceToBatchND");let o=w.sizeFromShape(s),l=[[0,0]];l.push(...i);for(let g=1+s.length;g<r.shape.length;++g)l.push([0,0]);let u=E_.kernelFunc({inputs:{x:r},backend:n,attrs:{paddings:l,constantValue:0}}),p=C.getReshaped(u.shape,s,o,!1),d=C.getPermuted(p.length,s.length,!1),c=C.getReshapedPermuted(u.shape,s,o,!1),h=xt({inputs:{x:u},backend:n,attrs:{shape:p}}),m=Vn({inputs:{x:h},backend:n,attrs:{perm:d}}),f=xt({inputs:{x:m},backend:n,attrs:{shape:c}});return n.disposeIntermediateTensorInfo(u),n.disposeIntermediateTensorInfo(h),n.disposeIntermediateTensorInfo(m),f}var MX={kernelName:Au,backendName:"cpu",kernelFunc:RX};function PX(e){let{inputs:t,backend:n}=e,{indices:a,values:r,denseShape:s,defaultValue:i}=t;if(s.shape.length!==1)throw new Error(`Dense shape must be a vector, saw:
${s.shape}`);if(a.shape.length!==2)throw new Error(`Indices must be a matrix, saw:
${a.shape}`);if(r.shape.length!==1)throw new Error(`Values must be a vector, saw:
${r.shape}`);if(i.shape.length!==0)throw new Error(`Default value must be a scalar, saw:
${i.shape}`);let o=n.data.get(a.dataId).values,l=n.data.get(r.dataId).values,u=n.data.get(s.dataId).values,p=n.data.get(i.dataId).values[0],[d,c,h,m,f]=n_(o,a.shape,a.dtype,l,r.dtype,u,p);return[n.makeTensorInfo(c,a.dtype,d),n.makeTensorInfo([c[0]],r.dtype,h),n.makeTensorInfo([m.length],"bool",new Uint8Array(m.map(g=>Number(g)))),n.makeTensorInfo([f.length],a.dtype,new Int32Array(f))]}var OX={kernelName:Ec,backendName:"cpu",kernelFunc:PX};function LX(e){let{inputs:t,backend:n}=e,{inputIndices:a,inputShape:r,newShape:s}=t;if(a.shape.length!==2)throw new Error(`Input indices should be a matrix but received shape
${a.shape}`);if(r.shape.length!==1)throw new Error(`Input shape should be a vector but received shape
${r.shape}`);if(s.shape.length!==1)throw new Error(`Target shape should be a vector but received shape ${s.shape}`);let i=Array.from(n.data.get(r.dataId).values),o=n.data.get(a.dataId).values,l=Array.from(n.data.get(s.dataId).values),[u,p,d]=a_(o,a.shape,a.dtype,i,l);return[n.makeTensorInfo(p,a.dtype,u),n.makeTensorInfo([d.length],s.dtype,new Int32Array(d))]}var zX={kernelName:Du,backendName:"cpu",kernelFunc:LX};function BX(e){let{inputs:t,backend:n}=e,{data:a,indices:r,segmentIds:s}=t;if(a.shape.length<1)throw new Error("Data should be at least 1 dimensional but received scalar");if(r.shape.length!==1)throw new Error(`Indices should be a vector but received shape
${r.shape}`);if(s.shape.length!==1)throw new Error(`Segment ids should be a vector but received shape
${s.shape}`);if(r.shape[0]!==s.shape[0])throw new Error("segmentIds and indices should have same size.");let i=n.data.get(a.dataId).values,o=n.data.get(r.dataId).values,l=n.data.get(s.dataId).values,[u,p]=T0(i,a.shape,a.dtype,o,l,!0);return n.makeTensorInfo(p,a.dtype,u)}var WX={kernelName:Fc,backendName:"cpu",kernelFunc:BX};function VX(e){let{inputs:t,backend:n}=e,{data:a,indices:r,segmentIds:s}=t;if(a.shape.length<1)throw new Error("Data should be at least 1 dimensional but received scalar");if(r.shape.length!==1)throw new Error(`Indices should be a vector but received shape
${r.shape}`);if(s.shape.length!==1)throw new Error(`Segment ids should be a vector but received shape
${s.shape}`);if(r.shape[0]!==s.shape[0])throw new Error("segmentIds and indices should have same size.");let i=n.data.get(a.dataId).values,o=n.data.get(r.dataId).values,l=n.data.get(s.dataId).values,[u,p]=T0(i,a.shape,a.dtype,o,l);return n.makeTensorInfo(p,a.dtype,u)}var UX={kernelName:Ac,backendName:"cpu",kernelFunc:VX};function GX(e){let{inputs:t,backend:n,attrs:a}=e,{sparseIndices:r,sparseValues:s,defaultValue:i}=t,{outputShape:o}=a,{sliceRank:l,numUpdates:u,sliceSize:p,strides:d,outputSize:c}=C.calculateShapes(s,r,o),h=!1,m=n.bufferSync(r),f;switch(s.dtype){case"bool":{let g=n.bufferSync(s),y=Boolean(n.data.get(i.dataId).values[0]);f=fl(m,g,o,c,p,u,l,d,y,h);break}case"float32":{let g=n.bufferSync(s),y=n.data.get(i.dataId).values[0];f=fl(m,g,o,c,p,u,l,d,y,h);break}case"int32":{let g=n.bufferSync(s),y=n.data.get(i.dataId).values[0];f=fl(m,g,o,c,p,u,l,d,y,h);break}case"string":{let g=n.bufferSync(s),y=w.decodeString(n.data.get(i.dataId).values[0]);f=fl(m,g,o,c,p,u,l,d,y,h);break}default:throw new Error(`Unsupported type ${s.dtype}`)}return n.makeTensorInfo(o,f.dtype,f.values)}var HX={kernelName:Xm,backendName:"cpu",kernelFunc:GX};function jX(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{numOrSizeSplits:s,axis:i}=a,o=w.parseAxisParam(i,r.shape)[0],l=C.prepareSplitSize(r,s,o),u=new Array(r.shape.length).fill(0),p=r.shape.slice();return l.map(d=>{let c=[...p];c[o]=d;let h=fi({inputs:{x:r},backend:n,attrs:{begin:u,size:c}});return u[o]+=d,h})}var qX={kernelName:$u,backendName:"cpu",kernelFunc:jX},KX={kernelName:$c,backendName:"cpu",kernelFunc:({inputs:e,backend:t})=>{let{x:n}=e,a=t;xe(n,"square");let r=a.data.get(n.dataId).values,s=new Float32Array(r.length);for(let i=0;i<r.length;++i){let o=r[i];s[i]=o*o}return{dataId:a.write(s,n.shape,n.dtype),shape:n.shape,dtype:n.dtype}}},XX=ot(ws,(e,t)=>{let n=t;return isNaN(e)?NaN:e>0?1:n.alpha}),YX={kernelName:ws,backendName:"cpu",kernelFunc:XX};function QX(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{begin:s,end:i,strides:o,beginMask:l,endMask:u,ellipsisMask:p,newAxisMask:d,shrinkAxisMask:c}=a;xe(r,"stridedSlice");let{finalShapeSparse:h,finalShape:m,isIdentity:f,sliceDim0:g,isSimpleSlice:y,begin:b,end:x,strides:v}=Kt.sliceInfo(r.shape,s,i,o,l,u,p,d,c),k;if(f)k=xt({inputs:{x:r},backend:n,attrs:{shape:m}});else if(g||y){w.assert(r.shape.length>=1,()=>`Input must have rank at least 1, got: ${r.shape.length}`);let T=Kt.computeOutShape(b,x,v),_=fi({inputs:{x:r},backend:n,attrs:{begin:b,size:T}});k=xt({inputs:{x:_},backend:n,attrs:{shape:m}}),n.disposeIntermediateTensorInfo(_)}else{let T=n.bufferSync(r),_=s_(h,T,v,b);k=n.makeTensorInfo(m,_.dtype,_.values)}return k}var JX={kernelName:Ru,backendName:"cpu",kernelFunc:QX};function ZX(e){let{inputs:t,backend:n,attrs:a}=e,{separator:r,nGramWidths:s,leftPad:i,rightPad:o,padWidth:l,preserveShortSequences:u}=a,{data:p,dataSplits:d}=t,c=n.data.get(p.dataId).values,h=n.data.get(d.dataId).values,[m,f]=i_(c,h,r,s,i,o,l,u);return[n.makeTensorInfo([m.length],"string",m),n.makeTensorInfo(d.shape,"int32",f)]}var e7={kernelName:Ym,backendName:"cpu",kernelFunc:ZX};function t7(e){let{inputs:t,backend:n,attrs:a}=e,{skipEmpty:r}=a,{input:s,delimiter:i}=t;if(s.dtype!=="string")throw new Error("Input must be of datatype string");if(s.shape.length!==1)throw new Error(`Input must be a vector, got shape: ${s.shape}`);if(i.shape.length!==0)throw new Error(`Delimiter must be a scalar, got shape: ${i.shape}`);let o=n.data.get(s.dataId).values,l=n.data.get(i.dataId).values[0],[u,p,d]=o_(o,l,r),c=p.length;return[n.makeTensorInfo([c,2],"int32",u),n.makeTensorInfo([c],"string",p),n.makeTensorInfo([2],"int32",new Int32Array(d))]}var n7={kernelName:Qm,backendName:"cpu",kernelFunc:t7};function a7(e){let{inputs:t,backend:n,attrs:a}=e,{numBuckets:r}=a,{input:s}=t;if(s.dtype!=="string")throw new Error("Input must be of datatype string");if(r<=0)throw new Error("Number of buckets must be at least 1");let i=n.data.get(s.dataId).values,o=l_(i,r);return n.makeTensorInfo(s.shape,"int32",o)}var r7={kernelName:Jm,backendName:"cpu",kernelFunc:a7},s7=ot(go,e=>Math.tan(e)),i7={kernelName:go,backendName:"cpu",kernelFunc:s7},o7=ot(yo,e=>Math.tanh(e)),l7={kernelName:yo,backendName:"cpu",kernelFunc:o7};function u7(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{reps:s}=a;xe(r,"tile");let i=p_(n.bufferSync(r),s);return n.makeTensorInfo(i.shape,i.dtype,i.values)}var p7={kernelName:vs,backendName:"cpu",kernelFunc:u7};function c7(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{k:s,sorted:i}=a;xe(r,"topk");let o=n.data.get(r.dataId).values,[l,u]=d_(o,r.shape,r.dtype,s,i);return[n.makeTensorInfo(l.shape,l.dtype,l.values),n.makeTensorInfo(u.shape,u.dtype,u.values)]}var d7={kernelName:Mu,backendName:"cpu",kernelFunc:c7};function h7(e){let{inputs:t,attrs:n,backend:a}=e,{image:r,transforms:s}=t,{interpolation:i,fillMode:o,fillValue:l,outputShape:u}=n,[p,d,c,h]=r.shape,[m,f]=u!=null?u:[d,c],g=[p,m,f,h],y=w.computeStrides(r.shape),b=y[0],x=y[1],v=y[2],k=w.getTypedArrayFromDType(r.dtype,w.sizeFromShape(g));k.fill(l);let T=a.data.get(r.dataId).values,_=a.data.get(s.dataId).values;for(let E=0;E<p;++E){let A=s.shape[0]===1?_:_.subarray(E*8,E*8+8);for(let M=0;M<m;++M)for(let $=0;$<f;++$)for(let S=0;S<h;++S){let P,V=A[6]*$+A[7]*M+1;if(V===0)continue;let j=(A[0]*$+A[1]*M+A[2])/V,q=(A[3]*$+A[4]*M+A[5])/V,K=Hk(j,c,o),Z=Hk(q,d,o);switch(i){case"nearest":P=x7(T,d,c,b,x,v,E,Z,K,S,l);break;case"bilinear":P=v7(T,d,c,b,x,v,E,Z,K,S,l);break;default:throw new Error(`Error in Transform: Expect 'nearest' or 'bilinear', but got ${i}`)}let ee=E*b+M*x+$*v+S;k[ee]=P}return a.makeTensorInfo(g,r.dtype,k)}return{dataId:a.write(k,g,r.dtype),shape:r.shape,dtype:r.dtype}}var m7={kernelName:Pu,backendName:"cpu",kernelFunc:h7};function Hk(e,t,n){switch(n){case"reflect":return f7(e,t);case"wrap":return g7(e,t);case"nearest":return b7(e,t);case"constant":default:return y7(e,t)}}function f7(e,t){let n=e;if(n<0)if(t<=1)n=0;else{let a=2*t;n<a&&(n=a*Math.trunc(-n/a)+n),n=n<-t?n+a:-n-1}else if(n>t-1)if(t<=1)n=0;else{let a=2*t;n-=a*Math.trunc(n/a),n>=t&&(n=a-n-1)}return w.clamp(0,n,t-1)}function g7(e,t){let n=e;if(n<0)if(t<=1)n=0;else{let a=t-1;n+=t*(Math.trunc(-n/a)+1)}else if(n>t-1)if(t<=1)n=0;else{let a=t-1;n-=t*Math.trunc(n/a)}return w.clamp(0,n,t-1)}function y7(e,t){return e}function b7(e,t){return w.clamp(0,e,t-1)}function Hp(e,t,n,a,r,s,i,o,l,u,p){let d=i*a+o*r+l*s+u;return 0<=o&&o<t&&0<=l&&l<n?e[d]:p}function x7(e,t,n,a,r,s,i,o,l,u,p){let d=Math.round(o),c=Math.round(l);return Hp(e,t,n,a,r,s,i,d,c,u,p)}function v7(e,t,n,a,r,s,i,o,l,u,p){let d=Math.floor(o),c=Math.floor(l),h=d+1,m=c+1,f=(m-l)*Hp(e,t,n,a,r,s,i,d,c,u,p)+(l-c)*Hp(e,t,n,a,r,s,i,d,m,u,p),g=(m-l)*Hp(e,t,n,a,r,s,i,h,c,u,p)+(l-c)*Hp(e,t,n,a,r,s,i,h,m,u,p);return(h-o)*f+(o-d)*g}function w7(e){let{inputs:t,attrs:n,backend:a}=e,{axis:r}=n,{x:s}=t;xe(s,"unique");let i=a.data.get(s.dataId).values,{outputValues:o,outputShape:l,indices:u}=h_(i,r,s.shape,s.dtype);return[a.makeTensorInfo(l,s.dtype,o),a.makeTensorInfo([u.length],"int32",u)]}var k7={kernelName:Zm,backendName:"cpu",kernelFunc:w7};function I7(e){let{inputs:t,backend:n,attrs:a}=e,{value:r}=t,{axis:s}=a;s<0&&(s+=r.shape.length);let i=r.shape.length,o=r.shape[s],l=new Array(i-1),u=0;for(let h=0;h<i;h++)h!==s&&(l[u++]=r.shape[h]);let p=new Array(i).fill(0),d=r.shape.slice();d[s]=1;let c=new Array(o);for(let h=0;h<c.length;h++){p[s]=h;let m=fi({inputs:{x:r},backend:n,attrs:{begin:p,size:d}});c[h]=xt({inputs:{x:m},backend:n,attrs:{shape:l}}),n.disposeIntermediateTensorInfo(m)}return c}var S7={kernelName:Ou,backendName:"cpu",kernelFunc:I7};function N7(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,segmentIds:s}=t,{numSegments:i}=a;xe(r,"unsortedSegmentSum");let o=r.shape.length,l=s.shape.length,u=[],p=[],d=o-l,c=s;for(let m=0;m<d;++m){let f=mm({inputs:{input:c},backend:n,attrs:{dim:m+1}});c=f,p.push(f)}for(let m=0;m<i;++m){let f=w.createScalarValue(m,"int32"),g=n.makeTensorInfo([],"int32",f),y=MT({inputs:{a:g,b:c},backend:n}),b=fs({inputs:{x:y},backend:n,attrs:{dtype:"float32"}}),x=ag({inputs:{a:b,b:r},backend:n}),v=ud({inputs:{x},backend:n,attrs:{axis:0,keepDims:!1}});u.push(v),p.push(g),p.push(y),p.push(b),p.push(x),p.push(v)}let h=C_({inputs:u,backend:n,attrs:{axis:0}});return p.forEach(m=>n.disposeIntermediateTensorInfo(m)),h}var T7={kernelName:Dc,backendName:"cpu",kernelFunc:N7},_7=[_5,vj,E5,A5,Tj,D5,M5,O5,z5,W5,U5,H5,q5,Y5,J5,tq,aq,sq,oq,N5,uq,cq,hq,fq,Sj,Cj,yq,wj,xq,wq,kq,Sq,Tq,Cq,Fq,$q,Rq,Pq,Lq,Bq,Vq,Gq,jq,qq,Xq,Qq,Zq,e8,t8,n8,s8,b5,o8,Ej,f8,Fj,g8,$j,k8,I8,N8,Rj,C8,F8,$8,R8,P8,Pj,Lj,kj,L8,vq,B8,V8,G8,x5,Bj,Vj,j8,Gj,K8,Q8,Z8,nK,rK,iK,oK,jj,uK,cK,hK,fK,yK,xK,wK,Kj,IK,TK,FK,Yj,Jj,DK,PK,zK,e5,WK,UK,GK,E_,KK,w5,a5,YK,Ij,fx,JK,k5,I5,S5,eX,nX,rX,iX,lX,uX,cX,s5,hX,bX,vX,SX,o5,TX,CX,FX,l5,CK,DX,MX,OX,zX,WX,UX,HX,qX,c5,KX,h5,YX,JX,e7,n7,r7,y5,a8,i7,l7,p7,d7,m7,t5,k7,S7,T7,VK];for(let e of _7)Rc(e);var F_={};Me(F_,{assertNotComplex:()=>Zu,bindCanvasToFramebuffer:()=>L7,bindColorTextureToFramebuffer:()=>Mh,bindTextureToProgramUniformSampler:()=>j_,bindTextureUnit:()=>U_,bindVertexBufferToProgramAttribute:()=>yx,callAndCheck:()=>ge,canBeRepresented:()=>$_,createFragmentShader:()=>M_,createFramebuffer:()=>V_,createProgram:()=>P_,createStaticIndexBuffer:()=>z_,createStaticVertexBuffer:()=>L_,createTexture:()=>B_,createVertexShader:()=>R_,getBatchDim:()=>gi,getExtensionOrThrow:()=>jp,getFramebufferErrorMessage:()=>q_,getMaxTexturesInShader:()=>Q_,getNumChannels:()=>P7,getProgramUniformLocation:()=>H_,getProgramUniformLocationOrThrow:()=>G_,getRowsCols:()=>yi,getShapeAs3D:()=>Ph,getTextureShapeFromLogicalShape:()=>X_,getWebGLDisjointQueryTimerVersion:()=>J_,getWebGLErrorMessage:()=>D_,getWebGLMaxTextureSize:()=>Y_,hasExtension:()=>ha,isCapableOfRenderingToFloatTexture:()=>Z_,isDownloadFloatTextureEnabled:()=>eC,isReshapeFree:()=>dc,isWebGLFenceEnabled:()=>tC,isWebGLVersionEnabled:()=>xx,linkProgram:()=>O_,logShaderSourceAndInfoLog:()=>$0,resetMaxTextureSize:()=>z7,resetMaxTexturesInShader:()=>B7,unbindColorTextureFromFramebuffer:()=>bx,unbindTextureUnit:()=>O7,validateFramebuffer:()=>qp,validateProgram:()=>Rh,validateTextureSize:()=>W_});var Xs={},Cb={alpha:!1,antialias:!1,premultipliedAlpha:!1,preserveDrawingBuffer:!1,depth:!1,stencil:!1,failIfMajorPerformanceCaveat:!0};function A_(e,t){Xs[e]=t}function Ya(e,t){if(!(e in Xs)||t!=null){let a=E7(e,t);if(a!==null)Xs[e]=a;else return console.log("Could not get context for WebGL version",e),null}let n=Xs[e];return n==null||n.isContextLost()?(delete Xs[e],Ya(e)):(n.disable(n.DEPTH_TEST),n.disable(n.STENCIL_TEST),n.disable(n.BLEND),n.disable(n.DITHER),n.disable(n.POLYGON_OFFSET_FILL),n.disable(n.SAMPLE_COVERAGE),n.enable(n.SCISSOR_TEST),n.enable(n.CULL_FACE),n.cullFace(n.BACK),Xs[e])}function C7(e){if(typeof OffscreenCanvas!="undefined"&&e===2)return new OffscreenCanvas(300,150);if(typeof document!="undefined")return document.createElement("canvas");throw new Error("Cannot create a canvas in this context")}function E7(e,t){if(e!==1&&e!==2)throw new Error("Cannot get WebGL rendering context, WebGL is disabled.");let n=t==null?C7(e):t;return n.addEventListener("webglcontextlost",a=>{a.preventDefault(),delete Xs[e]},!1),e===1?n.getContext("webgl",Cb)||n.getContext("experimental-webgl",Cb):n.getContext("webgl2",Cb)}var cc;(function(e){e[e.DENSE=0]="DENSE",e[e.SHARED_BATCH=1]="SHARED_BATCH"})(cc||(cc={}));var da;(function(e){e[e.RENDER=0]="RENDER",e[e.UPLOAD=1]="UPLOAD",e[e.PIXELS=2]="PIXELS",e[e.DOWNLOAD=3]="DOWNLOAD"})(da||(da={}));var un;(function(e){e[e.UNPACKED_FLOAT16=0]="UNPACKED_FLOAT16",e[e.UNPACKED_FLOAT32=1]="UNPACKED_FLOAT32",e[e.PACKED_4X1_UNSIGNED_BYTE=2]="PACKED_4X1_UNSIGNED_BYTE",e[e.PACKED_2X2_FLOAT32=3]="PACKED_2X2_FLOAT32",e[e.PACKED_2X2_FLOAT16=4]="PACKED_2X2_FLOAT16"})(un||(un={}));function pd(e,t){return[t,e]}function F7(e,t){return e*t}function _h(e){let t=w.sizeFromShape(e),n=Math.ceil(t/4);return w.sizeToSquarishShape(n)}function Ju(e,t){return[Math.max(1,Math.ceil(t/2)),Math.max(1,Math.ceil(e/2))]}function A7(e,t){let[n,a]=Ju(e,t);return n*a*4}function A0(e,t){let n=e,a,r,s,i,o,l,u,p,d,c;return X().getNumber("WEBGL_VERSION")===2?(a=n.R32F,r=n.R16F,s=n.RGBA16F,i=n.RGBA32F,o=n.RED,u=4,p=1,d=n.HALF_FLOAT,c=n.FLOAT,l=n.RGBA8):(a=e.RGBA,r=e.RGBA,s=e.RGBA,i=n.RGBA,o=e.RGBA,u=4,p=4,d=t!=null?t.HALF_FLOAT_OES:null,c=e.FLOAT,l=e.RGBA),{internalFormatFloat:a,internalFormatHalfFloat:r,internalFormatPackedHalfFloat:s,internalFormatPackedFloat:i,textureFormatFloat:o,downloadTextureFormat:l,downloadUnpackNumChannels:u,defaultNumChannels:p,textureTypeHalfFloat:d,textureTypeFloat:c}}function ge(e,t){let n=t();return X().getBool("DEBUG")&&$7(e),n}function $7(e){let t=e.getError();if(t!==e.NO_ERROR)throw new Error("WebGL Error: "+D_(e,t))}var D7=596e-10,R7=65504;function $_(e){return!!(X().getBool("WEBGL_RENDER_FLOAT32_ENABLED")||e===0||D7<Math.abs(e)&&Math.abs(e)<R7)}function D_(e,t){switch(t){case e.NO_ERROR:return"NO_ERROR";case e.INVALID_ENUM:return"INVALID_ENUM";case e.INVALID_VALUE:return"INVALID_VALUE";case e.INVALID_OPERATION:return"INVALID_OPERATION";case e.INVALID_FRAMEBUFFER_OPERATION:return"INVALID_FRAMEBUFFER_OPERATION";case e.OUT_OF_MEMORY:return"OUT_OF_MEMORY";case e.CONTEXT_LOST_WEBGL:return"CONTEXT_LOST_WEBGL";default:return`Unknown error code ${t}`}}function jp(e,t){return Pr(e,()=>e.getExtension(t),'Extension "'+t+'" not supported on this browser.')}function R_(e,t){let n=Pr(e,()=>e.createShader(e.VERTEX_SHADER),"Unable to create vertex WebGLShader.");if(ge(e,()=>e.shaderSource(n,t)),ge(e,()=>e.compileShader(n)),e.getShaderParameter(n,e.COMPILE_STATUS)===!1)throw console.log(e.getShaderInfoLog(n)),new Error("Failed to compile vertex shader.");return n}function M_(e,t){let n=Pr(e,()=>e.createShader(e.FRAGMENT_SHADER),"Unable to create fragment WebGLShader.");if(ge(e,()=>e.shaderSource(n,t)),ge(e,()=>e.compileShader(n)),X().get("ENGINE_COMPILE_ONLY"))return n;if(e.getShaderParameter(n,e.COMPILE_STATUS)===!1)throw $0(t,e.getShaderInfoLog(n)),new Error("Failed to compile fragment shader.");return n}var M7=/ERROR: [0-9]+:([0-9]+):/g;function $0(e,t){let n=M7.exec(t);if(n==null){console.log(`Couldn't parse line number in error: ${t}`),console.log(e);return}let a=+n[1],r=e.split(`
`),s=r.length.toString().length+2,i=r.map((d,c)=>w.rightPad((c+1).toString(),s)+d),o=0;for(let d=0;d<i.length;d++)o=Math.max(i[d].length,o);let l=i.slice(0,a-1),u=i.slice(a-1,a),p=i.slice(a);console.log(l.join(`
`)),console.log(t.split(`
`)[0]),console.log(`%c ${w.rightPad(u[0],o)}`,"border:1px solid red; background-color:#e3d2d2; color:#a61717"),console.log(p.join(`
`))}function P_(e){return Pr(e,()=>e.createProgram(),"Unable to create WebGLProgram.")}function O_(e,t){if(ge(e,()=>e.linkProgram(t)),!X().get("ENGINE_COMPILE_ONLY")&&e.getProgramParameter(t,e.LINK_STATUS)===!1)throw console.log(e.getProgramInfoLog(t)),new Error("Failed to link vertex and fragment shaders.")}function Rh(e,t){if(ge(e,()=>e.validateProgram(t)),e.getProgramParameter(t,e.VALIDATE_STATUS)===!1)throw console.log(e.getProgramInfoLog(t)),new Error("Shader program validation failed.")}function L_(e,t){let n=Pr(e,()=>e.createBuffer(),"Unable to create WebGLBuffer");return ge(e,()=>e.bindBuffer(e.ARRAY_BUFFER,n)),ge(e,()=>e.bufferData(e.ARRAY_BUFFER,t,e.STATIC_DRAW)),n}function z_(e,t){let n=Pr(e,()=>e.createBuffer(),"Unable to create WebGLBuffer");return ge(e,()=>e.bindBuffer(e.ELEMENT_ARRAY_BUFFER,n)),ge(e,()=>e.bufferData(e.ELEMENT_ARRAY_BUFFER,t,e.STATIC_DRAW)),n}function P7(){return X().getNumber("WEBGL_VERSION")===2?1:4}function B_(e){return Pr(e,()=>e.createTexture(),"Unable to create WebGLTexture.")}function W_(e,t){let n=X().getNumber("WEBGL_MAX_TEXTURE_SIZE");if(e<=0||t<=0){let a=`[${e}x${t}]`;throw new Error("Requested texture size "+a+" is invalid.")}if(e>n||t>n){let a=`[${e}x${t}]`,r=`[${n}x${n}]`;throw new Error("Requested texture size "+a+" greater than WebGL maximum on this browser / GPU "+r+".")}}function V_(e){return Pr(e,()=>e.createFramebuffer(),"Unable to create WebGLFramebuffer.")}function yx(e,t,n,a,r,s,i){let o=e.getAttribLocation(t,n);return o===-1?!1:(ge(e,()=>e.bindBuffer(e.ARRAY_BUFFER,a)),ge(e,()=>e.vertexAttribPointer(o,r,e.FLOAT,!1,s,i)),ge(e,()=>e.enableVertexAttribArray(o)),!0)}function U_(e,t,n){K_(e,n),ge(e,()=>e.activeTexture(e.TEXTURE0+n)),ge(e,()=>e.bindTexture(e.TEXTURE_2D,t))}function O7(e,t){K_(e,t),ge(e,()=>e.activeTexture(e.TEXTURE0+t)),ge(e,()=>e.bindTexture(e.TEXTURE_2D,null))}function G_(e,t,n){return Pr(e,()=>e.getUniformLocation(t,n),'uniform "'+n+'" not present in program.')}function H_(e,t,n){return e.getUniformLocation(t,n)}function j_(e,t,n,a){ge(e,()=>U_(e,t,a)),ge(e,()=>e.uniform1i(n,a))}function L7(e){ge(e,()=>e.bindFramebuffer(e.FRAMEBUFFER,null)),ge(e,()=>e.viewport(0,0,e.canvas.width,e.canvas.height)),ge(e,()=>e.scissor(0,0,e.canvas.width,e.canvas.height))}function Mh(e,t,n){ge(e,()=>e.bindFramebuffer(e.FRAMEBUFFER,n)),ge(e,()=>e.framebufferTexture2D(e.FRAMEBUFFER,e.COLOR_ATTACHMENT0,e.TEXTURE_2D,t,0))}function bx(e,t){ge(e,()=>e.bindFramebuffer(e.FRAMEBUFFER,t)),ge(e,()=>e.framebufferTexture2D(e.FRAMEBUFFER,e.COLOR_ATTACHMENT0,e.TEXTURE_2D,null,0))}function qp(e){let t=e.checkFramebufferStatus(e.FRAMEBUFFER);if(t!==e.FRAMEBUFFER_COMPLETE)throw new Error("Error binding framebuffer: "+q_(e,t))}function q_(e,t){switch(t){case e.FRAMEBUFFER_INCOMPLETE_ATTACHMENT:return"FRAMEBUFFER_INCOMPLETE_ATTACHMENT";case e.FRAMEBUFFER_INCOMPLETE_MISSING_ATTACHMENT:return"FRAMEBUFFER_INCOMPLETE_MISSING_ATTACHMENT";case e.FRAMEBUFFER_INCOMPLETE_DIMENSIONS:return"FRAMEBUFFER_INCOMPLETE_DIMENSIONS";case e.FRAMEBUFFER_UNSUPPORTED:return"FRAMEBUFFER_UNSUPPORTED";default:return`unknown error ${t}`}}function Pr(e,t,n){let a=ge(e,()=>t());if(a==null)throw new Error(n);return a}function K_(e,t){let n=e.MAX_COMBINED_TEXTURE_IMAGE_UNITS-1,a=t+e.TEXTURE0;if(a<e.TEXTURE0||a>n){let r=`[gl.TEXTURE0, gl.TEXTURE${n}]`;throw new Error(`textureUnit must be in ${r}.`)}}function gi(e,t=2){return w.sizeFromShape(e.slice(0,e.length-t))}function yi(e){if(e.length===0)throw Error("Cannot get rows and columns of an empty shape array.");return[e.length>1?e[e.length-2]:1,e[e.length-1]]}function Ph(e){let t=[1,1,1];return e.length===0||e.length===1&&e[0]===1||(t=[gi(e),...yi(e)]),t}function X_(e,t=!1){let n=X().getNumber("WEBGL_MAX_TEXTURE_SIZE");t&&(n=n*2,e=e.map((r,s)=>s>=e.length-2?w.nearestLargerEven(e[s]):e[s]),e.length===1&&(e=[2,e[0]])),e.length!==2&&(e=w.squeezeShape(e).newShape);let a=w.sizeFromShape(e);if(e.length<=1&&a<=n)return[1,a];if(e.length===2&&e[0]<=n&&e[1]<=n)return e;if(e.length===3&&e[0]*e[1]<=n&&e[2]<=n)return[e[0]*e[1],e[2]];if(e.length===3&&e[0]<=n&&e[1]*e[2]<=n)return[e[0],e[1]*e[2]];if(e.length===4&&e[0]*e[1]*e[2]<=n&&e[3]<=n)return[e[0]*e[1]*e[2],e[3]];if(e.length===4&&e[0]<=n&&e[1]*e[2]*e[3]<=n)return[e[0],e[1]*e[2]*e[3]];if(t){let r=gi(e),s=2,i=2;return e.length&&([s,i]=yi(e)),a=r*(s/2)*(i/2),w.sizeToSquarishShape(a).map(o=>o*2)}return w.sizeToSquarishShape(a)}function Ch(e){return e%2===0}function dc(e,t){if(e=e.slice(-2),t=t.slice(-2),w.arraysEqual(e,t)||!e.length||!t.length||e[0]===0||e[1]===0||t[0]===0||t[1]===0)return!0;if(e.length!==t.length){let n=e.slice(-1)[0],a=t.slice(-1)[0];if(n===a||Ch(n)&&Ch(a)&&(e[0]===1||t[0]===1))return!0}return e[1]===t[1]&&Ch(e[0])&&Ch(t[0])}var Oh,Lh;function Y_(e){if(Oh==null){let t=Ya(e);Oh=t.getParameter(t.MAX_TEXTURE_SIZE)}return Oh}function z7(){Oh=null}function B7(){Lh=null}function Q_(e){if(Lh==null){let t=Ya(e);Lh=t.getParameter(t.MAX_TEXTURE_IMAGE_UNITS)}return Math.min(16,Lh)}function J_(e){if(e===0)return 0;let t,n=Ya(e);return ha(n,"EXT_disjoint_timer_query_webgl2")&&e===2?t=2:ha(n,"EXT_disjoint_timer_query")?t=1:t=0,t}function ha(e,t){return e.getExtension(t)!=null}function xx(e){try{if(Ya(e)!=null)return!0}catch(t){return console.log("Error when getting WebGL context: ",t),!1}return!1}function Z_(e){if(e===0)return!1;let t=Ya(e);if(e===1){if(!ha(t,"OES_texture_float"))return!1}else if(!ha(t,"EXT_color_buffer_float"))return!1;return vx(t)}function eC(e){if(e===0)return!1;let t=Ya(e);if(e===1){if(!ha(t,"OES_texture_float")||!ha(t,"WEBGL_color_buffer_float"))return!1}else{if(ha(t,"EXT_color_buffer_float"))return vx(t);let n="EXT_color_buffer_half_float";if(ha(t,n)){let a=t.getExtension(n);return W7(t,a)}return!1}return vx(t)}function vx(e){let t=A0(e),n=e.createTexture();e.bindTexture(e.TEXTURE_2D,n);let a=1,r=1;e.texImage2D(e.TEXTURE_2D,0,t.internalFormatFloat,a,r,0,t.textureFormatFloat,t.textureTypeFloat,null);let s=e.createFramebuffer();e.bindFramebuffer(e.FRAMEBUFFER,s),e.framebufferTexture2D(e.FRAMEBUFFER,e.COLOR_ATTACHMENT0,e.TEXTURE_2D,n,0);let i=e.checkFramebufferStatus(e.FRAMEBUFFER)===e.FRAMEBUFFER_COMPLETE;return e.bindTexture(e.TEXTURE_2D,null),e.bindFramebuffer(e.FRAMEBUFFER,null),e.deleteTexture(n),e.deleteFramebuffer(s),i}function W7(e,t){let n=A0(e,t),a=e.createTexture();e.bindTexture(e.TEXTURE_2D,a);let r=1,s=1;e.texImage2D(e.TEXTURE_2D,0,n.internalFormatHalfFloat,r,s,0,n.textureFormatFloat,n.textureTypeHalfFloat,null);let i=e.createFramebuffer();e.bindFramebuffer(e.FRAMEBUFFER,i),e.framebufferTexture2D(e.FRAMEBUFFER,e.COLOR_ATTACHMENT0,e.TEXTURE_2D,a,0);let o=e.checkFramebufferStatus(e.FRAMEBUFFER)===e.FRAMEBUFFER_COMPLETE;return e.bindTexture(e.TEXTURE_2D,null),e.bindFramebuffer(e.FRAMEBUFFER,null),e.deleteTexture(a),e.deleteFramebuffer(i),o}function tC(e){return e!==2?!1:Ya(e).fenceSync!=null}function Zu(e,t){Array.isArray(e)||(e=[e]),e.forEach(n=>{n!=null&&w.assert(n.dtype!=="complex64",()=>`${t} does not support complex64 tensors in the WebGL backend.`)})}var Ne=X();Ne.registerFlag("HAS_WEBGL",()=>Ne.getNumber("WEBGL_VERSION")>0);Ne.registerFlag("WEBGL_VERSION",()=>xx(2)?2:xx(1)?1:0);Ne.registerFlag("WEBGL_CHECK_NUMERICAL_PROBLEMS",()=>!1);Ne.registerFlag("WEBGL_BUFFER_SUPPORTED",()=>Ne.get("WEBGL_VERSION")===2);Ne.registerFlag("WEBGL_CPU_FORWARD",()=>!0);Ne.registerFlag("WEBGL_FORCE_F16_TEXTURES",()=>!1);Ne.registerFlag("WEBGL_PACK",()=>Ne.getBool("HAS_WEBGL"));Ne.registerFlag("WEBGL_PACK_NORMALIZATION",()=>Ne.getBool("WEBGL_PACK"));Ne.registerFlag("WEBGL_PACK_CLIP",()=>Ne.getBool("WEBGL_PACK"));Ne.registerFlag("WEBGL_PACK_DEPTHWISECONV",()=>Ne.getBool("WEBGL_PACK"));Ne.registerFlag("WEBGL_PACK_BINARY_OPERATIONS",()=>Ne.getBool("WEBGL_PACK"));Ne.registerFlag("WEBGL_PACK_UNARY_OPERATIONS",()=>Ne.getBool("WEBGL_PACK"));Ne.registerFlag("WEBGL_PACK_ARRAY_OPERATIONS",()=>Ne.getBool("WEBGL_PACK"));Ne.registerFlag("WEBGL_PACK_IMAGE_OPERATIONS",()=>Ne.getBool("WEBGL_PACK"));Ne.registerFlag("WEBGL_PACK_REDUCE",()=>Ne.getBool("WEBGL_PACK"));Ne.registerFlag("WEBGL_LAZILY_UNPACK",()=>Ne.getBool("WEBGL_PACK"));Ne.registerFlag("WEBGL_CONV_IM2COL",()=>Ne.getBool("WEBGL_PACK"));Ne.registerFlag("WEBGL_MAX_TEXTURE_SIZE",()=>Y_(Ne.getNumber("WEBGL_VERSION")));Ne.registerFlag("WEBGL_MAX_TEXTURES_IN_SHADER",()=>Q_(Ne.getNumber("WEBGL_VERSION")));Ne.registerFlag("WEBGL_DISJOINT_QUERY_TIMER_EXTENSION_VERSION",()=>{let e=Ne.getNumber("WEBGL_VERSION");return e===0?0:J_(e)});Ne.registerFlag("WEBGL_DISJOINT_QUERY_TIMER_EXTENSION_RELIABLE",()=>Ne.getNumber("WEBGL_DISJOINT_QUERY_TIMER_EXTENSION_VERSION")>0&&!Oc.isMobile());Ne.registerFlag("WEBGL_RENDER_FLOAT32_CAPABLE",()=>Z_(Ne.getNumber("WEBGL_VERSION")));Ne.registerFlag("WEBGL_RENDER_FLOAT32_ENABLED",()=>Ne.getBool("WEBGL_FORCE_F16_TEXTURES")?!1:Ne.getBool("WEBGL_RENDER_FLOAT32_CAPABLE"));Ne.registerFlag("WEBGL_DOWNLOAD_FLOAT_ENABLED",()=>eC(Ne.getNumber("WEBGL_VERSION")));Ne.registerFlag("WEBGL_FENCE_API_ENABLED",()=>tC(Ne.getNumber("WEBGL_VERSION")));Ne.registerFlag("WEBGL_SIZE_UPLOAD_UNIFORM",()=>Ne.getBool("WEBGL_RENDER_FLOAT32_ENABLED")?4:0);Ne.registerFlag("WEBGL_DELETE_TEXTURE_THRESHOLD",()=>-1,e=>{if(e<0&&e!==-1)throw new Error(`WEBGL_DELETE_TEXTURE_THRESHOLD must be -1 (indicating never delete) or at least 0, but got ${e}.`)});Ne.registerFlag("WEBGL_FLUSH_THRESHOLD",()=>Oc.isMobile()?1:-1,e=>{if(e<0&&e!==-1)throw new Error(`WEBGL_FLUSH_THRESHOLD must be -1 (indicating never manual flush) or at least 0, but got ${e}.`)});Ne.registerFlag("CPU_HANDOFF_SIZE_THRESHOLD",()=>128);Ne.registerFlag("WEBGL_USE_SHAPES_UNIFORMS",()=>!1);Ne.registerFlag("TOPK_LAST_DIM_CPU_HANDOFF_SIZE_THRESHOLD",()=>1e5);Ne.registerFlag("TOPK_K_CPU_HANDOFF_THRESHOLD",()=>128);function En(){let e,t,n,a,r,s,i,o,l,u;return X().getNumber("WEBGL_VERSION")===2?(e="#version 300 es",t="in",n="out",a="in",r="texture",s="outputColor",i="out vec4 outputColor;",o=`
bool isnan_custom(float val) {
uint floatToUint = floatBitsToUint(val);
return (floatToUint & 0x7fffffffu) > 0x7f800000u;
}
bvec4 isnan_custom(vec4 val) {
return bvec4(isnan_custom(val.x),
isnan_custom(val.y), isnan_custom(val.z), isnan_custom(val.w));
}
#define isnan(value) isnan_custom(value)
`,l="",u=`
#define round(value) newRound(value)
int newRound(float value) {
return int(floor(value + 0.5));
}
ivec4 newRound(vec4 value) {
return ivec4(floor(value + vec4(0.5)));
}
`):(e="",t="attribute",n="varying",a="varying",r="texture2D",s="gl_FragColor",i="",o=`
#define isnan(value) isnan_custom(value)
bool isnan_custom(float val) {
return (val > 0. || val < 1. || val == 0.) ? false : true;
}
bvec4 isnan_custom(vec4 val) {
return bvec4(isnan(val.x), isnan(val.y), isnan(val.z), isnan(val.w));
}
`,l=`
uniform float INFINITY;
bool isinf(float val) {
return abs(val) == INFINITY;
}
bvec4 isinf(vec4 val) {
return equal(abs(val), vec4(INFINITY));
}
`,u=`
int round(float value) {
return int(floor(value + 0.5));
}
ivec4 round(vec4 value) {
return ivec4(floor(value + vec4(0.5)));
}
`),{version:e,attribute:t,varyingVs:n,varyingFs:a,texture2D:r,output:s,defineOutput:i,defineSpecialNaN:o,defineSpecialInf:l,defineRound:u}}function So(e,t,n="index"){let a=w.computeStrides(t);return a.map((r,s)=>{let i=`int ${e[s]} = ${n} / ${r}`,o=s===a.length-1?`int ${e[s+1]} = ${n} - ${e[s]} * ${r}`:`index -= ${e[s]} * ${r}`;return`${i}; ${o};`}).join("")}function rg(e,t,n="index"){let a=w.computeStrides(t);return a.map((r,s)=>{let i=`int ${e[s]} = ${n} / outShapeStrides[${s}]`,o=s===a.length-1?`int ${e[s+1]} = ${n} - ${e[s]} * outShapeStrides[${s}]`:`index -= ${e[s]} * outShapeStrides[${s}]`;return`${i}; ${o};`}).join("")}function V7(e,t){let n=e.length,a=e.map(s=>`${t}[${s}]`),r=new Array(n-1);r[n-2]=a[n-1];for(let s=n-3;s>=0;--s)r[s]=`(${r[s+1]} * ${a[s+1]})`;return r}function U7(e,t,n="index"){let a=e.map((s,i)=>i),r=V7(a,t);return r.map((s,i)=>{let o=`int ${e[i]} = ${n} / ${r[i]}`,l=i===r.length-1?`int ${e[i+1]} = ${n} - ${e[i]} * ${r[i]}`:`index -= ${e[i]} * ${r[i]}`;return`${o}; ${l};`}).join("")}function D0(e){let t=w.computeStrides(e).map(n=>n.toString());return`
int getFlatIndex(ivec3 coords) {
return coords.x * ${t[0]} + coords.y * ${t[1]} + coords.z;
}
`}function R0(){return`
int getFlatIndex(ivec3 coords) {
return coords.x * outShapeStrides[0] + coords.y * outShapeStrides[1] + coords.z;
}
`}var nC=`
const float FLOAT_MAX = 1.70141184e38;
const float FLOAT_MIN = 1.17549435e-38;
lowp vec4 encode_float(highp float v) {
if (isnan(v)) {
return vec4(255, 255, 255, 255);
}
highp float av = abs(v);
if(av < FLOAT_MIN) {
return vec4(0.0, 0.0, 0.0, 0.0);
} else if(v > FLOAT_MAX) {
return vec4(0.0, 0.0, 128.0, 127.0) / 255.0;
} else if(v < -FLOAT_MAX) {
return vec4(0.0, 0.0, 128.0, 255.0) / 255.0;
}
highp vec4 c = vec4(0,0,0,0);
highp float e = floor(log2(av));
highp float m = exp2(fract(log2(av))) - 1.0;
c[2] = floor(128.0 * m);
m -= c[2] / 128.0;
c[1] = floor(32768.0 * m);
m -= c[1] / 32768.0;
c[0] = floor(8388608.0 * m);
highp float ebias = e + 127.0;
c[3] = floor(ebias / 2.0);
ebias -= c[3] * 2.0;
c[2] += floor(ebias) * 128.0;
c[3] += 128.0 * step(0.0, -v);
return c / 255.0;
}
`,{getBroadcastDims:aC}=C;function G7(e,t,n){let a=[];if(e.forEach(c=>{let h=w.sizeFromShape(c.shapeInfo.logicalShape);if(c.shapeInfo.isUniform?a.push(`uniform float ${c.name}${h>1?`[${h}]`:""};`):(a.push(`uniform sampler2D ${c.name};`),a.push(`uniform int offset${c.name};`)),n.enableShapeUniforms){let{uniformShape:m}=M0(n.packedInputs,c.shapeInfo.logicalShape,c.shapeInfo.texShape);switch(m.length){case 1:a.push(`uniform int ${c.name}Shape;`);break;case 2:a.push(`uniform ivec2 ${c.name}Shape;`);break;case 3:a.push(`uniform ivec3 ${c.name}Shape;`);break;case 4:a.push(`uniform ivec4 ${c.name}Shape;`);break;default:break}a.push(`uniform ivec2 ${c.name}TexShape;`)}}),n.enableShapeUniforms){switch(t.logicalShape.length){case 1:a.push("uniform int outShape;");break;case 2:a.push("uniform ivec2 outShape;"),a.push("uniform int outShapeStrides;");break;case 3:a.push("uniform ivec3 outShape;"),a.push("uniform ivec2 outShapeStrides;");break;case 4:a.push("uniform ivec4 outShape;"),a.push("uniform ivec3 outShapeStrides;");break;default:break}a.push("uniform ivec2 outTexShape;")}n.customUniforms&&n.customUniforms.forEach(c=>{a.push(`uniform ${c.type} ${c.name}${c.arrayIndex?`[${c.arrayIndex}]`:""};`)});let r=a.join(`
`),s=e.map(c=>H7(c,t,n.packedInputs,n.enableShapeUniforms)).join(`
`),i=t.texShape,o=En(),l=K7(o),u,p,d=Q7(o);return t.isPacked?(u=j7(t.logicalShape,i,n.enableShapeUniforms),p=Y7(o)):(u=q7(t.logicalShape,i,n.enableShapeUniforms),p=X7(o)),n.packedInputs&&(d+=tY),[d,l,p,r,u,s,n.userCode].join(`
`)}function ep(e,t=!1){let n=e.shapeInfo.logicalShape;switch(n.length){case 0:return hY(e,t);case 1:return fY(e,t);case 2:return yY(e,t);case 3:return xY(e,t);case 4:return wY(e,t);case 5:return kY(e);case 6:return IY(e);default:throw new Error(`${n.length}-D input sampling is not yet supported`)}}function rC(e,t){switch(e.shapeInfo.logicalShape.length){case 0:return dY(e);case 1:return mY(e,t);case 2:return gY(e,t);case 3:return bY(e,t);default:return vY(e,t)}}function H7(e,t,n=!1,a){let r="";n?r+=rC(e,a):r+=ep(e,a);let s=e.shapeInfo.logicalShape,i=t.logicalShape;return s.length<=i.length&&(n?r+=SY(e,t):r+=NY(e,t)),r}function j7(e,t,n){switch(e.length){case 0:return sC();case 1:return nY(e,t,n);case 2:return pY(e,t,n);case 3:return rY(e,t,n);default:return iY(e,t,n)}}function q7(e,t,n){switch(e.length){case 0:return sC();case 1:return aY(e,t,n);case 2:return cY(e,t,n);case 3:return sY(e,t,n);case 4:return oY(e,t,n);case 5:return lY(e,t);case 6:return uY(e,t);default:throw new Error(`${e.length}-D output sampling is not yet supported`)}}function K7(e){return`
float sampleTexture(sampler2D textureSampler, vec2 uv) {
return ${e.texture2D}(textureSampler, uv).r;
}
`}function X7(e){return`
void setOutput(float val) {
${e.output} = vec4(val, 0, 0, 0);
}
`}function Y7(e){return`
void setOutput(vec4 val) {
${e.output} = val;
}
`}function Q7(e){return`${e.version}
precision highp float;
precision highp int;
precision highp sampler2D;
${e.varyingFs} vec2 resultUV;
${e.defineOutput}
const vec2 halfCR = vec2(0.5, 0.5);
struct ivec5
{
int x;
int y;
int z;
int w;
int u;
};
struct ivec6
{
int x;
int y;
int z;
int w;
int u;
int v;
};
uniform float NAN;
${e.defineSpecialNaN}
${e.defineSpecialInf}
${e.defineRound}
int imod(int x, int y) {
return x - y * (x / y);
}
int idiv(int a, int b, float sign) {
int res = a / b;
int mod = imod(a, b);
if (sign < 0. && mod != 0) {
res -= 1;
}
return res;
}
//Based on the work of Dave Hoskins
//https://www.shadertoy.com/view/4djSRW
#define HASHSCALE1 443.8975
float random(float seed){
vec2 p = resultUV * seed;
vec3 p3 = fract(vec3(p.xyx) * HASHSCALE1);
p3 += dot(p3, p3.yzx + 19.19);
return fract((p3.x + p3.y) * p3.z);
}
${J7}
${Z7}
${eY}
`}var J7=`
vec2 uvFromFlat(int texNumR, int texNumC, int index) {
int texR = index / texNumC;
int texC = index - texR * texNumC;
return (vec2(texC, texR) + halfCR) / vec2(texNumC, texNumR);
}
vec2 packedUVfrom1D(int texNumR, int texNumC, int index) {
int texelIndex = index / 2;
int texR = texelIndex / texNumC;
int texC = texelIndex - texR * texNumC;
return (vec2(texC, texR) + halfCR) / vec2(texNumC, texNumR);
}
`,Z7=`
vec2 packedUVfrom2D(int texelsInLogicalRow, int texNumR,
int texNumC, int row, int col) {
int texelIndex = (row / 2) * texelsInLogicalRow + (col / 2);
int texR = texelIndex / texNumC;
int texC = texelIndex - texR * texNumC;
return (vec2(texC, texR) + halfCR) / vec2(texNumC, texNumR);
}
`,eY=`
vec2 packedUVfrom3D(int texNumR, int texNumC,
int texelsInBatch, int texelsInLogicalRow, int b,
int row, int col) {
int index = b * texelsInBatch + (row / 2) * texelsInLogicalRow + (col / 2);
int texR = index / texNumC;
int texC = index - texR * texNumC;
return (vec2(texC, texR) + halfCR) / vec2(texNumC, texNumR);
}
`,tY=`
float getChannel(vec4 frag, vec2 innerDims) {
vec2 modCoord = mod(innerDims, 2.);
return modCoord.x == 0. ?
(modCoord.y == 0. ? frag.r : frag.g) :
(modCoord.y == 0. ? frag.b : frag.a);
}
float getChannel(vec4 frag, int dim) {
float modCoord = mod(float(dim), 2.);
return modCoord == 0. ? frag.r : frag.g;
}
`;function sC(){return`
int getOutputCoords() {
return 0;
}
`}function nY(e,t,n){let a=[Math.ceil(t[0]/2),Math.ceil(t[1]/2)];return a[0]===1?n?`
int getOutputCoords() {
return 2 * int(resultUV.x * ceil(float(outTexShape[1]) / 2.0));
}
`:`
int getOutputCoords() {
return 2 * int(resultUV.x * ${a[1]}.0);
}
`:a[1]===1?n?`
int getOutputCoords() {
return 2 * int(resultUV.y * ceil(float(outTexShape[0]) / 2.0));
}
`:`
int getOutputCoords() {
return 2 * int(resultUV.y * ${a[0]}.0);
}
`:n?`
int getOutputCoords() {
ivec2 packedTexShape = ivec2(ceil(float(outTexShape[0]) / 2.0), ceil(float(outTexShape[1]) / 2.0));
ivec2 resTexRC = ivec2(resultUV.yx *
vec2(packedTexShape[0], packedTexShape[1]));
return 2 * (resTexRC.x * packedTexShape[1] + resTexRC.y);
}
`:`
int getOutputCoords() {
ivec2 resTexRC = ivec2(resultUV.yx *
vec2(${a[0]}, ${a[1]}));
return 2 * (resTexRC.x * ${a[1]} + resTexRC.y);
}
`}function aY(e,t,n){return t[0]===1?n?`
int getOutputCoords() {
return int(resultUV.x * float(outTexShape[1]));
}
`:`
int getOutputCoords() {
return int(resultUV.x * ${t[1]}.0);
}
`:t[1]===1?n?`
int getOutputCoords() {
return int(resultUV.y * float(outTexShape[0]));
}
`:`
int getOutputCoords() {
return int(resultUV.y * ${t[0]}.0);
}
`:n?`
int getOutputCoords() {
ivec2 resTexRC = ivec2(resultUV.yx *
vec2(outTexShape[0], outTexShape[1]));
return resTexRC.x * outTexShape[1] + resTexRC.y;
}
`:`
int getOutputCoords() {
ivec2 resTexRC = ivec2(resultUV.yx *
vec2(${t[0]}, ${t[1]}));
return resTexRC.x * ${t[1]} + resTexRC.y;
}
`}function rY(e,t,n){if(n)return`
ivec3 getOutputCoords() {
ivec2 packedTexShape = ivec2(ceil(float(outTexShape[0]) / 2.0), ceil(float(outTexShape[1]) / 2.0));
int texelsInLogicalRow = int(ceil(float(outShape[2]) / 2.0));
int texelsInBatch = texelsInLogicalRow * int(ceil(float(outShape[1]) / 2.0));
ivec2 resTexRC = ivec2(resultUV.yx *
vec2(packedTexShape[0], packedTexShape[1]));
int index = resTexRC.x * packedTexShape[1] + resTexRC.y;
int b = index / texelsInBatch;
index -= b * texelsInBatch;
int r = 2 * (index / texelsInLogicalRow);
int c = imod(index, texelsInLogicalRow) * 2;
return ivec3(b, r, c);
}
`;let a=[Math.ceil(t[0]/2),Math.ceil(t[1]/2)],r=Math.ceil(e[2]/2),s=r*Math.ceil(e[1]/2);return`
ivec3 getOutputCoords() {
ivec2 resTexRC = ivec2(resultUV.yx *
vec2(${a[0]}, ${a[1]}));
int index = resTexRC.x * ${a[1]} + resTexRC.y;
int b = index / ${s};
index -= b * ${s};
int r = 2 * (index / ${r});
int c = imod(index, ${r}) * 2;
return ivec3(b, r, c);
}
`}function sY(e,t,n){if(n)return`
ivec3 getOutputCoords() {
ivec2 resTexRC = ivec2(resultUV.yx *
vec2(outTexShape[0], outTexShape[1]));
int index = resTexRC.x * outTexShape[1] + resTexRC.y;
${rg(["r","c","d"],e)}
return ivec3(r, c, d);
}
`;let a=So(["r","c","d"],e);return`
ivec3 getOutputCoords() {
ivec2 resTexRC = ivec2(resultUV.yx *
vec2(${t[0]}, ${t[1]}));
int index = resTexRC.x * ${t[1]} + resTexRC.y;
${a}
return ivec3(r, c, d);
}
`}function iY(e,t,n){if(n)return`
ivec4 getOutputCoords() {
ivec2 packedTexShape = ivec2(ceil(float(outTexShape[0]) / 2.0), ceil(float(outTexShape[1]) / 2.0));
ivec2 resTexRC = ivec2(resultUV.yx *
vec2(packedTexShape[0], packedTexShape[1]));
int index = resTexRC.x * packedTexShape[1] + resTexRC.y;
int texelsInLogicalRow = int(ceil(float(outShape[3]) / 2.0));
int texelsInBatch = texelsInLogicalRow * int(ceil(float(outShape[2]) / 2.0));
int texelsInBatchN = texelsInBatch * outShape[1];
int b2 = index / texelsInBatchN;
index -= b2 * texelsInBatchN;
int b = index / texelsInBatch;
index -= b * texelsInBatch;
int r = 2 * (index / texelsInLogicalRow);
int c = imod(index, texelsInLogicalRow) * 2;
return ivec4(b2, b, r, c);
}
`;let a=[Math.ceil(t[0]/2),Math.ceil(t[1]/2)],r=Math.ceil(e[e.length-1]/2),s=r*Math.ceil(e[e.length-2]/2),i=s,o="",l="b, r, c";for(let u=2;u<e.length-1;u++)i*=e[e.length-u-1],o=`
int b${u} = index / ${i};
index -= b${u} * ${i};
`+o,l=`b${u}, `+l;return`
ivec${e.length} getOutputCoords() {
ivec2 resTexRC = ivec2(resultUV.yx *
vec2(${a[0]}, ${a[1]}));
int index = resTexRC.x * ${a[1]} + resTexRC.y;
${o}
int b = index / ${s};
index -= b * ${s};
int r = 2 * (index / ${r});
int c = imod(index, ${r}) * 2;
return ivec${e.length}(${l});
}
`}function oY(e,t,n){if(n)return`
ivec4 getOutputCoords() {
ivec2 resTexRC = ivec2(resultUV.yx *
vec2(outTexShape[0], outTexShape[1]));
int index = resTexRC.x * outTexShape[1] + resTexRC.y;
${rg(["r","c","d","d2"],e)}
return ivec4(r, c, d, d2);
}
`;let a=So(["r","c","d","d2"],e);return`
ivec4 getOutputCoords() {
ivec2 resTexRC = ivec2(resultUV.yx *
vec2(${t[0]}, ${t[1]}));
int index = resTexRC.x * ${t[1]} + resTexRC.y;
${a}
return ivec4(r, c, d, d2);
}
`}function lY(e,t){let n=So(["r","c","d","d2","d3"],e);return`
ivec5 getOutputCoords() {
ivec2 resTexRC = ivec2(resultUV.yx * vec2(${t[0]},
${t[1]}));
int index = resTexRC.x * ${t[1]} + resTexRC.y;
${n}
ivec5 outShape = ivec5(r, c, d, d2, d3);
return outShape;
}
`}function uY(e,t){let n=So(["r","c","d","d2","d3","d4"],e);return`
ivec6 getOutputCoords() {
ivec2 resTexRC = ivec2(resultUV.yx *
vec2(${t[0]}, ${t[1]}));
int index = resTexRC.x * ${t[1]} + resTexRC.y;
${n}
ivec6 result = ivec6(r, c, d, d2, d3, d4);
return result;
}
`}function pY(e,t,n){let a=[Math.ceil(t[0]/2),Math.ceil(t[1]/2)];if(w.arraysEqual(e,t))return n?`
ivec2 getOutputCoords() {
ivec2 packedTexShape = ivec2(ceil(float(outTexShape[0]) / 2.0), ceil(float(outTexShape[1]) / 2.0));
return 2 * ivec2(resultUV.yx * vec2(packedTexShape[0], packedTexShape[1]));
}
`:`
ivec2 getOutputCoords() {
return 2 * ivec2(resultUV.yx * vec2(${a[0]}, ${a[1]}));
}
`;let r=Math.ceil(e[1]/2);return n?`
ivec2 getOutputCoords() {
ivec2 packedTexShape = ivec2(ceil(float(outTexShape[0]) / 2.0), ceil(float(outTexShape[1]) / 2.0));
int texelsInLogicalRow = int(ceil(float(outShape[1]) / 2.0));
ivec2 resTexRC = ivec2(resultUV.yx *
vec2(packedTexShape[0], packedTexShape[1]));
int index = resTexRC.x * packedTexShape[1] + resTexRC.y;
int r = 2 * (index / texelsInLogicalRow);
int c = imod(index, texelsInLogicalRow) * 2;
return ivec2(r, c);
}
`:`
ivec2 getOutputCoords() {
ivec2 resTexRC = ivec2(resultUV.yx *
vec2(${a[0]}, ${a[1]}));
int index = resTexRC.x * ${a[1]} + resTexRC.y;
int r = 2 * (index / ${r});
int c = imod(index, ${r}) * 2;
return ivec2(r, c);
}
`}function cY(e,t,n){return w.arraysEqual(e,t)?n?`
ivec2 getOutputCoords() {
return ivec2(resultUV.yx * vec2(outTexShape[0], outTexShape[1]));
}
`:`
ivec2 getOutputCoords() {
return ivec2(resultUV.yx * vec2(${t[0]}, ${t[1]}));
}
`:e[1]===1?n?`
ivec2 getOutputCoords() {
ivec2 resTexRC = ivec2(resultUV.yx *
vec2(outTexShape[0], outTexShape[1]));
int index = resTexRC.x * outTexShape[1] + resTexRC.y;
return ivec2(index, 0);
}
`:`
ivec2 getOutputCoords() {
ivec2 resTexRC = ivec2(resultUV.yx *
vec2(${t[0]}, ${t[1]}));
int index = resTexRC.x * ${t[1]} + resTexRC.y;
return ivec2(index, 0);
}
`:e[0]===1?n?`
ivec2 getOutputCoords() {
ivec2 resTexRC = ivec2(resultUV.yx *
vec2(outTexShape[0], outTexShape[1]));
int index = resTexRC.x * outTexShape[1] + resTexRC.y;
return ivec2(0, index);
}
`:`
ivec2 getOutputCoords() {
ivec2 resTexRC = ivec2(resultUV.yx *
vec2(${t[0]}, ${t[1]}));
int index = resTexRC.x * ${t[1]} + resTexRC.y;
return ivec2(0, index);
}
`:n?`
ivec2 getOutputCoords() {
ivec2 resTexRC = ivec2(resultUV.yx *
vec2(outTexShape[0], outTexShape[1]));
int index = resTexRC.x * outTexShape[1] + resTexRC.y;
int r = index / outShape[1];
int c = index - r * outShape[1];
return ivec2(r, c);
}
`:`
ivec2 getOutputCoords() {
ivec2 resTexRC = ivec2(resultUV.yx *
vec2(${t[0]}, ${t[1]}));
int index = resTexRC.x * ${t[1]} + resTexRC.y;
int r = index / ${e[1]};
int c = index - r * ${e[1]};
return ivec2(r, c);
}
`}function No(e){return`offset${e}`}function dY(e){let t=e.name,n="get"+t.charAt(0).toUpperCase()+t.slice(1),a=En();return`
vec4 ${n}() {
return ${a.texture2D}(${t}, halfCR);
}
`}function hY(e,t){let n=e.name,a="get"+n.charAt(0).toUpperCase()+n.slice(1);if(e.shapeInfo.isUniform)return`float ${a}() {return ${n};}`;let[r,s]=e.shapeInfo.texShape;if(r===1&&s===1)return`
float ${a}() {
return sampleTexture(${n}, halfCR);
}
`;let i=No(n);if(t)return`
float ${a}() {
vec2 uv = uvFromFlat(${n}TexShape[0], ${n}TexShape[1], ${i});
return sampleTexture(${n}, uv);
}
`;let[o,l]=e.shapeInfo.texShape;return`
float ${a}() {
vec2 uv = uvFromFlat(${o}, ${l}, ${i});
return sampleTexture(${n}, uv);
}
`}function mY(e,t){let n=e.name,a="get"+n.charAt(0).toUpperCase()+n.slice(1),r=e.shapeInfo.texShape,s=En();if(t)return`
vec4 ${a}(int index) {
ivec2 packedTexShape = ivec2(ceil(float(${n}TexShape[0]) / 2.0), ceil(float(${n}TexShape[1]) / 2.0));
vec2 uv = packedUVfrom1D(
packedTexShape[0], packedTexShape[1], index);
return ${s.texture2D}(${n}, uv);
}
`;let i=[Math.ceil(r[0]/2),Math.ceil(r[1]/2)];return`
vec4 ${a}(int index) {
vec2 uv = packedUVfrom1D(
${i[0]}, ${i[1]}, index);
return ${s.texture2D}(${n}, uv);
}
`}function fY(e,t){let n=e.name,a="get"+n.charAt(0).toUpperCase()+n.slice(1);if(e.shapeInfo.isUniform)return`
float ${a}(int index) {
${tp(e)}
}
`;let r=e.shapeInfo.texShape,s=r[0],i=r[1];if(i===1&&s===1)return`
float ${a}(int index) {
return sampleTexture(${n}, halfCR);
}
`;let o=No(n);return i===1?t?`
float ${a}(int index) {
vec2 uv = vec2(0.5, (float(index + ${o}) + 0.5) / float(${n}TexShape[0]));
return sampleTexture(${n}, uv);
}
`:`
float ${a}(int index) {
vec2 uv = vec2(0.5, (float(index + ${o}) + 0.5) / ${s}.0);
return sampleTexture(${n}, uv);
}
`:s===1?t?`
float ${a}(int index) {
vec2 uv = vec2((float(index + ${o}) + 0.5) / float(${n}TexShape[1]), 0.5);
return sampleTexture(${n}, uv);
}
`:`
float ${a}(int index) {
vec2 uv = vec2((float(index + ${o}) + 0.5) / ${i}.0, 0.5);
return sampleTexture(${n}, uv);
}
`:t?`
float ${a}(int index) {
vec2 uv = uvFromFlat(${n}TexShape[0], ${n}TexShape[1], index + ${o});
return sampleTexture(${n}, uv);
}
`:`
float ${a}(int index) {
vec2 uv = uvFromFlat(${s}, ${i}, index + ${o});
return sampleTexture(${n}, uv);
}
`}function gY(e,t){let n=e.shapeInfo.logicalShape,a=e.name,r="get"+a.charAt(0).toUpperCase()+a.slice(1),s=e.shapeInfo.texShape,i=s[0],o=s[1],l=En();if(s!=null&&w.arraysEqual(n,s))return t?`
vec4 ${r}(int row, int col) {
vec2 uv = (vec2(col, row) + halfCR) / vec2(${a}TexShape[1], ${a}TexShape[0]);
return ${l.texture2D}(${a}, uv);
}
`:`
vec4 ${r}(int row, int col) {
vec2 uv = (vec2(col, row) + halfCR) / vec2(${o}.0, ${i}.0);
return ${l.texture2D}(${a}, uv);
}
`;if(t)return`
vec4 ${r}(int row, int col) {
ivec2 packedTexShape = ivec2(ceil(float(${a}TexShape[0]) / 2.0), ceil(float(${a}TexShape[1]) / 2.0));
int valuesPerRow = int(ceil(float(${a}Shape[1]) / 2.0));
vec2 uv = packedUVfrom2D(valuesPerRow, packedTexShape[0], packedTexShape[1], row, col);
return ${l.texture2D}(${a}, uv);
}
`;let u=[Math.ceil(s[0]/2),Math.ceil(s[1]/2)],p=Math.ceil(n[1]/2);return`
vec4 ${r}(int row, int col) {
vec2 uv = packedUVfrom2D(${p}, ${u[0]}, ${u[1]}, row, col);
return ${l.texture2D}(${a}, uv);
}
`}function yY(e,t){let n=e.shapeInfo.logicalShape,a=e.name,r="get"+a.charAt(0).toUpperCase()+a.slice(1),s=e.shapeInfo.texShape;if(s!=null&&w.arraysEqual(n,s)){if(t)return`
float ${r}(int row, int col) {
vec2 uv = (vec2(col, row) + halfCR) / vec2(${a}TexShape[1], ${a}TexShape[0]);
return sampleTexture(${a}, uv);
}
`;let c=s[0],h=s[1];return`
float ${r}(int row, int col) {
vec2 uv = (vec2(col, row) + halfCR) / vec2(${h}.0, ${c}.0);
return sampleTexture(${a}, uv);
}
`}let{newShape:i,keptDims:o}=w.squeezeShape(n),l=i;if(l.length<n.length){let c=np(e,l),h=["row","col"];return`
${ep(c,t)}
float ${r}(int row, int col) {
return ${r}(${ap(h,o)});
}
`}if(e.shapeInfo.isUniform)return`
float ${r}(int row, int col) {
int index = round(dot(vec2(row, col), vec2(${n[1]}, 1)));
${tp(e)}
}
`;let u=s[0],p=s[1],d=No(a);return p===1?t?`
float ${r}(int row, int col) {
float index = dot(vec3(row, col, ${d}), vec3(${a}Shape[1], 1, 1));
vec2 uv = vec2(0.5, (index + 0.5) / float(${a}TexShape[0]));
return sampleTexture(${a}, uv);
}
`:`
float ${r}(int row, int col) {
float index = dot(vec3(row, col, ${d}), vec3(${n[1]}, 1, 1));
vec2 uv = vec2(0.5, (index + 0.5) / ${u}.0);
return sampleTexture(${a}, uv);
}
`:u===1?t?`
float ${r}(int row, int col) {
float index = dot(vec3(row, col, ${d}), vec3(${a}Shape[1], 1, 1));
vec2 uv = vec2((index + 0.5) / float(${a}TexShape[1]), 0.5);
return sampleTexture(${a}, uv);
}
`:`
float ${r}(int row, int col) {
float index = dot(vec3(row, col, ${d}), vec3(${n[1]}, 1, 1));
vec2 uv = vec2((index + 0.5) / ${p}.0, 0.5);
return sampleTexture(${a}, uv);
}
`:t?`
float ${r}(int row, int col) {
// Explicitly use integer operations as dot() only works on floats.
int index = row * ${a}Shape[1] + col + ${d};
vec2 uv = uvFromFlat(${a}TexShape[0], ${a}TexShape[1], index);
return sampleTexture(${a}, uv);
}
`:`
float ${r}(int row, int col) {
// Explicitly use integer operations as dot() only works on floats.
int index = row * ${n[1]} + col + ${d};
vec2 uv = uvFromFlat(${u}, ${p}, index);
return sampleTexture(${a}, uv);
}
`}function bY(e,t){let n=e.shapeInfo.logicalShape,a=e.name,r="get"+a.charAt(0).toUpperCase()+a.slice(1),s=e.shapeInfo.texShape,i=[Math.ceil(s[0]/2),Math.ceil(s[1]/2)];if(n[0]===1){let c=n.slice(1),h=[1,2],m=np(e,c),f=["b","row","col"];return`
${rC(m,t)}
vec4 ${r}(int b, int row, int col) {
return ${r}(${ap(f,h)});
}
`}let o=En();if(t)return`
vec4 ${r}(int b, int row, int col) {
ivec2 packedTexShape = ivec2(ceil(float(${a}TexShape[0]) / 2.0), ceil(float(${a}TexShape[1]) / 2.0));
int valuesPerRow = int(ceil(float(${a}Shape[2]) / 2.0));
int texelsInBatch = valuesPerRow * int(ceil(float(${a}Shape[1]) / 2.0));
vec2 uv = packedUVfrom3D(
packedTexShape[0], packedTexShape[1], texelsInBatch, valuesPerRow, b, row, col);
return ${o.texture2D}(${a}, uv);
}
`;let l=i[0],u=i[1],p=Math.ceil(n[2]/2),d=p*Math.ceil(n[1]/2);return`
vec4 ${r}(int b, int row, int col) {
vec2 uv = packedUVfrom3D(
${l}, ${u}, ${d}, ${p}, b, row, col);
return ${o.texture2D}(${a}, uv);
}
`}function xY(e,t){let n=e.shapeInfo.logicalShape,a=e.name,r="get"+a.charAt(0).toUpperCase()+a.slice(1),s=n[1]*n[2],i=n[2],{newShape:o,keptDims:l}=w.squeezeShape(n),u=o;if(u.length<n.length){let f=np(e,u),g=["row","col","depth"];return`
${ep(f,t)}
float ${r}(int row, int col, int depth) {
return ${r}(${ap(g,l)});
}
`}if(e.shapeInfo.isUniform)return`
float ${r}(int row, int col, int depth) {
int index = round(dot(vec3(row, col, depth),
vec3(${s}, ${i}, 1)));
${tp(e)}
}
`;let p=e.shapeInfo.texShape,d=p[0],c=p[1],h=e.shapeInfo.flatOffset;if(c===s&&h==null)return t?`
float ${r}(int row, int col, int depth) {
int stride1 = ${a}Shape[2];
float texR = float(row);
float texC = dot(vec2(col, depth), vec2(stride1, 1));
vec2 uv = (vec2(texC, texR) + halfCR) /
vec2(${a}TexShape[1], ${a}TexShape[0]);
return sampleTexture(${a}, uv);
}
`:`
float ${r}(int row, int col, int depth) {
float texR = float(row);
float texC = dot(vec2(col, depth), vec2(${i}, 1));
vec2 uv = (vec2(texC, texR) + halfCR) /
vec2(${c}.0, ${d}.0);
return sampleTexture(${a}, uv);
}
`;if(c===i&&h==null)return t?`
float ${r}(int row, int col, int depth) {
float texR = dot(vec2(row, col), vec2(${a}Shape[1], 1));
float texC = float(depth);
vec2 uv = (vec2(texC, texR) + halfCR) / vec2(${a}TexShape[1], ${a}TexShape[0]);
return sampleTexture(${a}, uv);
}
`:`
float ${r}(int row, int col, int depth) {
float texR = dot(vec2(row, col), vec2(${n[1]}, 1));
float texC = float(depth);
vec2 uv = (vec2(texC, texR) + halfCR) / vec2(${c}.0, ${d}.0);
return sampleTexture(${a}, uv);
}
`;let m=No(a);return t?`
float ${r}(int row, int col, int depth) {
// Explicitly use integer operations as dot() only works on floats.
int stride0 = ${a}Shape[1] * ${a}Shape[2];
int stride1 = ${a}Shape[2];
int index = row * ${s} + col * ${i} + depth + ${m};
vec2 uv = uvFromFlat(${a}TexShape[0], ${a}TexShape[1], index);
return sampleTexture(${a}, uv);
}
`:`
float ${r}(int row, int col, int depth) {
// Explicitly use integer operations as dot() only works on floats.
int index = row * ${s} + col * ${i} + depth + ${m};
vec2 uv = uvFromFlat(${d}, ${c}, index);
return sampleTexture(${a}, uv);
}
`}function vY(e,t){let n=e.name,a="get"+n.charAt(0).toUpperCase()+n.slice(1),r=En();if(t)return`
vec4 ${a}(int b2, int b, int row, int col) {
int valuesPerRow = int(ceil(float(${n}Shape[3]) / 2.0));
int texelsInBatch = valuesPerRow * int(ceil(float(${n}Shape[2]) / 2.0));
int index = b * texelsInBatch + (row / 2) * valuesPerRow + (col / 2);
texelsInBatch *= ${n}Shape[1];
index = b2 * texelsInBatch + index;
ivec2 packedTexShape = ivec2(ceil(float(${n}TexShape[0]) / 2.0), ceil(float(${n}TexShape[1]) / 2.0));
int texR = index / packedTexShape[1];
int texC = index - texR * packedTexShape[1];
vec2 uv = (vec2(texC, texR) + halfCR) / vec2(packedTexShape[1], packedTexShape[0]); return ${r.texture2D}(${n}, uv);
}
`;let s=e.shapeInfo.logicalShape,i=s.length,o=e.shapeInfo.texShape,l=[Math.ceil(o[0]/2),Math.ceil(o[1]/2)],u=l[0],p=l[1],d=Math.ceil(s[i-1]/2),c=d*Math.ceil(s[i-2]/2),h="int b, int row, int col",m=`b * ${c} + (row / 2) * ${d} + (col / 2)`;for(let f=2;f<i-1;f++)h=`int b${f}, `+h,c*=s[i-f-1],m=`b${f} * ${c} + `+m;return`
vec4 ${a}(${h}) {
int index = ${m};
int texR = index / ${p};
int texC = index - texR * ${p};
vec2 uv = (vec2(texC, texR) + halfCR) / vec2(${p}, ${u});
return ${r.texture2D}(${n}, uv);
}
`}function wY(e,t){let n=e.shapeInfo.logicalShape,a=e.name,r="get"+a.charAt(0).toUpperCase()+a.slice(1),s=n[3],i=n[2]*s,o=n[1]*i,{newShape:l,keptDims:u}=w.squeezeShape(n);if(l.length<n.length){let b=np(e,l),x=["row","col","depth","depth2"];return`
${ep(b,t)}
float ${r}(int row, int col, int depth, int depth2) {
return ${r}(${ap(x,u)});
}
`}if(e.shapeInfo.isUniform)return`
float ${r}(int row, int col, int depth, int depth2) {
int index = round(dot(vec4(row, col, depth, depth2),
vec4(${o}, ${i}, ${s}, 1)));
${tp(e)}
}
`;let p=e.shapeInfo.flatOffset,d=e.shapeInfo.texShape,c=d[0],h=d[1],m=`int stride2 = ${a}Shape[3];`,f=`int stride1 = ${a}Shape[2] * stride2;`,g=`int stride0 = ${a}Shape[1] * stride1;`;if(h===o&&p==null)return t?`
float ${r}(int row, int col, int depth, int depth2) {
${m}
${f}
float texR = float(row);
float texC =
dot(vec3(col, depth, depth2),
vec3(stride1, stride2, 1));
vec2 uv = (vec2(texC, texR) + halfCR) /
vec2(${a}TexShape[1], ${a}TexShape[0]);
return sampleTexture(${a}, uv);
}
`:`
float ${r}(int row, int col, int depth, int depth2) {
float texR = float(row);
float texC =
dot(vec3(col, depth, depth2),
vec3(${i}, ${s}, 1));
vec2 uv = (vec2(texC, texR) + halfCR) /
vec2(${h}.0, ${c}.0);
return sampleTexture(${a}, uv);
}
`;if(h===s&&p==null)return t?`
float ${r}(int row, int col, int depth, int depth2) {
float texR = dot(vec3(row, col, depth),
vec3(${a}Shape[1] * ${a}Shape[2], ${a}Shape[2], 1));
float texC = float(depth2);
vec2 uv = (vec2(texC, texR) + halfCR) /
vec2(${a}TexShape[1], ${a}TexShape[0]);
return sampleTexture(${a}, uv);
}
`:`
float ${r}(int row, int col, int depth, int depth2) {
float texR = dot(vec3(row, col, depth),
vec3(${n[1]*n[2]}, ${n[2]}, 1));
float texC = float(depth2);
vec2 uv = (vec2(texC, texR) + halfCR) /
vec2(${h}.0, ${c}.0);
return sampleTexture(${a}, uv);
}
`;let y=No(a);return t?`
float ${r}(int row, int col, int depth, int depth2) {
// Explicitly use integer operations as dot() only works on floats.
${m}
${f}
${g}
int index = row * stride0 + col * stride1 +
depth * stride2 + depth2;
vec2 uv = uvFromFlat(${a}TexShape[0], ${a}TexShape[1], index + ${y});
return sampleTexture(${a}, uv);
}
`:`
float ${r}(int row, int col, int depth, int depth2) {
// Explicitly use integer operations as dot() only works on floats.
int index = row * ${o} + col * ${i} +
depth * ${s} + depth2;
vec2 uv = uvFromFlat(${c}, ${h}, index + ${y});
return sampleTexture(${a}, uv);
}
`}function kY(e){let t=e.shapeInfo.logicalShape,n=e.name,a="get"+n.charAt(0).toUpperCase()+n.slice(1),r=t[4],s=t[3]*r,i=t[2]*s,o=t[1]*i,{newShape:l,keptDims:u}=w.squeezeShape(t);if(l.length<t.length){let f=np(e,l),g=["row","col","depth","depth2","depth3"];return`
${ep(f)}
float ${a}(int row, int col, int depth, int depth2, int depth3) {
return ${a}(${ap(g,u)});
}
`}if(e.shapeInfo.isUniform)return`
float ${a}(int row, int col, int depth, int depth2, int depth3) {
float index = dot(
vec4(row, col, depth, depth2),
vec4(${o}, ${i}, ${s}, ${r})) +
depth3;
${tp(e)}
}
`;let p=e.shapeInfo.flatOffset,d=e.shapeInfo.texShape,c=d[0],h=d[1];if(h===o&&p==null)return`
float ${a}(int row, int col, int depth, int depth2, int depth3) {
int texR = row;
float texC = dot(vec4(col, depth, depth2, depth3),
vec4(${i}, ${s}, ${r}, 1));
vec2 uv = (vec2(texC, texR) + halfCR) /
vec2(${h}.0, ${c}.0);
return sampleTexture(${n}, uv);
}
`;if(h===r&&p==null)return`
float ${a}(int row, int col, int depth, int depth2, int depth3) {
float texR = dot(
vec4(row, col, depth, depth2),
vec4(${t[1]*t[2]*t[3]},
${t[2]*t[3]}, ${t[3]}, 1));
int texC = depth3;
vec2 uv = (vec2(texC, texR) + halfCR) /
vec2(${h}.0, ${c}.0);
return sampleTexture(${n}, uv);
}
`;let m=No(n);return`
float ${a}(int row, int col, int depth, int depth2, int depth3) {
// Explicitly use integer operations as dot() only works on floats.
int index = row * ${o} + col * ${i} + depth * ${s} +
depth2 * ${r} + depth3 + ${m};
vec2 uv = uvFromFlat(${c}, ${h}, index);
return sampleTexture(${n}, uv);
}
`}function IY(e){let t=e.shapeInfo.logicalShape,n=e.name,a="get"+n.charAt(0).toUpperCase()+n.slice(1),{newShape:r,keptDims:s}=w.squeezeShape(t);if(r.length<t.length){let g=np(e,r),y=["row","col","depth","depth2","depth3","depth4"];return`
${ep(g)}
float ${a}(int row, int col, int depth,
int depth2, int depth3, int depth4) {
return ${a}(${ap(y,s)});
}
`}let i=t[5],o=t[4]*i,l=t[3]*o,u=t[2]*l,p=t[1]*u;if(e.shapeInfo.isUniform)return`
float ${a}(int row, int col, int depth,
int depth2, int depth3, int depth4) {
int index = round(dot(
vec4(row, col, depth, depth2),
vec4(${p}, ${u}, ${l}, ${o})) +
dot(
vec2(depth3, depth4),
vec2(${i}, 1)));
${tp(e)}
}
`;let d=e.shapeInfo.flatOffset,c=e.shapeInfo.texShape,h=c[0],m=c[1];if(m===p&&d==null)return`
float ${a}(int row, int col, int depth,
int depth2, int depth3, int depth4) {
int texR = row;
float texC = dot(vec4(col, depth, depth2, depth3),
vec4(${u}, ${l}, ${o}, ${i})) +
float(depth4);
vec2 uv = (vec2(texC, texR) + halfCR) /
vec2(${m}.0, ${h}.0);
return sampleTexture(${n}, uv);
}
`;if(m===i&&d==null)return`
float ${a}(int row, int col, int depth,
int depth2, int depth3, int depth4) {
float texR = dot(vec4(row, col, depth, depth2),
vec4(${t[1]*t[2]*t[3]*t[4]},
${t[2]*t[3]*t[4]},
${t[3]*t[4]},
${t[4]})) + float(depth3);
int texC = depth4;
vec2 uv = (vec2(texC, texR) + halfCR) /
vec2(${m}.0, ${h}.0);
return sampleTexture(${n}, uv);
}
`;let f=No(n);return`
float ${a}(int row, int col, int depth,
int depth2, int depth3, int depth4) {
// Explicitly use integer operations as dot() only works on floats.
int index = row * ${p} + col * ${u} + depth * ${l} +
depth2 * ${o} + depth3 * ${i} + depth4 + ${f};
vec2 uv = uvFromFlat(${h}, ${m}, index);
return sampleTexture(${n}, uv);
}
`}function tp(e){let t=e.name,n=w.sizeFromShape(e.shapeInfo.logicalShape);return n<2?`return ${t};`:`
for (int i = 0; i < ${n}; i++) {
if (i == index) {
return ${t}[i];
}
}
`}function SY(e,t){let n=e.name,a=n.charAt(0).toUpperCase()+n.slice(1),r="get"+a+"AtOutCoords",s=e.shapeInfo.logicalShape.length,i=t.logicalShape.length,o=aC(e.shapeInfo.logicalShape,t.logicalShape),l=mt(i),u=i-s,p,d=["x","y","z","w","u","v"];s===0?p="":i<2&&o.length>=1?p="coords = 0;":p=o.map(g=>`coords.${d[g+u]} = 0;`).join(`
`);let c="";i<2&&s>0?c="coords":c=e.shapeInfo.logicalShape.map((g,y)=>`coords.${d[y+u]}`).join(", ");let h="return outputValue;",m=w.sizeFromShape(e.shapeInfo.logicalShape)===1,f=w.sizeFromShape(t.logicalShape)===1;if(s===1&&!m&&!f)h=`
return vec4(outputValue.xy, outputValue.xy);
`;else if(m&&!f)i===1?h=`
return vec4(outputValue.x, outputValue.x, 0., 0.);
`:h=`
return vec4(outputValue.x);
`;else if(o.length){let g=s-2,y=s-1;o.indexOf(g)>-1&&o.indexOf(y)>-1?h="return vec4(outputValue.x);":o.indexOf(g)>-1?h="return vec4(outputValue.x, outputValue.y, outputValue.x, outputValue.y);":o.indexOf(y)>-1&&(h="return vec4(outputValue.xx, outputValue.zz);")}return`
vec4 ${r}() {
${l} coords = getOutputCoords();
${p}
vec4 outputValue = get${a}(${c});
${h}
}
`}function NY(e,t){let n=e.name,a=n.charAt(0).toUpperCase()+n.slice(1),r="get"+a+"AtOutCoords",s=t.texShape,i=e.shapeInfo.texShape,o=e.shapeInfo.logicalShape.length,l=t.logicalShape.length;if(!e.shapeInfo.isUniform&&o===l&&e.shapeInfo.flatOffset==null&&w.arraysEqual(i,s))return`
float ${r}() {
return sampleTexture(${n}, resultUV);
}
`;let u=mt(l),p=aC(e.shapeInfo.logicalShape,t.logicalShape),d=l-o,c,h=["x","y","z","w","u","v"];o===0?c="":l<2&&p.length>=1?c="coords = 0;":c=p.map(f=>`coords.${h[f+d]} = 0;`).join(`
`);let m="";return l<2&&o>0?m="coords":m=e.shapeInfo.logicalShape.map((f,g)=>`coords.${h[g+d]}`).join(", "),`
float ${r}() {
${u} coords = getOutputCoords();
${c}
return get${a}(${m});
}
`}function mt(e){if(e<=1)return"int";if(e===2)return"ivec2";if(e===3)return"ivec3";if(e===4)return"ivec4";if(e===5)return"ivec5";if(e===6)return"ivec6";throw Error(`GPU for rank ${e} is not yet supported`)}function M0(e,t,n){let{newShape:a,keptDims:r}=w.squeezeShape(t),s=t.length,i=e&&s===3&&t[0]===1,o=i?t.slice(1):a,l=!e&&s>1&&!w.arraysEqual(t,n)&&a.length<s||i;return{useSqueezeShape:l,uniformShape:l?o:t,keptDims:r}}function np(e,t){let n=JSON.parse(JSON.stringify(e));return n.shapeInfo.logicalShape=t,n}function ap(e,t){return t.map(n=>e[n]).join(", ")}function TY(e,t,n,a){let r=n.map((p,d)=>{let c={logicalShape:p.shape,texShape:p.isUniform?null:p.texData.texShape,isUniform:p.isUniform,isPacked:p.isUniform?!1:p.texData.isPacked,flatOffset:null};return p.texData!=null&&p.texData.slice!=null&&p.texData.slice.flatOffset>0&&(c.flatOffset=p.texData.slice.flatOffset),{name:t.variableNames[d],shapeInfo:c}}),s=r.map(p=>p.shapeInfo),i={logicalShape:a.shape,texShape:a.texData.texShape,isUniform:!1,isPacked:a.texData.isPacked,flatOffset:null},o=G7(r,i,t),l=M_(e.gl,o),u=e.createProgram(l);return X().get("ENGINE_COMPILE_ONLY")?{program:t,fragmentShader:l,source:o,webGLProgram:u,inShapeInfos:s,outShapeInfo:i,uniformLocations:null,customUniformLocations:null,infLoc:null,nanLoc:null,inShapesLocations:null,inTexShapesLocations:null,outShapeLocation:null,outShapeStridesLocation:null,outTexShapeLocation:null}:Object.assign({program:t,fragmentShader:l,source:o,webGLProgram:u,inShapeInfos:s,outShapeInfo:i},iC(e,t,u))}function iC(e,t,n){let a={},r={},s={},i=[],o,l,u,p=null,d=null;d=e.getUniformLocation(n,"NAN",!1),X().getNumber("WEBGL_VERSION")===1&&(p=e.getUniformLocation(n,"INFINITY",!1));let c=!1;for(let h=0;h<t.variableNames.length;h++){let m=t.variableNames[h];a[m]=e.getUniformLocation(n,m,c),a[`offset${m}`]=e.getUniformLocation(n,`offset${m}`,c),t.enableShapeUniforms&&(r[`${m}Shape`]=e.getUniformLocation(n,`${m}Shape`,c),s[`${m}TexShape`]=e.getUniformLocation(n,`${m}TexShape`,c))}return t.enableShapeUniforms&&(o=e.getUniformLocation(n,"outShape",c),u=e.getUniformLocation(n,"outShapeStrides",c),l=e.getUniformLocation(n,"outTexShape",c)),t.customUniforms&&t.customUniforms.forEach((h,m)=>{i[m]=e.getUniformLocation(n,h.name,c)}),{uniformLocations:a,customUniformLocations:i,infLoc:p,nanLoc:d,inShapesLocations:r,inTexShapesLocations:s,outShapeLocation:o,outShapeStridesLocation:u,outTexShapeLocation:l}}function jk(e,t){if(e.length!==t.length)throw Error(`Binary was compiled with ${e.length} inputs, but was executed with ${t.length} inputs`);e.forEach((n,a)=>{let r=n.logicalShape,s=t[a],i=s.shape;if(!w.arraysEqual(r,i))throw Error(`Binary was compiled with different shapes than the current args. Shapes ${r} and ${i} must match`);if(n.isUniform&&s.isUniform)return;let o=n.texShape,l=s.isUniform?null:s.texData.texShape;if(!w.arraysEqual(o,l))throw Error(`Binary was compiled with different texture shapes than the current args. Shape ${o} and ${l} must match`)})}function _Y(e,t,n,a,r){t.program.enableShapeUniforms||(jk(t.inShapeInfos,n),jk([t.outShapeInfo],[a]));let s=a.texData.texture,i=a.texData.texShape;a.texData.isPacked?e.setOutputPackedMatrixTexture(s.texture,i[0],i[1]):e.setOutputMatrixTexture(s.texture,i[0],i[1]),e.setProgram(t.webGLProgram),X().getNumber("WEBGL_VERSION")===1&&t.infLoc!==null&&e.gl.uniform1f(t.infLoc,1/0),t.nanLoc!==null&&e.gl.uniform1f(t.nanLoc,NaN),n.forEach((l,u)=>{let p=t.program.variableNames[u],d=t.uniformLocations[p],c=t.uniformLocations[`offset${p}`],h=t.inShapesLocations[`${p}Shape`],m=t.inTexShapesLocations[`${p}TexShape`];if(h){let{uniformShape:f}=M0(t.program.packedInputs,l.shape,l.texData.texShape);switch(f.length){case 1:e.gl.uniform1iv(h,new Int32Array(f));break;case 2:e.gl.uniform2iv(h,new Int32Array(f));break;case 3:e.gl.uniform3iv(h,new Int32Array(f));break;case 4:e.gl.uniform4iv(h,new Int32Array(f));break;default:break}}if(m&&e.gl.uniform2i(m,l.texData.texShape[0],l.texData.texShape[1]),d!=null){if(l.isUniform){if(w.sizeFromShape(l.shape)<2)e.gl.uniform1f(d,l.uniformValues[0]);else{let f=l.uniformValues;f instanceof Float32Array||(f=new Float32Array(f)),e.gl.uniform1fv(d,f)}return}l.texData.slice!=null&&c!=null&&e.gl.uniform1i(c,l.texData.slice.flatOffset),e.setInputMatrixTexture(l.texData.texture.texture,d,u)}});let o=t.outShapeLocation;if(o)switch(a.shape.length){case 1:e.gl.uniform1iv(o,new Int32Array(a.shape));break;case 2:e.gl.uniform2iv(o,new Int32Array(a.shape));break;case 3:e.gl.uniform3iv(o,new Int32Array(a.shape));break;case 4:e.gl.uniform4iv(o,new Int32Array(a.shape));break;default:break}if(t.outShapeStridesLocation){let l=w.computeStrides(a.shape);switch(a.shape.length){case 2:e.gl.uniform1iv(t.outShapeStridesLocation,new Int32Array(l));break;case 3:e.gl.uniform2iv(t.outShapeStridesLocation,new Int32Array(l));break;case 4:e.gl.uniform3iv(t.outShapeStridesLocation,new Int32Array(l));break;default:break}}t.outTexShapeLocation&&e.gl.uniform2i(t.outTexShapeLocation,a.texData.texShape[0],a.texData.texShape[1]),t.program.customUniforms&&r&&t.program.customUniforms.forEach((l,u)=>{let p=t.customUniformLocations[u],d=r[u];if(l.type==="float")e.gl.uniform1fv(p,d);else if(l.type==="vec2")e.gl.uniform2fv(p,d);else if(l.type==="vec3")e.gl.uniform3fv(p,d);else if(l.type==="vec4")e.gl.uniform4fv(p,d);else if(l.type==="int")e.gl.uniform1iv(p,d);else if(l.type==="ivec2")e.gl.uniform2iv(p,d);else if(l.type==="ivec3")e.gl.uniform3iv(p,d);else if(l.type==="ivec4")e.gl.uniform4iv(p,d);else throw Error(`uniform type ${l.type} is not supported yet.`)}),e.executeProgram()}function CY(e,t,n){let a="";t.concat(n).forEach(i=>{let o=i.texData!=null&&i.texData.slice!=null&&i.texData.slice.flatOffset>0;if(e.enableShapeUniforms&&!i.isUniform){let l=i.texData.texShape,{useSqueezeShape:u,uniformShape:p,keptDims:d}=M0(e.packedInputs,i.shape,l),c="",h="",m="";if(p.length===1&&e.packedInputs){let k=[Math.ceil(l[0]/2),Math.ceil(l[1]/2)];c=`${k[0]>1}_${k[1]>1}`}else if(p.length===2&&!e.packedInputs)h=`${p[0]>1}_${p[1]>1}`;else if(p.length>2&&!e.packedInputs){let k=w.computeStrides(p);m=`${k[0]===l[1]}_${k[k.length-1]===l[1]}`}let f=i.shape.length,g=p.length===2&&w.arraysEqual(i.shape,l),y=w.sizeFromShape(i.shape)===1,b=C.getBroadcastDims(i.shape,n.shape),x=!e.packedInputs&&f===n.shape.length&&w.arraysEqual(l,n.texData.texShape),v=e.packedInputs||p.length>2?"":`${l[0]>1}_${l[1]>1}`;a+=`${f}_${x}_${u?d:""}_${p.length}_${y}_${b}_${g}_${c}_${h}_${m}_${v}_${o}`}else{let l=i.isUniform?"uniform":i.texData.texShape;a+=`${i.shape}_${l}_${o}`}});let r=e.userCode,s=e.constructor.name;return s+="_"+a+"_"+r+`${X().getNumber("WEBGL_VERSION")}`,s}function Hn(e){return X().getBool("WEBGL_USE_SHAPES_UNIFORMS")&&e<=4}var EY=class{constructor(e){this.variableNames=["A"],this.packedInputs=!1,this.packedOutput=!0,this.outPackingScheme=cc.DENSE,this.customUniforms=[{name:"texShape",type:"ivec2"}];let t=En();this.outputShape=e,this.enableShapeUniforms=Hn(this.outputShape.length),this.userCode=`
ivec3 outCoordsFromFlatIndex(int index) {
${this.enableShapeUniforms?rg(["r","c","d"],e):So(["r","c","d"],e)}
return ivec3(r, c, d);
}
void main() {
ivec2 resTexRC = ivec2(resultUV.yx * vec2(texShape[0], texShape[1]));
int index = 4 * (resTexRC.x * texShape[1] + resTexRC.y);
vec4 result = vec4(0.);
for (int i=0; i<4; i++) {
int flatIndex = index + i;
ivec3 rc = outCoordsFromFlatIndex(flatIndex);
result[i] = getA(rc.x, rc.y, rc.z);
}
${t.output} = result;
}
`}},FY=class{constructor(e){this.variableNames=["A"],this.packedInputs=!0,this.packedOutput=!0,this.outPackingScheme=cc.DENSE,this.customUniforms=[{name:"texShape",type:"ivec2"}];let t=En();this.outputShape=e,this.enableShapeUniforms=Hn(this.outputShape.length),this.userCode=`
ivec3 outCoordsFromFlatIndex(int index) {
${this.enableShapeUniforms?rg(["r","c","d"],e):So(["r","c","d"],e)}
return ivec3(r, c, d);
}
void main() {
ivec2 resTexRC = ivec2(resultUV.yx * vec2(texShape[0], texShape[1]));
int index = 4 * (resTexRC.x * texShape[1] + resTexRC.y);
vec4 result = vec4(0.);
for (int i=0; i<4; i++) {
int flatIndex = index + i;
ivec3 rc = outCoordsFromFlatIndex(flatIndex);
result[i] = getChannel(getA(rc.x, rc.y, rc.z), vec2(rc.y, rc.z));
}
${t.output} = result;
}
`}},AY=class{constructor(e){this.variableNames=["A"],this.outTexUsage=da.DOWNLOAD;let t=En();this.outputShape=e,this.userCode=`
${nC}
void main() {
float x = getAAtOutCoords();
${t.output} = encode_float(x);
}
`}},$Y=class{constructor(e){this.variableNames=["A"],this.packedInputs=!0,this.packedOutput=!1,this.outTexUsage=da.DOWNLOAD;let t=En();this.outputShape=e,this.userCode=`
${nC}
void main() {
ivec3 coords = getOutputCoords();
float x = getChannel(getAAtOutCoords(), vec2(coords.y, coords.z));
${t.output} = encode_float(x);
}
`}},DY=class{constructor(e,t=!1){this.variableNames=["A"],this.customUniforms=[{name:"texShape",type:"ivec2"}];let n=En();this.outputShape=e,this.enableShapeUniforms=Hn(this.outputShape.length);let a="result";t&&(a="floor(result * 255. + 0.5)"),this.userCode=`
${this.enableShapeUniforms?R0():D0(e)}
void main() {
ivec3 coords = getOutputCoords();
int flatIndex = getFlatIndex(coords);
int offset = imod(flatIndex, 4);
flatIndex = idiv(flatIndex, 4, 1.);
int r = flatIndex / texShape[1];
int c = imod(flatIndex, texShape[1]);
vec2 uv = (vec2(c, r) + halfCR) / vec2(texShape[1], texShape[0]);
vec4 values = ${n.texture2D}(A, uv);
float result;
if(offset == 0) {
result = values[0];
} else if(offset == 1) {
result = values[1];
} else if(offset == 2) {
result = values[2];
} else {
result = values[3];
}
${n.output} = vec4(${a}, 0., 0., 0.);
}
`}},RY=class{constructor(e,t=!1){this.variableNames=["A"],this.packedInputs=!1,this.packedOutput=!0,this.customUniforms=[{name:"texShape",type:"ivec2"}];let n=En();this.outputShape=e,this.enableShapeUniforms=Hn(this.outputShape.length);let a="",r="result";t&&(r="floor(result * 255. + 0.5)");for(let s=0;s<=1;s++)for(let i=0;i<=1;i++){let o=s*2+i;a+=`
localCoords = coords;
if(localCoords[2] + ${i} < ${this.enableShapeUniforms?"outShape[2]":`${e[2]}`}) {
localCoords[2] += ${i};
if (localCoords[1] + ${s} < ${this.enableShapeUniforms?"outShape[1]":`${e[1]}`}) {
localCoords[1] += ${s};
flatIndex = getFlatIndex(localCoords);
offset = imod(flatIndex, 4);
flatIndex = idiv(flatIndex, 4, 1.);
int r = flatIndex / texShape[1];
int c = imod(flatIndex, texShape[1]);
vec2 uv = (vec2(c, r) + halfCR) / vec2(texShape[1], texShape[0]);
values = ${n.texture2D}(A, uv);
if (offset == 0) {
result[${o}] = values[0];
} else if (offset == 1) {
result[${o}] = values[1];
} else if (offset == 2) {
result[${o}] = values[2];
} else {
result[${o}] = values[3];
}
}
}
`}this.userCode=`
${this.enableShapeUniforms?R0():D0(e)}
void main() {
ivec3 coords = getOutputCoords();
vec4 result = vec4(0.);
int flatIndex, r, c, offset;
ivec3 localCoords;
vec2 uv;
vec4 values;
${a}
${n.output} = ${r};
}
`}},oC={};Me(oC,{bindVertexProgramAttributeStreams:()=>gC,createBufferFromOutputTexture:()=>xC,createFloat16MatrixTexture:()=>dC,createFloat16PackedMatrixTexture:()=>fC,createFloat32MatrixTexture:()=>cC,createIndexBuffer:()=>pC,createPackedMatrixTexture:()=>mC,createUnsignedBytesMatrixTexture:()=>hC,createVertexBuffer:()=>uC,createVertexShader:()=>lC,downloadByteEncodedFloatMatrixFromOutputTexture:()=>wC,downloadFloat32MatrixFromBuffer:()=>vC,downloadMatrixFromPackedOutputTexture:()=>IC,downloadPackedMatrixFromBuffer:()=>kC,getInternalFormatForFloat16MatrixTexture:()=>O0,getInternalFormatForFloat16PackedMatrixTexture:()=>B0,getInternalFormatForFloat32MatrixTexture:()=>P0,getInternalFormatForPackedMatrixTexture:()=>z0,getInternalFormatForUnsignedBytesMatrixTexture:()=>L0,uploadDenseMatrixToTexture:()=>yC,uploadPixelDataToTexture:()=>bC});function lC(e){let t=En(),n=`${t.version}
precision highp float;
${t.attribute} vec3 clipSpacePos;
${t.attribute} vec2 uv;
${t.varyingVs} vec2 resultUV;
void main() {
gl_Position = vec4(clipSpacePos, 1);
resultUV = uv;
}`;return R_(e,n)}function uC(e){let t=new Float32Array([-1,1,0,0,1,-1,-1,0,0,0,1,1,0,1,1,1,-1,0,1,0]);return L_(e,t)}function pC(e){let t=new Uint16Array([0,1,2,2,1,3]);return z_(e,t)}function cd(e,t,n,a,r,s){W_(t,n);let i=B_(e),o=e.TEXTURE_2D;return ge(e,()=>e.bindTexture(o,i)),ge(e,()=>e.texParameteri(o,e.TEXTURE_WRAP_S,e.CLAMP_TO_EDGE)),ge(e,()=>e.texParameteri(o,e.TEXTURE_WRAP_T,e.CLAMP_TO_EDGE)),ge(e,()=>e.texParameteri(o,e.TEXTURE_MIN_FILTER,e.NEAREST)),ge(e,()=>e.texParameteri(o,e.TEXTURE_MAG_FILTER,e.NEAREST)),X().getNumber("WEBGL_VERSION")===1?ge(e,()=>e.texImage2D(o,0,a,t,n,0,r,s,null)):ge(e,()=>e.texStorage2D(o,1,a,t,n)),ge(e,()=>e.bindTexture(e.TEXTURE_2D,null)),{texture:i,texShape:[n,t]}}function P0(e){return e.internalFormatFloat}function cC(e,t,n,a){let[r,s]=pd(t,n);return cd(e,r,s,P0(a),a.textureFormatFloat,e.FLOAT)}function O0(e){return e.internalFormatHalfFloat}function dC(e,t,n,a){let[r,s]=pd(t,n);return cd(e,r,s,O0(a),a.textureFormatFloat,a.textureTypeHalfFloat)}function L0(e){return e.downloadTextureFormat}function hC(e,t,n,a){let[r,s]=pd(t,n);return cd(e,r,s,L0(a),e.RGBA,e.UNSIGNED_BYTE)}function z0(e){return e.internalFormatPackedFloat}function mC(e,t,n,a){let[r,s]=Ju(t,n);return cd(e,r,s,z0(a),e.RGBA,e.FLOAT)}function B0(e){return e.internalFormatPackedHalfFloat}function fC(e,t,n,a){let[r,s]=Ju(t,n);return cd(e,r,s,B0(a),e.RGBA,a.textureTypeHalfFloat)}function gC(e,t,n){return ge(e,()=>e.bindBuffer(e.ARRAY_BUFFER,n)),yx(e,t,"clipSpacePos",n,3,20,0)&&yx(e,t,"uv",n,2,20,12)}function yC(e,t,n,a,r,s){ge(e,()=>e.bindTexture(e.TEXTURE_2D,t));let i,o,l;r instanceof Uint8Array?(i=new Uint8Array(n*a*4),o=e.UNSIGNED_BYTE,l=e.RGBA):(i=new Float32Array(n*a*4),o=e.FLOAT,l=s.internalFormatPackedFloat),i.set(r),X().getNumber("WEBGL_VERSION")===2?ge(e,()=>e.texSubImage2D(e.TEXTURE_2D,0,0,0,n,a,e.RGBA,o,i)):ge(e,()=>e.texImage2D(e.TEXTURE_2D,0,l,n,a,0,e.RGBA,o,i)),ge(e,()=>e.bindTexture(e.TEXTURE_2D,null))}function bC(e,t,n){ge(e,()=>e.bindTexture(e.TEXTURE_2D,t)),n.data instanceof Uint8Array?X().getNumber("WEBGL_VERSION")===2?ge(e,()=>e.texSubImage2D(e.TEXTURE_2D,0,0,0,n.width,n.height,e.RGBA,e.UNSIGNED_BYTE,n.data)):ge(e,()=>e.texImage2D(e.TEXTURE_2D,0,e.RGBA,n.width,n.height,0,e.RGBA,e.UNSIGNED_BYTE,n.data)):X().getNumber("WEBGL_VERSION")===2?ge(e,()=>e.texSubImage2D(e.TEXTURE_2D,0,0,0,e.RGBA,e.UNSIGNED_BYTE,n)):ge(e,()=>e.texImage2D(e.TEXTURE_2D,0,e.RGBA,e.RGBA,e.UNSIGNED_BYTE,n)),ge(e,()=>e.bindTexture(e.TEXTURE_2D,null))}function xC(e,t,n,a){let r=e.createBuffer();ge(e,()=>e.bindBuffer(e.PIXEL_PACK_BUFFER,r));let s=4*4*t*n;return ge(e,()=>e.bufferData(e.PIXEL_PACK_BUFFER,s,e.STREAM_READ)),ge(e,()=>e.readPixels(0,0,n,t,e.RGBA,e.FLOAT,0)),ge(e,()=>e.bindBuffer(e.PIXEL_PACK_BUFFER,null)),r}function vC(e,t,n){let a=e,r=new Float32Array(n);return a.bindBuffer(a.PIXEL_PACK_BUFFER,t),a.getBufferSubData(a.PIXEL_PACK_BUFFER,0,r),a.bindBuffer(a.PIXEL_PACK_BUFFER,null),r}function wC(e,t,n,a){let[r,s]=pd(t,n),i=4,o=new Uint8Array(F7(t*n,i));return ge(e,()=>e.readPixels(0,0,r,s,a.downloadTextureFormat,e.UNSIGNED_BYTE,o)),new Float32Array(o.buffer)}function kC(e,t,n,a,r,s,i,o){let l=e,u=new Float32Array(A7(s,i));return l.bindBuffer(l.PIXEL_PACK_BUFFER,t),l.getBufferSubData(l.PIXEL_PACK_BUFFER,0,u),l.bindBuffer(l.PIXEL_PACK_BUFFER,null),u}function IC(e,t,n){let a=new Float32Array(t*n*4);return ge(e,()=>e.readPixels(0,0,n,t,e.RGBA,e.FLOAT,a)),a}var zh=class{constructor(e){this.outputTexture=null,this.program=null,this.disposed=!1,this.vertexAttrsAreBound=!1,this.itemsToPoll=[];let t=X().getNumber("WEBGL_VERSION");e!=null?(this.gl=e,A_(t,e)):this.gl=Ya(t);let n="WEBGL_color_buffer_float",a="EXT_color_buffer_half_float";if(this.parallelCompilationExtension=this.gl.getExtension("KHR_parallel_shader_compile"),X().getNumber("WEBGL_VERSION")===1){let r="OES_texture_float",s="OES_texture_half_float";if(this.textureFloatExtension=jp(this.gl,r),ha(this.gl,s))this.textureHalfFloatExtension=jp(this.gl,s);else if(X().get("WEBGL_FORCE_F16_TEXTURES"))throw new Error("GL context does not support half float textures, yet the environment flag WEBGL_FORCE_F16_TEXTURES is set to true.");if(this.colorBufferFloatExtension=this.gl.getExtension(n),ha(this.gl,a))this.colorBufferHalfFloatExtension=jp(this.gl,a);else if(X().get("WEBGL_FORCE_F16_TEXTURES"))throw new Error("GL context does not support color renderable half floats, yet the environment flag WEBGL_FORCE_F16_TEXTURES is set to true.")}else if(n="EXT_color_buffer_float",ha(this.gl,n))this.colorBufferFloatExtension=this.gl.getExtension(n);else if(ha(this.gl,a))this.colorBufferHalfFloatExtension=this.gl.getExtension(a);else throw new Error("GL context does not support color renderable floats");this.vertexBuffer=uC(this.gl),this.indexBuffer=pC(this.gl),this.framebuffer=V_(this.gl),this.textureConfig=A0(this.gl,this.textureHalfFloatExtension)}get debug(){return X().getBool("DEBUG")}dispose(){if(this.disposed)return;this.program!=null&&console.warn("Disposing a GPGPUContext that still has a bound WebGLProgram. This is probably a resource leak, delete the program with GPGPUContext.deleteProgram before disposing."),this.outputTexture!=null&&console.warn("Disposing a GPGPUContext that still has a bound output matrix texture. This is probably a resource leak, delete the output matrix texture with GPGPUContext.deleteMatrixTexture before disposing.");let e=this.gl;ge(e,()=>e.finish()),ge(e,()=>e.bindFramebuffer(e.FRAMEBUFFER,null)),ge(e,()=>e.deleteFramebuffer(this.framebuffer)),ge(e,()=>e.bindBuffer(e.ARRAY_BUFFER,null)),ge(e,()=>e.bindBuffer(e.ELEMENT_ARRAY_BUFFER,null)),ge(e,()=>e.deleteBuffer(this.indexBuffer)),this.disposed=!0}createFloat32MatrixTexture(e,t){return this.throwIfDisposed(),cC(this.gl,e,t,this.textureConfig)}createFloat16MatrixTexture(e,t){return this.throwIfDisposed(),dC(this.gl,e,t,this.textureConfig)}createUnsignedBytesMatrixTexture(e,t){return this.throwIfDisposed(),hC(this.gl,e,t,this.textureConfig)}uploadPixelDataToTexture(e,t){this.throwIfDisposed(),bC(this.gl,e,t)}uploadDenseMatrixToTexture(e,t,n,a){this.throwIfDisposed(),yC(this.gl,e,t,n,a,this.textureConfig)}createFloat16PackedMatrixTexture(e,t){return this.throwIfDisposed(),fC(this.gl,e,t,this.textureConfig)}createPackedMatrixTexture(e,t){return this.throwIfDisposed(),mC(this.gl,e,t,this.textureConfig)}deleteMatrixTexture(e){this.throwIfDisposed(),this.outputTexture===e&&(bx(this.gl,this.framebuffer),this.outputTexture=null),ge(this.gl,()=>this.gl.deleteTexture(e))}downloadByteEncodedFloatMatrixFromOutputTexture(e,t,n){return this.downloadMatrixDriver(e,()=>wC(this.gl,t,n,this.textureConfig))}downloadPackedMatrixFromBuffer(e,t,n,a,r,s){return kC(this.gl,e,t,n,a,r,s,this.textureConfig)}downloadFloat32MatrixFromBuffer(e,t){return vC(this.gl,e,t)}createBufferFromTexture(e,t,n){this.bindTextureToFrameBuffer(e);let a=xC(this.gl,t,n,this.textureConfig);return this.unbindTextureToFrameBuffer(),a}createAndWaitForFence(){let e=this.createFence(this.gl);return this.pollFence(e)}createFence(e){let t,n;if(X().getBool("WEBGL_FENCE_API_ENABLED")){let a=e,r=a.fenceSync(a.SYNC_GPU_COMMANDS_COMPLETE,0);e.flush(),n=()=>{let s=a.clientWaitSync(r,0,0);return s===a.ALREADY_SIGNALED||s===a.CONDITION_SATISFIED},t=r}else X().getNumber("WEBGL_DISJOINT_QUERY_TIMER_EXTENSION_VERSION")>0?(t=this.beginQuery(),this.endQuery(),n=()=>this.isQueryAvailable(t,X().getNumber("WEBGL_DISJOINT_QUERY_TIMER_EXTENSION_VERSION"))):n=()=>!0;return{query:t,isFencePassed:n}}downloadMatrixFromPackedTexture(e,t,n){return this.downloadMatrixDriver(e,()=>IC(this.gl,t,n))}createProgram(e){this.throwIfDisposed();let t=this.gl;this.vertexShader==null&&(this.vertexShader=lC(t));let n=P_(t);return ge(t,()=>t.attachShader(n,this.vertexShader)),ge(t,()=>t.attachShader(n,e)),O_(t,n),this.debug&&Rh(t,n),this.vertexAttrsAreBound||(this.setProgram(n),this.vertexAttrsAreBound=gC(t,this.program,this.vertexBuffer)),n}deleteProgram(e){this.throwIfDisposed(),e===this.program&&(this.program=null),e!=null&&ge(this.gl,()=>this.gl.deleteProgram(e))}setProgram(e){this.throwIfDisposed(),this.program=e,this.program!=null&&this.debug&&Rh(this.gl,this.program),ge(this.gl,()=>this.gl.useProgram(e))}getUniformLocation(e,t,n=!0){return this.throwIfDisposed(),n?G_(this.gl,e,t):H_(this.gl,e,t)}getAttributeLocation(e,t){return this.throwIfDisposed(),ge(this.gl,()=>this.gl.getAttribLocation(e,t))}getUniformLocationNoThrow(e,t){return this.throwIfDisposed(),this.gl.getUniformLocation(e,t)}setInputMatrixTexture(e,t,n){this.throwIfDisposed(),this.throwIfNoProgram(),j_(this.gl,e,t,n)}setOutputMatrixTexture(e,t,n){this.setOutputMatrixTextureDriver(e,n,t)}setOutputPackedMatrixTexture(e,t,n){this.throwIfDisposed();let[a,r]=Ju(t,n);this.setOutputMatrixTextureDriver(e,a,r)}setOutputMatrixWriteRegion(e,t,n,a){this.setOutputMatrixWriteRegionDriver(n,e,a,t)}setOutputPackedMatrixWriteRegion(e,t,n,a){throw new Error("setOutputPackedMatrixWriteRegion not implemented.")}debugValidate(){this.program!=null&&Rh(this.gl,this.program),qp(this.gl)}executeProgram(){this.throwIfDisposed(),this.throwIfNoProgram();let e=this.gl;this.debug&&this.debugValidate(),ge(e,()=>e.drawElements(e.TRIANGLES,6,e.UNSIGNED_SHORT,0))}blockUntilAllProgramsCompleted(){this.throwIfDisposed(),ge(this.gl,()=>this.gl.finish())}getQueryTimerExtension(){return this.disjointQueryTimerExtension==null&&(this.disjointQueryTimerExtension=jp(this.gl,X().getNumber("WEBGL_DISJOINT_QUERY_TIMER_EXTENSION_VERSION")===2?"EXT_disjoint_timer_query_webgl2":"EXT_disjoint_timer_query")),this.disjointQueryTimerExtension}getQueryTimerExtensionWebGL2(){return this.getQueryTimerExtension()}getQueryTimerExtensionWebGL1(){return this.getQueryTimerExtension()}beginQuery(){if(X().getNumber("WEBGL_DISJOINT_QUERY_TIMER_EXTENSION_VERSION")===2){let n=this.gl,a=this.getQueryTimerExtensionWebGL2(),r=n.createQuery();return n.beginQuery(a.TIME_ELAPSED_EXT,r),r}let e=this.getQueryTimerExtensionWebGL1(),t=e.createQueryEXT();return e.beginQueryEXT(e.TIME_ELAPSED_EXT,t),t}endQuery(){if(X().getNumber("WEBGL_DISJOINT_QUERY_TIMER_EXTENSION_VERSION")===2){let t=this.gl,n=this.getQueryTimerExtensionWebGL2();t.endQuery(n.TIME_ELAPSED_EXT);return}let e=this.getQueryTimerExtensionWebGL1();e.endQueryEXT(e.TIME_ELAPSED_EXT)}async waitForQueryAndGetTime(e){return await w.repeatedTry(()=>this.disposed||this.isQueryAvailable(e,X().getNumber("WEBGL_DISJOINT_QUERY_TIMER_EXTENSION_VERSION"))),this.getQueryTime(e,X().getNumber("WEBGL_DISJOINT_QUERY_TIMER_EXTENSION_VERSION"))}getQueryTime(e,t){if(t===0)return null;if(t===2){let n=this.gl;return n.getQueryParameter(e,n.QUERY_RESULT)/1e6}else{let n=this.getQueryTimerExtensionWebGL1();return n.getQueryObjectEXT(e,n.QUERY_RESULT_EXT)/1e6}}isQueryAvailable(e,t){if(t===0)return!0;if(t===2){let n=this.gl,a=this.getQueryTimerExtensionWebGL2(),r=n.getQueryParameter(e,n.QUERY_RESULT_AVAILABLE);return this.disjoint==null&&(this.disjoint=this.gl.getParameter(a.GPU_DISJOINT_EXT)),r&&!this.disjoint}else{let n=this.getQueryTimerExtensionWebGL1(),a=n.getQueryObjectEXT(e,n.QUERY_RESULT_AVAILABLE_EXT);return this.disjoint==null&&(this.disjoint=this.gl.getParameter(n.GPU_DISJOINT_EXT)),a&&!this.disjoint}}pollFence(e){return new Promise(t=>{this.addItemToPoll(()=>e.isFencePassed(),()=>t())})}pollItems(){let e=MY(this.itemsToPoll.map(t=>t.isDoneFn));for(let t=0;t<=e;++t){let{resolveFn:n}=this.itemsToPoll[t];n()}this.itemsToPoll=this.itemsToPoll.slice(e+1)}addItemToPoll(e,t){this.itemsToPoll.push({isDoneFn:e,resolveFn:t}),!(this.itemsToPoll.length>1)&&w.repeatedTry(()=>(this.pollItems(),this.itemsToPoll.length===0))}bindTextureToFrameBuffer(e){this.throwIfDisposed(),Mh(this.gl,e,this.framebuffer),this.debug&&qp(this.gl)}unbindTextureToFrameBuffer(){this.outputTexture!=null?(Mh(this.gl,this.outputTexture,this.framebuffer),this.debug&&qp(this.gl)):bx(this.gl,this.framebuffer)}downloadMatrixDriver(e,t){this.bindTextureToFrameBuffer(e);let n=t();return this.unbindTextureToFrameBuffer(),n}setOutputMatrixTextureDriver(e,t,n){this.throwIfDisposed();let a=this.gl;Mh(a,e,this.framebuffer),this.debug&&qp(a),this.outputTexture=e,ge(a,()=>a.viewport(0,0,t,n)),ge(a,()=>a.scissor(0,0,t,n))}setOutputMatrixWriteRegionDriver(e,t,n,a){this.throwIfDisposed(),ge(this.gl,()=>this.gl.scissor(e,t,n,a))}throwIfDisposed(){if(this.disposed)throw new Error("Attempted to use disposed GPGPUContext.")}throwIfNoProgram(){if(this.program==null)throw new Error("No GPU program is currently set.")}};function MY(e){let t=0;for(;t<e.length&&e[t]();++t);return t-1}var{addImpl:PY,bincountImpl:SC,bincountReduceImpl:OY,ceilImpl:LY,concatImpl:zY,equalImpl:BY,expImpl:WY,expm1Impl:VY,floorImpl:UY,gatherNdImpl:GY,gatherV2Impl:HY,greaterImpl:jY,greaterEqualImpl:qY,lessImpl:KY,lessEqualImpl:XY,linSpaceImpl:YY,logImpl:QY,maxImpl:JY,maximumImpl:ZY,minimumImpl:e9,multiplyImpl:t9,negImpl:n9,notEqualImpl:a9,prodImpl:r9,rangeImpl:s9,rsqrtImpl:i9,scatterImpl:o9,sigmoidImpl:l9,simpleAbsImpl:NC,sliceImpl:u9,sparseFillEmptyRowsImpl:p9,sparseReshapeImpl:c9,sparseSegmentReductionImpl:TC,sqrtImpl:d9,stridedSliceImpl:h9,stringNGramsImpl:m9,stringSplitImpl:f9,stringToHashBucketFastImpl:g9,subImpl:y9,tileImpl:b9,topKImpl:x9,transposeImpl:W0,uniqueImpl:v9}=ET;function _C(e,t){return["x","y","z","w","u","v"].slice(0,t).map(n=>`${e}.${n}`)}function Nn(e,t){return t===1?[e]:_C(e,t)}function w9(e,t){if(e===1)return"rc";let n="";for(let a=0;a<e;a++)n+=t[a],a<e-1&&(n+=",");return n}var k9=class{constructor(e){if(this.variableNames=["A"],this.packedInputs=!1,this.packedOutput=!0,this.outputShape=e,this.rank=e.length,this.enableShapeUniforms=Hn(this.outputShape.length),this.rank===0)this.userCode=`
void main() {
setOutput(vec4(getA(), 0., 0., 0.));
}
`;else{let t=Nn("rc",this.rank),n=mt(this.rank),a=this.getOutOfBoundsCondition(t),r=this.getSetup(t),s=this.getOutput(t);this.userCode=`
void main() {
${n} rc = getOutputCoords();
if(${a}) {
setOutput(vec4(0));
} else {
${r}
setOutput(vec4(${s}));
}
}
`}}getSourceCoordsArr(e){let t=[];for(let n=0;n<=1;n++)for(let a=0;a<=1;a++){let r=`${n===0?"r":"rp1"}, ${a===0?"c":"cp1"}`;for(let s=2;s<this.rank;s++)r=`${e[e.length-1-s]},`+r;t.push(r)}return t}getOutOfBoundsCondition(e){if(this.rank===1)return`rc > ${this.enableShapeUniforms?"outShape":this.outputShape[0]}`;let t="";for(let n=this.rank-2;n<this.rank;n++)t+=`${e[n]} >= ${this.enableShapeUniforms?`outShape[${n}]`:this.outputShape[n]}`,n<this.rank-1&&(t+="||");return t}getSetup(e){if(this.rank===1)return"";let t=e.slice(-2),n=this.enableShapeUniforms?`outShape[${this.rank} - 1]`:this.outputShape[this.rank-1],a=this.enableShapeUniforms?`outShape[${this.rank} - 2]`:this.outputShape[this.rank-2];return`
int r = ${t[0]};
int c = ${t[1]};
int rp1 = r + 1;
int cp1 = c + 1;
bool cEdge = cp1 >= ${n};
bool rEdge = rp1 >= ${a};
`}getOutput(e){let t=this.getSourceCoordsArr(e);return this.rank===1?`getA(rc), (rc + 1 >= ${this.enableShapeUniforms?"outShape":this.outputShape[0]} ? 0. : getA(rc + 1)), 0, 0`:`getA(${t[0]}),
cEdge ? 0. : getA(${t[1]}),
rEdge ? 0. : getA(${t[2]}),
rEdge || cEdge ? 0. : getA(${t[3]})`}},CC=class{constructor(e,t){this.variableNames=["A"],this.packedInputs=!0,this.packedOutput=!0,this.customUniforms=[{name:"inputShape",type:"ivec3"}],this.outputShape=e,this.enableShapeUniforms=Hn(this.outputShape.length);let n="";for(let a=0;a<4;a++){let r="thisRC = rc;";a%2===1&&(r+="thisRC.z += 1;"),a>1&&(r+="thisRC.y += 1;"),n+=`
${r}
${a>0?"if(thisRC.y < rows && thisRC.z < cols){":""}
int flatIndex = getFlatIndex(thisRC);
ivec3 inputRC = inputCoordsFromReshapedOutCoords(flatIndex);
vec2 inputRCInnerDims = vec2(float(inputRC.y),float(inputRC.z));
result[${a}] =
getChannel(getA(inputRC.x, inputRC.y, inputRC.z), inputRCInnerDims);
${a>0?"}":""}
`}this.userCode=`
${I9(t,this.enableShapeUniforms)}
${this.enableShapeUniforms?R0():D0(e)}
void main() {
ivec3 rc = getOutputCoords();
vec4 result = vec4(0.);
ivec3 thisRC;
int rows = ${this.enableShapeUniforms?"outShape[1]":e[1]};
int cols = ${this.enableShapeUniforms?"outShape[2]":e[2]};
${n}
setOutput(result);
}
`}};function I9(e,t){return`
ivec3 inputCoordsFromReshapedOutCoords(int index) {
${t?U7(["r","c","d"],"inputShape"):So(["r","c","d"],e)}
return ivec3(r, c, d);
}
`}var S9=class{constructor(e){this.gpgpu=e,this.numUsedTextures=0,this.numFreeTextures=0,this._numBytesAllocated=0,this._numBytesFree=0,this.freeTextures={},this.logEnabled=!1,this.usedTextures={}}acquireTexture(e,t,n){let a=Kk(t,n),r=Xk(e,a,n);r in this.freeTextures||(this.freeTextures[r]=[]),r in this.usedTextures||(this.usedTextures[r]=[]);let s=qk(e,a,this.gpgpu.gl,this.gpgpu.textureConfig,n);if(this.freeTextures[r].length>0){this.numFreeTextures--,this.numUsedTextures++,this._numBytesFree-=s,this.log();let o=this.freeTextures[r].shift();return this.usedTextures[r].push(o),o}let i;return a===un.PACKED_2X2_FLOAT32?i=this.gpgpu.createPackedMatrixTexture(e[0],e[1]):a===un.PACKED_2X2_FLOAT16?i=this.gpgpu.createFloat16PackedMatrixTexture(e[0],e[1]):a===un.UNPACKED_FLOAT32?i=this.gpgpu.createFloat32MatrixTexture(e[0],e[1]):a===un.UNPACKED_FLOAT16?i=this.gpgpu.createFloat16MatrixTexture(e[0],e[1]):a===un.PACKED_4X1_UNSIGNED_BYTE&&(i=this.gpgpu.createUnsignedBytesMatrixTexture(e[0],e[1])),this.usedTextures[r].push(i),this.numUsedTextures++,this._numBytesAllocated+=s,this.log(),i}releaseTexture(e,t,n,a){if(this.freeTextures==null)return;let r=Kk(n,a),s=Xk(t,r,a);s in this.freeTextures||(this.freeTextures[s]=[]);let i=qk(t,r,this.gpgpu.gl,this.gpgpu.textureConfig,a),o=X().get("WEBGL_DELETE_TEXTURE_THRESHOLD");o!==-1&&this._numBytesAllocated>o?(this.gpgpu.deleteMatrixTexture(e.texture),this._numBytesAllocated-=i):(this.freeTextures[s].push(e),this.numFreeTextures++,this._numBytesFree+=i),this.numUsedTextures--;let l=this.usedTextures[s],u=l.indexOf(e);if(u<0)throw new Error("Cannot release a texture that was never provided by this texture manager");l.splice(u,1),this.log()}log(){if(!this.logEnabled)return;let e=this.numFreeTextures+this.numUsedTextures;console.log("Free/Used",`${this.numFreeTextures} / ${this.numUsedTextures}`,`(${e})`);let t=this._numBytesFree/this._numBytesAllocated;console.log(`Bytes allocated: ${this._numBytesAllocated}`),console.log(`Bytes unused: ${this._numBytesFree} (${Math.round(100*t)}%)`)}get numBytesAllocated(){return this._numBytesAllocated}get numBytesFree(){return this._numBytesFree}getNumUsedTextures(){return this.numUsedTextures}getNumFreeTextures(){return this.numFreeTextures}dispose(){if(this.freeTextures!=null){for(let e in this.freeTextures)this.freeTextures[e].forEach(t=>{this.gpgpu.deleteMatrixTexture(t.texture)});for(let e in this.usedTextures)this.usedTextures[e].forEach(t=>{this.gpgpu.deleteMatrixTexture(t.texture)});this.freeTextures=null,this.usedTextures=null,this.numUsedTextures=0,this.numFreeTextures=0,this._numBytesAllocated=0,this._numBytesFree=0}}};function N9(e,t){let n=e;if(t===n.R32F)return 4;if(t===n.R16F)return 2;if(t===n.RGBA32F||t===e.RGBA)return 16;if(t===n.RGBA16F)return 8;if(t===n.RGBA8)return 4;throw new Error(`Unknown internal format ${t}`)}function qk(e,t,n,a,r){let s=T9(t,a),i;if(r){let[l,u]=Ju(e[0],e[1]);i=l*u}else{let[l,u]=pd(e[0],e[1]);i=l*u}let o=N9(n,s);return i*o}function T9(e,t){switch(e){case un.PACKED_2X2_FLOAT32:return z0(t);case un.PACKED_2X2_FLOAT16:return B0(t);case un.UNPACKED_FLOAT32:return P0(t);case un.UNPACKED_FLOAT16:return O0(t);case un.PACKED_4X1_UNSIGNED_BYTE:return L0(t);default:throw new Error(`Unknown physical texture type ${e}`)}}function _9(e){return X().getBool("WEBGL_RENDER_FLOAT32_ENABLED")?e?un.PACKED_2X2_FLOAT32:un.UNPACKED_FLOAT32:e?un.PACKED_2X2_FLOAT16:un.UNPACKED_FLOAT16}function Kk(e,t){if(e===da.UPLOAD)return un.PACKED_2X2_FLOAT32;if(e===da.RENDER||e==null)return _9(t);if(e===da.DOWNLOAD||e===da.PIXELS)return un.PACKED_4X1_UNSIGNED_BYTE;throw new Error(`Unknown logical texture type ${e}`)}function Xk(e,t,n){return`${e[0]}_${e[1]}_${t}_${n}`}var _r=class{constructor(e,t){this.variableNames=["A"],this.outputShape=e,this.enableShapeUniforms=Hn(this.outputShape.length),this.userCode=`
float unaryOperation(float x) {
${t}
}
void main() {
float x = getAAtOutCoords();
float y = unaryOperation(x);
setOutput(y);
}
`}},Aa="if (isnan(x)) return x;",C9="return x;",Yk="return abs(x);",E9="return (x >= 0.0) ? x : (exp(x) - 1.0);",F9=Aa+`
return (x < 0.0) ? 0.0 : x;
`,A9=Aa+`
return (x < 0.0) ? 0.0 : min(6.0, x);
`,pl="return x;",$9="return 1.0 / (1.0 + exp(-1.0 * x));",D9="return x;",R9=`
vec4 result;
result.r = (x.r >= 0.0) ? x.r : (exp(x.r) - 1.0);
result.g = (x.g >= 0.0) ? x.g : (exp(x.g) - 1.0);
result.b = (x.b >= 0.0) ? x.b : (exp(x.b) - 1.0);
result.a = (x.a >= 0.0) ? x.a : (exp(x.a) - 1.0);
return result;
`,M9=`
vec4 result = x * vec4(greaterThanEqual(x, vec4(0.0)));
bvec4 isNaN = isnan(x);
result.r = isNaN.r ? x.r : result.r;
result.g = isNaN.g ? x.g : result.g;
result.b = isNaN.b ? x.b : result.b;
result.a = isNaN.a ? x.a : result.a;
return result;
`,P9=`
vec4 result = min(x, vec4(6.)) * vec4(greaterThanEqual(x, vec4(0.0)));
bvec4 isNaN = isnan(x);
result.r = isNaN.r ? x.r : result.r;
result.g = isNaN.g ? x.g : result.g;
result.b = isNaN.b ? x.b : result.b;
result.a = isNaN.a ? x.a : result.a;
return result;
`,O9="return 1.0 / (1.0 + exp(-1.0 * x));",Js=class{constructor(e,t){this.variableNames=["A"],this.packedInputs=!0,this.packedOutput=!0,this.outputShape=e,this.enableShapeUniforms=Hn(this.outputShape.length),this.userCode=`
vec4 unaryOperation(vec4 x) {
${t}
}
void main() {
vec4 x = getAAtOutCoords();
vec4 y = unaryOperation(x);
setOutput(y);
}
`}},L9=class{constructor(e){this.variableNames=["A"],this.packedInputs=!0,this.packedOutput=!1,this.outputShape=e,this.enableShapeUniforms=Hn(this.outputShape.length);let t=e.length,n=Nn("rc",t),a=mt(t),r=w9(t,n),s=n.slice(-2),i=t<=1?"rc":`vec2(${s.join(",")})`;this.userCode=`
void main() {
${a} rc = getOutputCoords();
vec4 packedInput = getA(${r});
setOutput(getChannel(packedInput, ${i}));
}
`}},z9=gr.whereImpl,B9=1e-7,W9=1e-4,Eb={};function V9(e){return e in Eb||(Eb[e]={}),Eb[e]}var U9=X().getNumber("CPU_HANDOFF_SIZE_THRESHOLD"),G9=600;function H9(){return X().global.screen==null?1024:X().global.screen.height*X().global.screen.width*window.devicePixelRatio*G9/1024/1024}var sg=class extends gc{constructor(e){if(super(),this.pendingRead=new WeakMap,this.pendingDisposal=new WeakSet,this.dataRefCount=new WeakMap,this.numBytesInGPU=0,this.uploadWaitMs=0,this.downloadWaitMs=0,this.lastGlFlushTime=0,this.warnedAboutMemory=!1,this.pendingDeletes=0,this.disposed=!1,!X().getBool("HAS_WEBGL"))throw new Error("WebGL is not supported on this device");let t;if(e!=null){if(e instanceof zh)t=e;else{let n=Ya(X().getNumber("WEBGL_VERSION"),e);t=new zh(n)}this.binaryCache={},this.gpgpuCreatedLocally=!1}else{let n=Ya(X().getNumber("WEBGL_VERSION"));t=new zh(n),this.binaryCache=V9(X().getNumber("WEBGL_VERSION")),this.gpgpuCreatedLocally=!0}this.gpgpu=t,this.canvas=this.gpgpu.gl.canvas,this.textureManager=new S9(this.gpgpu),this.numMBBeforeWarning=H9(),this.texData=new xm(this,sr())}nextDataId(){return sg.nextDataId++}numDataIds(){return this.texData.numDataIds()-this.pendingDeletes}write(e,t,n){if((X().getBool("WEBGL_CHECK_NUMERICAL_PROBLEMS")||X().getBool("DEBUG"))&&this.checkNumericalProblems(e),n==="complex64"&&e!=null)throw new Error("Cannot write to a complex64 dtype. Please use tf.complex(real, imag).");let a={id:this.nextDataId()};return this.texData.set(a,{shape:t,dtype:n,values:e,usage:da.UPLOAD,refCount:1}),a}refCount(e){return this.texData.has(e)?this.texData.get(e).refCount:0}incRef(e){let t=this.texData.get(e);t.refCount++}decRef(e){if(this.texData.has(e)){let t=this.texData.get(e);t.refCount--}}move(e,t,n,a,r){if(X().getBool("DEBUG")&&this.checkNumericalProblems(t),a==="complex64")throw new Error("Cannot write to a complex64 dtype. Please use tf.complex(real, imag).");this.texData.set(e,{shape:n,dtype:a,values:t,usage:da.UPLOAD,refCount:r})}disposeIntermediateTensorInfo(e){this.disposeData(e.dataId)}readSync(e){let t=this.texData.get(e),{values:n,dtype:a,complexTensorInfos:r,slice:s,shape:i,isPacked:o}=t;if(s!=null){let d;o?d=new Js(i,pl):d=new _r(i,pl);let c=this.runWebGLProgram(d,[{dataId:e,shape:i,dtype:a}],a),h=this.readSync(c.dataId);return this.disposeIntermediateTensorInfo(c),h}if(n!=null)return this.convertAndCacheOnCPU(e);if(a==="string")return n;let l=this.activeTimers!=null,u;l&&(u=w.now());let p;if(a==="complex64"){let d=this.readSync(r.real.dataId),c=this.readSync(r.imag.dataId);p=C.mergeRealAndImagArrays(d,c)}else p=this.getValuesFromTexture(e);return l&&(this.downloadWaitMs+=w.now()-u),this.convertAndCacheOnCPU(e,p)}async read(e){if(this.pendingRead.has(e)){let h=this.pendingRead.get(e);return new Promise(m=>h.push(m))}let t=this.texData.get(e),{values:n,shape:a,slice:r,dtype:s,complexTensorInfos:i,isPacked:o}=t;if(r!=null){let h;o?h=new Js(a,pl):h=new _r(a,pl);let m=this.runWebGLProgram(h,[{dataId:e,shape:a,dtype:s}],s),f=this.read(m.dataId);return this.disposeIntermediateTensorInfo(m),f}if(n!=null)return this.convertAndCacheOnCPU(e);if(X().getBool("DEBUG")&&!X().getBool("WEBGL_DOWNLOAD_FLOAT_ENABLED")&&X().getNumber("WEBGL_VERSION")===2)throw new Error("tensor.data() with WEBGL_DOWNLOAD_FLOAT_ENABLED=false and WEBGL_VERSION=2 not yet supported.");let l=null,u;if(s!=="complex64"&&X().get("WEBGL_BUFFER_SUPPORTED")){u=this.decode(e);let h=this.texData.get(u.dataId);l=this.gpgpu.createBufferFromTexture(h.texture.texture,..._h(a))}this.pendingRead.set(e,[]),s!=="complex64"&&await this.gpgpu.createAndWaitForFence();let p;if(s==="complex64"){let h=await Promise.all([this.read(i.real.dataId),this.read(i.imag.dataId)]),m=h[0],f=h[1];p=C.mergeRealAndImagArrays(m,f)}else if(l==null)p=this.getValuesFromTexture(e);else{let h=w.sizeFromShape(a);p=this.gpgpu.downloadFloat32MatrixFromBuffer(l,h)}if(u!=null&&this.disposeIntermediateTensorInfo(u),l!=null){let h=this.gpgpu.gl;ge(h,()=>h.deleteBuffer(l))}let d=this.convertAndCacheOnCPU(e,p),c=this.pendingRead.get(e);return this.pendingRead.delete(e),c.forEach(h=>h(d)),this.pendingDisposal.has(e)&&(this.pendingDisposal.delete(e),this.disposeData(e)&&sr().removeDataId(e,this),this.pendingDeletes--),d}readToGPU(e,t={}){let n=this.texData.get(e),{values:a,shape:r,slice:s,dtype:i,isPacked:o,texture:l}=n;if(i==="complex64")throw new Error("Does not support reading texture for complex64 dtype.");if(s!=null){let c;o?c=new Js(r,pl):c=new _r(r,pl);let h=this.runWebGLProgram(c,[{dataId:e,shape:r,dtype:i}],i),m=this.readToGPU(h,t);return this.disposeIntermediateTensorInfo(h),m}if(l==null)throw a!=null?new Error("Data is not on GPU but on CPU."):new Error("There is no data on GPU or CPU.");let u=this.decode(e,t.customTexShape),p=sr().makeTensorFromTensorInfo(u),d=this.texData.get(u.dataId);return Object.assign({tensorRef:p},d.texture)}bufferSync(e){let t=this.readSync(e.dataId);if(e.dtype==="string")try{let n=t.map(a=>w.decodeString(a));return Ve(e.shape,e.dtype,n)}catch(n){throw new Error("Failed to decode encoded string bytes into utf-8")}return Ve(e.shape,e.dtype,t)}checkNumericalProblems(e){if(e!=null)for(let t=0;t<e.length;t++){let n=e[t];if(!$_(n))throw X().getBool("WEBGL_RENDER_FLOAT32_CAPABLE")?Error(`The value ${n} cannot be represented with your current settings. Consider enabling float32 rendering: 'tf.env().set('WEBGL_RENDER_FLOAT32_ENABLED', true);'`):Error(`The value ${n} cannot be represented on this device.`)}}getValuesFromTexture(e){let{shape:t,dtype:n,isPacked:a}=this.texData.get(e),r=w.sizeFromShape(t);if(X().getBool("WEBGL_DOWNLOAD_FLOAT_ENABLED")){let d=this.decode(e),c=this.texData.get(d.dataId),h=this.gpgpu.downloadMatrixFromPackedTexture(c.texture.texture,..._h(t)).subarray(0,r);return this.disposeIntermediateTensorInfo(d),h}let s=X().getBool("WEBGL_PACK")&&a===!0,i=s?Ph(t):t,o=s?new $Y(i):new AY(i),l=this.runWebGLProgram(o,[{shape:i,dtype:n,dataId:e}],"float32"),u=this.texData.get(l.dataId),p=this.gpgpu.downloadByteEncodedFloatMatrixFromOutputTexture(u.texture.texture,u.texShape[0],u.texShape[1]).subarray(0,r);return this.disposeIntermediateTensorInfo(l),p}timerAvailable(){return X().getNumber("WEBGL_DISJOINT_QUERY_TIMER_EXTENSION_RELIABLE")>0}time(e){let t=this.activeTimers,n=[],a=!1;this.programTimersStack==null?(this.programTimersStack=n,a=!0):this.activeTimers.push(n),this.activeTimers=n,e();let r=w.flatten(this.activeTimers.map(o=>o.query)).filter(o=>o!=null),s=w.flatten(this.activeTimers.map(o=>o.name)).filter(o=>o!=null);this.activeTimers=t,a&&(this.programTimersStack=null);let i={uploadWaitMs:this.uploadWaitMs,downloadWaitMs:this.downloadWaitMs,kernelMs:null,wallMs:null};return(async()=>{if(X().getNumber("WEBGL_DISJOINT_QUERY_TIMER_EXTENSION_RELIABLE")>0){let o=await Promise.all(r);i.kernelMs=w.sum(o),i.getExtraProfileInfo=()=>o.map((l,u)=>({name:s[u],ms:l})).map(l=>`${l.name}: ${l.ms}`).join(", ")}else i.kernelMs={error:"WebGL query timers are not supported in this environment."};return this.uploadWaitMs=0,this.downloadWaitMs=0,i})()}memory(){return{unreliable:!1,numBytesInGPU:this.numBytesInGPU,numBytesInGPUAllocated:this.textureManager.numBytesAllocated,numBytesInGPUFree:this.textureManager.numBytesFree}}startTimer(){return X().getNumber("WEBGL_DISJOINT_QUERY_TIMER_EXTENSION_RELIABLE")>0?this.gpgpu.beginQuery():{startMs:w.now(),endMs:null}}endTimer(e){return X().getNumber("WEBGL_DISJOINT_QUERY_TIMER_EXTENSION_RELIABLE")>0?(this.gpgpu.endQuery(),e):(e.endMs=w.now(),e)}async getQueryTime(e){if(X().getNumber("WEBGL_DISJOINT_QUERY_TIMER_EXTENSION_RELIABLE")>0)return this.gpgpu.waitForQueryAndGetTime(e);let t=e;return t.endMs-t.startMs}disposeData(e,t=!1){if(this.pendingDisposal.has(e))return!1;if(!this.texData.has(e))return!0;if(t?this.texData.get(e).refCount=0:this.texData.get(e).refCount--,!t&&this.texData.get(e).refCount>0)return!1;if(this.pendingRead.has(e))return this.pendingDisposal.add(e),this.pendingDeletes++,!1;this.releaseGPUData(e);let{complexTensorInfos:n}=this.texData.get(e);return n!=null&&(this.disposeData(n.real.dataId,t),this.disposeData(n.imag.dataId,t)),this.texData.delete(e),!0}releaseGPUData(e){let{texture:t,dtype:n,texShape:a,usage:r,isPacked:s,slice:i}=this.texData.get(e),o=i&&i.origDataId||e,l=this.dataRefCount.get(o);l>1?this.dataRefCount.set(o,l-1):(this.dataRefCount.delete(o),t!=null&&(this.numBytesInGPU-=this.computeBytes(a,n),this.textureManager.releaseTexture(t,a,r,s)));let u=this.texData.get(e);u.texture=null,u.texShape=null,u.isPacked=!1,u.slice=null}getTexture(e){return this.uploadToGPU(e),this.texData.get(e).texture.texture}getDataInfo(e){return this.texData.get(e)}shouldExecuteOnCPU(e,t=U9){return X().getBool("WEBGL_CPU_FORWARD")&&e.every(n=>this.texData.get(n.dataId).texture==null&&w.sizeFromShape(n.shape)<t)}getGPGPUContext(){return this.gpgpu}where(e){C.warn("tf.where() in webgl locks the UI thread. Call tf.whereAsync() instead");let t=e.dataSync();return z9(e.shape,t)}packedUnaryOp(e,t,n){let a=new Js(e.shape,t),r=this.compileAndRun(a,[e],n);return sr().makeTensorFromTensorInfo(r)}abs(e){if(this.shouldExecuteOnCPU([e])&&e.dtype!=="complex64"){let a=NC(this.texData.get(e.dataId).values);return this.makeOutput(e.shape,e.dtype,a)}if(X().getBool("WEBGL_PACK_UNARY_OPERATIONS"))return this.packedUnaryOp(e,Yk,e.dtype);let t=new _r(e.shape,Yk),n=this.compileAndRun(t,[e]);return sr().makeTensorFromTensorInfo(n)}makeTensorInfo(e,t,n){let a;if(t==="string"&&n!=null&&n.length>0&&w.isString(n[0])){let r=n.map(s=>w.encodeString(s));a=this.write(r,e,t)}else a=this.write(n,e,t);return this.texData.get(a).usage=null,{dataId:a,shape:e,dtype:t}}makeOutput(e,t,n){return sr().makeTensorFromTensorInfo(this.makeTensorInfo(e,t,n),this)}unpackTensor(e){let t=new L9(e.shape);return this.runWebGLProgram(t,[e],e.dtype)}packTensor(e){let t=new k9(e.shape),n=!0;return this.runWebGLProgram(t,[e],e.dtype,null,n)}packedReshape(e,t){let n=[gi(e.shape),...yi(e.shape)],a={dtype:e.dtype,shape:n,dataId:e.dataId},r=[gi(t),...yi(t)],s=new CC(r,n),i=!0,o=[n],l=this.runWebGLProgram(s,[a],e.dtype,o,i);return{dataId:l.dataId,shape:t,dtype:l.dtype}}decode(e,t){let n=this.texData.get(e),{isPacked:a,shape:r,dtype:s}=n;if(t!=null){let d=w.sizeFromShape(r),c=t[0]*t[1]*4;w.assert(d<=c,()=>"customTexShape is too small. Row * Column * 4 should be equal or larger than the size of the tensor data.")}let i=Ph(r),o;a?o=new FY(i):o=new EY(i);let l=!0,u=[t!=null?t:_h(i)],p=this.runWebGLProgram(o,[{shape:i,dtype:s,dataId:e}],s,u,l,t);return{dtype:s,shape:r,dataId:p.dataId}}runWebGLProgram(e,t,n,a,r=!1,s){let i=this.makeTensorInfo(e.outputShape,n),o=this.texData.get(i.dataId);if(e.packedOutput&&(o.isPacked=!0),e.outPackingScheme===cc.DENSE){let g=s!=null?s:_h(e.outputShape);o.texShape=g.map(y=>y*2)}if(e.outTexUsage!=null&&(o.usage=e.outTexUsage),w.sizeFromShape(i.shape)===0)return o.values=w.getTypedArrayFromDType(i.dtype,0),i;let l=[],u=t.map(g=>{if(g.dtype==="complex64")throw new Error("GPGPUProgram does not support complex64 input. For complex64 dtypes, please separate the program into real and imaginary parts.");let y=this.texData.get(g.dataId);if(y.texture==null){if(!e.packedInputs&&w.sizeFromShape(g.shape)<=X().getNumber("WEBGL_SIZE_UPLOAD_UNIFORM"))return{shape:g.shape,texData:null,isUniform:!0,uniformValues:y.values};e.packedInputs&&(y.isPacked=!0,y.shape=g.shape)}if(this.uploadToGPU(g.dataId),!!y.isPacked!=!!e.packedInputs)g=y.isPacked?this.unpackTensor(g):this.packTensor(g),l.push(g),y=this.texData.get(g.dataId);else if(y.isPacked&&!dc(y.shape,g.shape)){let b=g,x=g.shape;g.shape=y.shape,g=this.packedReshape(g,x),l.push(g),y=this.texData.get(g.dataId),b.shape=x}return{shape:g.shape,texData:y,isUniform:!1}});this.uploadToGPU(i.dataId);let p={shape:i.shape,texData:o,isUniform:!1},d=CY(e,u,p),c=this.getAndSaveBinary(d,()=>TY(this.gpgpu,e,u,p)),h=this.activeTimers!=null,m;h&&(m=this.startTimer()),X().get("ENGINE_COMPILE_ONLY")||_Y(this.gpgpu,c,u,p,a),l.forEach(g=>this.disposeIntermediateTensorInfo(g)),h&&(m=this.endTimer(m),this.activeTimers.push({name:e.constructor.name,query:this.getQueryTime(m)}));let f=X().get("WEBGL_FLUSH_THRESHOLD");if(f>0){let g=w.now();g-this.lastGlFlushTime>f&&(this.gpgpu.gl.flush(),this.lastGlFlushTime=g)}if(!X().getBool("WEBGL_LAZILY_UNPACK")&&o.isPacked&&r===!1){let g=this.unpackTensor(i);return this.disposeIntermediateTensorInfo(i),g}return i}compileAndRun(e,t,n,a,r=!1){return n=n||t[0].dtype,this.runWebGLProgram(e,t,n,a,r)}getAndSaveBinary(e,t){return e in this.binaryCache||(this.binaryCache[e]=t()),this.binaryCache[e]}getTextureManager(){return this.textureManager}dispose(){this.disposed||(X().getBool("IS_TEST")||Object.keys(this.binaryCache).forEach(e=>{this.gpgpu.deleteProgram(this.binaryCache[e].webGLProgram),delete this.binaryCache[e]}),this.textureManager.dispose(),this.canvas!=null&&typeof HTMLCanvasElement!="undefined"&&this.canvas instanceof HTMLCanvasElement?this.canvas.remove():this.canvas=null,this.gpgpuCreatedLocally&&(this.gpgpu.program=null,this.gpgpu.dispose()),this.disposed=!0)}floatPrecision(){return this.floatPrecisionValue==null&&(this.floatPrecisionValue=O(()=>{if(!X().get("WEBGL_RENDER_FLOAT32_ENABLED")){let e=X().getBool("DEBUG");X().set("DEBUG",!1);let t=this.abs(we(1e-8)).dataSync()[0];if(X().set("DEBUG",e),t>0)return 32}return 16})),this.floatPrecisionValue}epsilon(){return this.floatPrecision()===32?B9:W9}uploadToGPU(e){let t=this.texData.get(e),{shape:n,dtype:a,values:r,texture:s,usage:i,isPacked:o}=t;if(s!=null)return;let l=this.activeTimers!=null,u;l&&(u=w.now());let p=t.texShape;if(p==null&&(p=X_(n,o),t.texShape=p),r!=null){let d=Ph(n),c,h=p[1],m=p[0],f=r instanceof Uint8Array||r instanceof Uint8ClampedArray;(o||!f)&&([h,m]=Ju(p[0],p[1])),o?c=new RY(d,f):c=new DY(d,f);let g=f?[m,h]:p,y=this.makeTensorInfo(g,a),b=this.texData.get(y.dataId);f?b.usage=da.PIXELS:b.usage=da.UPLOAD,b.texShape=g,this.gpgpu.uploadDenseMatrixToTexture(this.getTexture(y.dataId),h,m,r);let x=[[m,h]],v=!0,k=this.runWebGLProgram(c,[y],a,x,v),T=this.texData.get(k.dataId);t.texShape=T.texShape,t.isPacked=T.isPacked,t.usage=T.usage,X().get("ENGINE_COMPILE_ONLY")?this.disposeData(k.dataId):(t.texture=T.texture,t.values=null,this.texData.delete(k.dataId)),this.disposeIntermediateTensorInfo(y),l&&(this.uploadWaitMs+=w.now()-u)}else{let d=this.acquireTexture(p,i,a,o);t.texture=d}}convertAndCacheOnCPU(e,t){let n=this.texData.get(e),{dtype:a}=n;return this.releaseGPUData(e),t!=null&&(n.values=j9(t,a)),n.values}acquireTexture(e,t,n,a){if(this.numBytesInGPU+=this.computeBytes(e,n),!this.warnedAboutMemory&&this.numBytesInGPU>this.numMBBeforeWarning*1024*1024){let r=(this.numBytesInGPU/1024/1024).toFixed(2);this.warnedAboutMemory=!0,console.warn(`High memory usage in GPU: ${r} MB, most likely due to a memory leak`)}return this.textureManager.acquireTexture(e,t,a)}computeBytes(e,t){return e[0]*e[1]*w.bytesPerElement(t)}checkCompileCompletion(){for(let[,e]of Object.entries(this.binaryCache))this.checkCompletion_(e)}async checkCompileCompletionAsync(){let e=[];if(this.gpgpu.parallelCompilationExtension){for(let[,t]of Object.entries(this.binaryCache))e.push(this.checkCompletionAsync_(t));return Promise.all(e)}else{for(let[,t]of Object.entries(this.binaryCache)){let n=new Promise(a=>{try{this.checkCompletion_(t),a(!0)}catch(r){throw r}});e.push(n)}return Promise.all(e)}}async checkCompletionAsync_(e){return this.gpgpu.gl.getProgramParameter(e.webGLProgram,this.gpgpu.parallelCompilationExtension.COMPLETION_STATUS_KHR)?this.checkCompletion_(e):(await Bv(),this.checkCompletionAsync_(e))}checkCompletion_(e){if(this.gpgpu.gl.getProgramParameter(e.webGLProgram,this.gpgpu.gl.LINK_STATUS)===!1)throw console.log(this.gpgpu.gl.getProgramInfoLog(e.webGLProgram)),this.gpgpu.gl.getShaderParameter(e.fragmentShader,this.gpgpu.gl.COMPILE_STATUS)===!1?($0(e.source,this.gpgpu.gl.getShaderInfoLog(e.fragmentShader)),new Error("Failed to compile fragment shader.")):new Error("Failed to link vertex and fragment shaders.");return!0}getUniformLocations(){for(let[,e]of Object.entries(this.binaryCache)){let{uniformLocations:t,customUniformLocations:n,infLoc:a,nanLoc:r,inShapesLocations:s,inTexShapesLocations:i,outShapeLocation:o,outShapeStridesLocation:l,outTexShapeLocation:u}=iC(this.gpgpu,e.program,e.webGLProgram);e.uniformLocations=t,e.customUniformLocations=n,e.infLoc=a,e.nanLoc=r,e.inShapesLocations=s,e.inTexShapesLocations=i,e.outShapeLocation=o,e.outShapeStridesLocation=l,e.outTexShapeLocation=u}}};sg.nextDataId=0;function j9(e,t){if(t==="float32"||t==="complex64")return e;if(t==="int32"||t==="bool"){let n=t==="int32"?new Int32Array(e.length):new Uint8Array(e.length);for(let a=0;a<n.length;++a)n[a]=Math.round(e[a]);return n}else throw new Error(`Unknown dtype ${t}`)}var q9="3.18.0";function EC(){X().set("WEBGL_FORCE_F16_TEXTURES",!0)}Oc.isBrowser()&&nf("webgl",()=>new sg,2);var K9={forceHalfFloat:EC},FC=`
if (isnan(a)) return a;
if (isnan(b)) return b;
`,Pl=class{constructor(e,t,n){this.variableNames=["A","B"],this.outputShape=C.assertAndGetBroadcastShape(t,n),this.enableShapeUniforms=Hn(this.outputShape.length),this.userCode=`
float binaryOperation(float a, float b) {
${e}
}
void main() {
float a = getAAtOutCoords();
float b = getBAtOutCoords();
setOutput(binaryOperation(a, b));
}
`}},ig=`
result.r = isNaN.r > 0. ? NAN : result.r;
result.g = isNaN.g > 0. ? NAN : result.g;
result.b = isNaN.b > 0. ? NAN : result.b;
result.a = isNaN.a > 0. ? NAN : result.a;
`,dd=class{constructor(e,t,n,a=!1){this.variableNames=["A","B"],this.supportsBroadcasting=!0,this.packedInputs=!0,this.packedOutput=!0,this.outputShape=C.assertAndGetBroadcastShape(t,n);let r=this.outputShape.length;this.enableShapeUniforms=Hn(r);let s="";if(a)if(r===0||w.sizeFromShape(this.outputShape)===1)s=`
result.y = 0.;
result.z = 0.;
result.w = 0.;
`;else if(s=`
${mt(r)} coords = getOutputCoords();
`,r===1)this.enableShapeUniforms?s+=`
result.y = (coords + 1) >= outShape ? 0. : result.y;
result.z = 0.;
result.w = 0.;
`:s+=`
result.y = (coords + 1) >= ${this.outputShape[0]} ? 0. : result.y;
result.z = 0.;
result.w = 0.;
`;else{let i=Nn("coords",r);this.enableShapeUniforms?s+=`
bool nextRowOutOfBounds =
(${i[r-2]} + 1) >= outShape[${r} - 2];
bool nextColOutOfBounds =
(${i[r-1]} + 1) >= outShape[${r} - 1];
result.y = nextColOutOfBounds ? 0. : result.y;
result.z = nextRowOutOfBounds ? 0. : result.z;
result.w = nextColOutOfBounds || nextRowOutOfBounds ? 0. : result.w;
`:s+=`
bool nextRowOutOfBounds =
(${i[r-2]} + 1) >= ${this.outputShape[r-2]};
bool nextColOutOfBounds =
(${i[r-1]} + 1) >= ${this.outputShape[r-1]};
result.y = nextColOutOfBounds ? 0. : result.y;
result.z = nextRowOutOfBounds ? 0. : result.z;
result.w = nextColOutOfBounds || nextRowOutOfBounds ? 0. : result.w;
`}this.userCode=`
vec4 binaryOperation(vec4 a, vec4 b) {
${e}
}
void main() {
vec4 a = getAAtOutCoords();
vec4 b = getBAtOutCoords();
vec4 result = binaryOperation(a, b);
${s}
setOutput(result);
}
`}};function aa(e){let{inputs:t,backend:n}=e,{x:a}=t;return n.incRef(a.dataId),{dataId:a.dataId,shape:a.shape,dtype:a.dtype}}var X9={kernelName:Bi,backendName:"webgl",kernelFunc:aa};function Cs(e){let{inputs:t,backend:n}=e,{real:a,imag:r}=t,s=n.makeTensorInfo(a.shape,"complex64"),i=n.texData.get(s.dataId),o=aa({inputs:{x:a},backend:n}),l=aa({inputs:{x:r},backend:n});return i.complexTensorInfos={real:o,imag:l},s}var Y9={kernelName:Tm,backendName:"webgl",kernelFunc:Cs},AC="return (a < 0.) ? b * a : a;",$C=`
vec4 aLessThanZero = vec4(lessThan(a, vec4(0.)));
return (aLessThanZero * (b * a)) + ((vec4(1.0) - aLessThanZero) * a);
`;function Q9(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{alpha:s}=a,i=n.makeTensorInfo([],"float32",w.createScalarValue(s,"float32")),o=X().getBool("WEBGL_PACK_BINARY_OPERATIONS")?new dd($C,r.shape,i.shape):new Pl(AC,r.shape,i.shape),l=n.runWebGLProgram(o,[r,i],"float32");return n.disposeIntermediateTensorInfo(i),l}var J9={kernelName:Wi,backendName:"webgl",kernelFunc:Q9},DC="return (a < 0.) ? b * a : a;",RC=`
vec4 aLessThanZero = vec4(lessThan(a, vec4(0.)));
return (aLessThanZero * (b * a)) + ((vec4(1.0) - aLessThanZero) * a);
`;function Z9(e){let{inputs:t,backend:n}=e,{x:a,alpha:r}=t,s=X().getBool("WEBGL_PACK_BINARY_OPERATIONS")?new dd(RC,a.shape,r.shape):new Pl(DC,a.shape,r.shape);return n.runWebGLProgram(s,[a,r],"float32")}var eQ={kernelName:eo,backendName:"webgl",kernelFunc:Z9},rp="if (isnan(x)) return x;",tQ=`
if (isnan(a)) return a;
if (isnan(b)) return b;
`,nQ=`
result.r = isNaN.r > 0. ? NAN : result.r;
result.g = isNaN.g > 0. ? NAN : result.g;
result.b = isNaN.b > 0. ? NAN : result.b;
result.a = isNaN.a > 0. ? NAN : result.a;
`;function Qe({opSnippet:e,packedOpSnippet:t,cpuKernelImpl:n,dtype:a}){return({inputs:r,backend:s})=>{let{x:i}=r,o=s,l=a||i.dtype;if(o.shouldExecuteOnCPU([i])&&n!=null){let d=o.texData.get(i.dataId),c=n(d.values,l);return o.makeTensorInfo(i.shape,l,c)}let u=X().getBool("WEBGL_PACK_UNARY_OPERATIONS")&&t!=null,p;return u?p=new Js(i.shape,t):p=new _r(i.shape,e),o.runWebGLProgram(p,[i],l)}}function cn({opSnippet:e,packedOpSnippet:t,checkOutOfBounds:n=!1,supportsComplex:a=!1,cpuKernelImpl:r,dtype:s}){return({inputs:i,backend:o})=>{let{a:l,b:u}=i,p=o;if(a&&l.dtype==="complex64"){let m=p.texData.get(l.dataId),f=p.texData.get(u.dataId),[g,y]=[[m.complexTensorInfos.real,f.complexTensorInfos.real],[m.complexTensorInfos.imag,f.complexTensorInfos.imag]].map(x=>{let[v,k]=x,T={dataId:v.dataId,dtype:v.dtype,shape:l.shape},_={dataId:k.dataId,dtype:k.dtype,shape:u.shape},E=new Pl(e,l.shape,u.shape);return p.runWebGLProgram(E,[T,_],fa(v.dtype,k.dtype))}),b=Cs({inputs:{real:g,imag:y},backend:p});return p.disposeIntermediateTensorInfo(g),p.disposeIntermediateTensorInfo(y),b}let d=s||fa(l.dtype,u.dtype);if((l.dtype==="string"||u.dtype==="string"||p.shouldExecuteOnCPU([l,u]))&&r!=null){let m=p.texData.get(l.dataId).values,f=p.texData.get(u.dataId).values,g=l.dtype==="string"?C.fromUint8ToStringArray(m):m,y=l.dtype==="string"?C.fromUint8ToStringArray(f):f,[b,x]=r(l.shape,u.shape,g,y,d),v=p.makeTensorInfo(x,d),k=p.texData.get(v.dataId);return k.values=b,v}let c=X().getBool("WEBGL_PACK_BINARY_OPERATIONS")&&t!=null,h;return c?h=new dd(t,l.shape,u.shape,n):h=new Pl(e,l.shape,u.shape),p.runWebGLProgram(h,[l,u],d)}}function og(e,t=!1){if(e==="linear")return t?D9:C9;if(e==="relu")return t?M9:F9;if(e==="elu")return t?R9:E9;if(e==="relu6")return t?P9:A9;if(e==="prelu")return t?RC:DC;if(e==="leakyrelu")return t?$C:AC;if(e==="sigmoid")return t?O9:$9;throw new Error(`Activation ${e} has not been implemented for the WebGL backend.`)}var MC=class{constructor(e,t,n,a=!1,r=!1,s=!1,i=null,o=!1,l=!1){this.variableNames=["matrixA","matrixB"],this.packedInputs=!0,this.packedOutput=!0,this.outputShape=n,this.enableShapeUniforms=Hn(this.outputShape.length);let u=a?e[1]:e[2],p=Math.ceil(u/2),d=a?"i * 2, rc.y":"rc.y, i * 2",c=r?"rc.z, i * 2":"i * 2, rc.z",h=a?["a.xxyy","a.zzww"]:["a.xxzz","a.yyww"],m=r?["b.xzxz","b.ywyw"]:["b.xyxy","b.zwzw"],f="",g="";i&&(o?f=`vec4 activation(vec4 a) {
vec4 b = getPreluActivationWeightsAtOutCoords();
${i}
}`:l?f=`vec4 activation(vec4 a) {
vec4 b = getLeakyreluAlphaAtOutCoords();
${i}
}`:f=`vec4 activation(vec4 x) {
${i}
}`,g="result = activation(result);");let y=s?"result += getBiasAtOutCoords();":"";s&&this.variableNames.push("bias"),o&&this.variableNames.push("preluActivationWeights"),l&&this.variableNames.push("leakyreluAlpha");let b="rc.x",x="rc.x";e[0]<t[0]?b=`int(min(float(rc.x), ${e[0]-1}.))`:t[0]<e[0]&&(x=`int(min(float(rc.x), ${t[0]-1}.))`),this.userCode=`
${f}
// Don't use uniform for sharedDimensionPacked for performance.
const float sharedDimension = ${p}.0;
vec4 dot2x2ARowBCol(ivec3 rc) {
vec4 result = vec4(0);
for (int i = 0; i < ${p}; i++) {
int batchA = ${b};
int batchB = ${x};
vec4 a = getMatrixA(batchA, ${d});
vec4 b = getMatrixB(batchB, ${c});
// These swizzled products need to be separately added.
// See: https://github.com/tensorflow/tfjs/issues/1735
result += (${h[0]} * ${m[0]});
result += (${h[1]} * ${m[1]});
}
return result;
}
void main() {
ivec3 rc = getOutputCoords();
vec4 result = dot2x2ARowBCol(rc);
${y}
${g}
setOutput(result);
}
`}},Qk={REAL:"return areal * breal - aimag * bimag;",IMAG:"return areal * bimag + aimag * breal;"},Jk=class{constructor(e,t,n){this.variableNames=["AReal","AImag","BReal","BImag"],this.outputShape=C.assertAndGetBroadcastShape(t,n),this.userCode=`
float binaryOpComplex(
float areal, float aimag, float breal, float bimag) {
${e}
}
void main() {
float areal = getARealAtOutCoords();
float aimag = getAImagAtOutCoords();
float breal = getBRealAtOutCoords();
float bimag = getBImagAtOutCoords();
setOutput(binaryOpComplex(areal, aimag, breal, bimag));
}
`}},Zk="return a * b;";function V0(e){let{inputs:t,backend:n}=e,{a,b:r}=t,s=C.upcastType(a.dtype,r.dtype);if(a.dtype==="complex64"){let o=n.texData.get(a.dataId),l=n.texData.get(r.dataId),u=new Jk(Qk.REAL,a.shape,r.shape),p=new Jk(Qk.IMAG,a.shape,r.shape),d=[{dataId:o.complexTensorInfos.real.dataId,dtype:o.complexTensorInfos.real.dtype,shape:a.shape},{dataId:o.complexTensorInfos.imag.dataId,dtype:o.complexTensorInfos.imag.dtype,shape:a.shape},{dataId:l.complexTensorInfos.real.dataId,dtype:l.complexTensorInfos.real.dtype,shape:r.shape},{dataId:l.complexTensorInfos.imag.dataId,dtype:l.complexTensorInfos.imag.dtype,shape:r.shape}],c=n.runWebGLProgram(u,d,"float32"),h=n.runWebGLProgram(p,d,"float32"),m=Cs({inputs:{real:c,imag:h},backend:n});return n.disposeIntermediateTensorInfo(c),n.disposeIntermediateTensorInfo(h),m}if(n.shouldExecuteOnCPU([a,r])){let o=n.texData.get(a.dataId),l=n.texData.get(r.dataId),[u,p]=t9(a.shape,r.shape,o.values,l.values,s),d=n.makeTensorInfo(p,s),c=n.texData.get(d.dataId);return c.values=u,d}let i;return X().getBool("WEBGL_PACK_BINARY_OPERATIONS")?i=new dd(Zk,a.shape,r.shape):i=new Pl(Zk,a.shape,r.shape),n.runWebGLProgram(i,[a,r],s)}var aQ={kernelName:Yi,backendName:"webgl",kernelFunc:V0};function rQ(e,t,n){let a=[gi(e.shape),...yi(e.shape)],r={dtype:e.dtype,shape:a,dataId:e.dataId},s=[gi(t),...yi(t)],i=new CC(s,a),o=!0,l=[a],u=n.runWebGLProgram(i,[r],e.dtype,l,o);return{dataId:u.dataId,shape:t,dtype:u.dtype}}function me(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{shape:s}=a,i=n,o=w.sizeFromShape(r.shape),l=w.inferFromImplicitShape(s,o),u=w.sizeFromShape(l);w.assert(o===u,()=>`The new shape (${l}) has ${u} elements and the old shape (${r.shape}) has ${o} elements. The new shape and old shape must have the same number of elements.`);let p=i.texData.get(r.dataId);return p.isPacked&&!dc(r.shape,l)&&!(p.texture!==null&&dc(p.shape,l))?rQ(r,l,i):(i.incRef(r.dataId),{dataId:r.dataId,shape:l,dtype:r.dtype})}var sQ={kernelName:Iu,backendName:"webgl",kernelFunc:me},eI=class{constructor(e,t){this.variableNames=["x"];let{windowSize:n,batchSize:a,inSize:r,outSize:s}=e;this.outputShape=[a,s];let i=Math.floor(n/4)*4,o=n%4,l="sumValue += dot(values, ones);";if(t!=null){let p=1/t;l=`sumValue += dot(values * ${w.isInt(p)?p.toPrecision(2):p}, ones);`}let u="";r%n>0&&(u=`
if (inIdx < 0 || inIdx >= ${r}) {
return 0.0;
}
`),this.userCode=`
const vec4 ones = vec4(1.0, 1.0, 1.0, 1.0);
float getValue(int batch, int inIdx) {
${u}
return getX(batch, inIdx);
}
void main() {
ivec2 coords = getOutputCoords();
int batch = coords[0];
int outIdx = coords[1];
int inOffset = outIdx * ${n};
float sumValue = 0.0;
for (int i = 0; i < ${i}; i += 4) {
int inIdx = inOffset + i;
vec4 values = vec4(
getValue(batch, inIdx),
getValue(batch, inIdx + 1),
getValue(batch, inIdx + 2),
getValue(batch, inIdx + 3)
);
${l}
}
int inIdx = inOffset + ${i};
if (${o===1}) {
vec4 values = vec4(getValue(batch, inIdx), 0.0, 0.0, 0.0);
${l}
} else if (${o===2}) {
vec4 values = vec4(
getValue(batch, inIdx),
getValue(batch, inIdx + 1), 0.0, 0.0);
${l}
} else if (${o===3}) {
vec4 values = vec4(
getValue(batch, inIdx),
getValue(batch, inIdx + 1),
getValue(batch, inIdx + 2), 0.0);
${l}
}
setOutput(sumValue);
}
`}},iQ=class{constructor(e,t){this.variableNames=["x"];let{windowSize:n,batchSize:a,inSize:r,outSize:s}=e;this.outputShape=[a,s];let i="0.0",o="";t==="prod"?i="1.0":t==="min"?(i="1.0 / 1e-20",o="min"):t==="max"&&(i="-1.0 / 1e-20",o="max");let l=`${t}(${t}(${t}(minMaxValue[0], minMaxValue[1]), minMaxValue[2]), minMaxValue[3])`;t==="sum"?l="sumValue":t==="prod"?l="prodValue":t==="all"?l="allValue":t==="any"&&(l="anyValue");let u=Math.floor(n/4)*4,p=n%4,d=`
if (${t==="sum"}) {
sumValue += dot(values, ones);
} else if (${t==="prod"}) {
vec2 tmp = vec2(values[0], values[1]) * vec2(values[2], values[3]);
prodValue *= tmp[0] * tmp[1];
} else {
minMaxValue = ${o}(values, minMaxValue);
if (${t==="min"} || ${t==="max"}) {
minMaxValue = ${o}(values, minMaxValue);
bvec4 isNaN = isnan(values);
if (isNaN.r || isNaN.g || isNaN.b || isNaN.a) {
minMaxValue = vec4(NAN);
}
}
}
`,c="vec4";t==="all"?(i="1.0",d=`
bool reducedAllValue = all(values);
float floatedReducedAllValue = float(reducedAllValue);
allValue = float(allValue >= 1.0 && floatedReducedAllValue >= 1.0);
`,c="bvec4"):t==="any"&&(i="0.0",d=`
bool reducedAnyValue = any(values);
float floatedReducedAnyValue = float(reducedAnyValue);
anyValue = float(anyValue >= 1.0 || floatedReducedAnyValue >= 1.0);
`,c="bvec4");let h="";r%n>0&&(h=`
if (inIdx < 0 || inIdx >= ${r}) {
return initializationValue;
}
`),this.userCode=`
const float initializationValue = ${i};
const vec4 ones = vec4(1.0, 1.0, 1.0, 1.0);
float getValue(int batch, int inIdx) {
${h}
return getX(batch, inIdx);
}
void main() {
ivec2 coords = getOutputCoords();
int batch = coords[0];
int outIdx = coords[1];
int inOffset = outIdx * ${n};
vec4 minMaxValue = vec4(${i});
float prodValue = 1.0;
float sumValue = 0.0;
float allValue = 1.0;
float anyValue = 0.0;
for (int i = 0; i < ${u}; i += 4) {
int inIdx = inOffset + i;
${c} values = ${c}(
getValue(batch, inIdx),
getValue(batch, inIdx + 1),
getValue(batch, inIdx + 2),
getValue(batch, inIdx + 3)
);
${d}
}
int inIdx = inOffset + ${u};
if (${p===1}) {
${c} values = ${c}(
getValue(batch, inIdx),
initializationValue,
initializationValue,
initializationValue
);
${d}
} else if (${p===2}) {
${c} values = ${c}(
getValue(batch, inIdx),
getValue(batch, inIdx + 1),
initializationValue,
initializationValue
);
${d}
} else if (${p===3}) {
${c} values = ${c}(
getValue(batch, inIdx),
getValue(batch, inIdx + 1),
getValue(batch, inIdx + 2),
initializationValue
);
${d}
}
setOutput(${l});
}
`}};function oQ(e){let t=[];for(;t.length===0||t[t.length-1].outSize!==1;){let n=t.length?t[t.length-1].outSize:e[1],a=C.computeOptimalWindowSize(n);t.push({inSize:n,windowSize:a,outSize:Math.ceil(n/a)})}return t}function To(e,t,n,a){let r=oQ(e.shape),s=e;for(let i=0;i<r.length;i++){let{inSize:o,windowSize:l,outSize:u}=r[i],p,d;n==="mean"?p=i===0?new eI({windowSize:l,inSize:o,batchSize:e.shape[0],outSize:u},o):new eI({windowSize:l,inSize:o,batchSize:e.shape[0],outSize:u}):p=new iQ({windowSize:l,inSize:o,batchSize:e.shape[0],outSize:u},n),d=s,s=a.runWebGLProgram(p,[s],t),d.dataId!==e.dataId&&a.disposeIntermediateTensorInfo(d)}return s}var lQ=class{constructor(e,t){this.variableNames=["A"];let n=new Array(e.length);for(let s=0;s<n.length;s++)n[s]=e[t[s]];this.outputShape=n,this.rank=n.length;let a=mt(this.rank),r=uQ(t);this.userCode=`
void main() {
${a} resRC = getOutputCoords();
setOutput(getA(${r}));
}
`}};function uQ(e){let t=e.length;if(t>6)throw Error(`Transpose for rank ${t} is not yet supported`);let n=["resRC.x","resRC.y","resRC.z","resRC.w","resRC.u","resRC.v"],a=new Array(t);for(let r=0;r<e.length;r++)a[e[r]]=n[r];return a.join()}var pQ=class{constructor(e,t){this.variableNames=["A"],this.packedInputs=!0,this.packedOutput=!0;let n=new Array(e.length);for(let u=0;u<n.length;u++)n[u]=e[t[u]];if(this.outputShape=n,this.rank=n.length,this.rank>6)throw Error(`Packed transpose for rank ${this.rank} is not yet supported.`);let a=mt(this.rank),r=_C("rc",this.rank),s=new Array(this.rank);for(let u=0;u<t.length;u++)s[t[u]]=r[u];let i=`vec2(${s.slice(-2).join()})`,o=`++${r[this.rank-1]} < ${n[this.rank-1]}`,l=`getChannel(getA(${s.join()}), ${i})`;this.userCode=`
void main() {
${a} rc = getOutputCoords();
vec4 result = vec4(0.);
result[0] = ${l};
if(${o}) {
result[1] = ${l};
}
--${r[this.rank-1]};
if(++${r[this.rank-2]} < ${n[this.rank-2]}) {
result[2] = ${l};
if(${o}) {
result[3] = ${l};
}
}
setOutput(result);
}
`}};function lg(e,t,n){let a=X().getBool("WEBGL_PACK_ARRAY_OPERATIONS")?new pQ(e.shape,t):new lQ(e.shape,t);return n.runWebGLProgram(a,[e],e.dtype)}function cQ(e,t,n,a){let r=t,s=e.shape.length,i=w.parseAxisParam(r,e.shape),o=i,l=C.getAxesPermutation(o,s),u=l!=null,p=e;u&&(p=lg(e,l,a),o=C.getInnerMostAxes(o.length,s)),C.assertAxesAreInnerMostDims("sum",o,s);let[d,c]=C.computeOutAndReduceShapes(p.shape,o),h=d;n&&(h=C.expandShapeToKeepDim(d,i));let m=w.sizeFromShape(c),f=w.sizeFromShape(e.shape)/m,g=me({inputs:{x:p},attrs:{shape:[f,m]},backend:a}),y=tf(e.dtype),b=To(g,y,"sum",a),x=me({inputs:{x:b},attrs:{shape:h},backend:a});return a.disposeIntermediateTensorInfo(g),a.disposeIntermediateTensorInfo(b),u&&a.disposeIntermediateTensorInfo(p),x}function ug(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{axis:s,keepDims:i}=a;return cQ(r,s,i,n)}var dQ={kernelName:co,backendName:"webgl",kernelFunc:ug};function Wt(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{perm:s}=a,i=n,o=r.shape.length,l=new Array(o);for(let p=0;p<l.length;p++)l[p]=r.shape[s[p]];let u;if(i.shouldExecuteOnCPU([r])){let p=i.texData.get(r.dataId).values,d=W0(p,r.shape,r.dtype,s,l);u=i.makeTensorInfo(l,r.dtype);let c=i.texData.get(u.dataId);c.values=d}else u=lg(r,s,i);return u}var hQ={kernelName:Cr,backendName:"webgl",kernelFunc:Wt},PC=1e3;function gm({a:e,b:t,transposeA:n,transposeB:a,backend:r,bias:s=null,preluActivationWeights:i=null,leakyreluAlpha:o=0,activation:l=null}){let u=e.shape.length,p=t.shape.length,d=n?e.shape[u-2]:e.shape[u-1],c=a?t.shape[p-1]:t.shape[p-2],h=n?e.shape[u-1]:e.shape[u-2],m=a?t.shape[p-2]:t.shape[p-1],f=e.shape.slice(0,-2),g=t.shape.slice(0,-2),y=w.sizeFromShape(f),b=w.sizeFromShape(g),x=Bu.assertAndGetBroadcastShape(e.shape.slice(0,-2),t.shape.slice(0,-2)).concat([h,m]);w.assert(d===c,()=>`Error in matMul: inner shapes (${d}) and (${c}) of Tensors with shapes ${e.shape} and ${t.shape} and transposeA=${n} and transposeB=${a} must match.`);let v=n?[y,d,h]:[y,h,d],k=a?[b,m,c]:[b,c,m],T=me({inputs:{x:e},backend:r,attrs:{shape:v}}),_=me({inputs:{x:t},backend:r,attrs:{shape:k}}),E=[T,_],A=Math.max(y,b),M=n?T.shape[1]:T.shape[2],$=s!=null,S=i!=null,P=l==="leakyrelu",V=l!=null?og(l,!0):null,j=$||S||P||V!=null,q;if((h===1||m===1)&&M>PC&&j===!1){let Z=T,ee=_;n&&(Z=Wt({inputs:{x:T},backend:r,attrs:{perm:[0,2,1]}}),E.push(Z)),a&&(ee=Wt({inputs:{x:_},backend:r,attrs:{perm:[0,2,1]}}),E.push(ee));let re=m!==1,Y=m===1,ie=Z;re&&(ie=me({inputs:{x:Z},backend:r,attrs:{shape:[A,M,1]}}),E.push(ie));let ae=m===1?2:1,le=ee;Y&&(le=me({inputs:{x:ee},backend:r,attrs:{shape:[A,1,M]}}),E.push(le));let ue=V0({inputs:{a:ie,b:le},backend:r});q=ug({inputs:{x:ue},backend:r,attrs:{axis:ae,keepDims:!0}}),E.push(ue)}else{let Z=fa(e.dtype,t.dtype),ee=new MC(v,k,[A,h,m],n,a,$,V,S,P),re=[T,_];if(s!=null&&re.push(s),S&&re.push(i),P){let Y=r.makeTensorInfo([],"float32",w.createScalarValue(o,"float32"));re.push(Y),E.push(Y)}q=r.runWebGLProgram(ee,re,Z)}let K=me({inputs:{x:q},backend:r,attrs:{shape:x}});E.push(q);for(let Z of E)r.disposeIntermediateTensorInfo(Z);return K}function mQ(e){let{inputs:t,backend:n,attrs:a}=e,{a:r,b:s,bias:i,preluActivationWeights:o}=t,{transposeA:l,transposeB:u,activation:p,leakyreluAlpha:d}=a;return gm({a:r,b:s,transposeA:l,transposeB:u,backend:n,bias:i,preluActivationWeights:o,leakyreluAlpha:d,activation:p})}var fQ={kernelName:ai,backendName:"webgl",kernelFunc:mQ},tI="return abs(x);";function gQ(e){let{inputs:t,backend:n}=e,{x:a}=t;if(n.shouldExecuteOnCPU([a])&&a.dtype!=="complex64"){let s=n.texData.get(a.dataId),i=NC(s.values);return n.makeTensorInfo(a.shape,a.dtype,i)}let r;return X().getBool("WEBGL_PACK_UNARY_OPERATIONS")?r=new Js(a.shape,tI):r=new _r(a.shape,tI),n.runWebGLProgram(r,[a],a.dtype)}var yQ={kernelName:Ll,backendName:"webgl",kernelFunc:gQ},bQ=Aa+`
if (abs(x) > 1.) {
return NAN;
}
return acos(x);
`,xQ=Qe({opSnippet:bQ}),vQ={kernelName:zl,backendName:"webgl",kernelFunc:xQ},wQ=Aa+`
if (x < 1.0) return NAN;
return log(x + sqrt(x * x - 1.0));`,kQ=Qe({opSnippet:wQ}),IQ={kernelName:Bl,backendName:"webgl",kernelFunc:kQ},nI="return a + b;",SQ=cn({opSnippet:nI,packedOpSnippet:nI,supportsComplex:!0,cpuKernelImpl:PY}),NQ={kernelName:bs,backendName:"webgl",kernelFunc:SQ},TQ=class{constructor(e,t){this.outputShape=[],this.outputShape=e,this.variableNames=t.map((r,s)=>`T${s}`);let n=[];this.variableNames.forEach(r=>{n.push(`float v${r} = get${r}AtOutCoords();`)});let a=this.variableNames.map(r=>`v${r}`).join(" + ");this.userCode=`
void main() {
${n.join(`
`)}
float result = ${a};
setOutput(result);
}
`}},_Q=class{constructor(e,t){this.outputShape=[],this.packedInputs=!0,this.packedOutput=!0,this.outputShape=e,this.variableNames=t.map((r,s)=>`T${s}`);let n=[];this.variableNames.forEach(r=>{n.push(`vec4 v${r} = get${r}AtOutCoords();`)});let a=this.variableNames.map(r=>`v${r}`).join(" + ");this.userCode=`
void main() {
${n.join(`
`)}
vec4 result = ${a};
setOutput(result);
}
`}};function Bh(e){let{inputs:t,backend:n}=e,a=t;if(a.length===1)return aa({inputs:{x:a[0]},backend:n});if(a.length>X().get("WEBGL_MAX_TEXTURES_IN_SHADER")){let o=Math.floor(a.length/2),l=Bh({inputs:a.slice(0,o),backend:n}),u=Bh({inputs:a.slice(o),backend:n});return Bh({inputs:[l,u],backend:n})}let r=a.map(o=>o.dtype).reduce((o,l)=>fa(o,l)),s=a.map(o=>o.shape),i=X().getBool("WEBGL_PACK")?new _Q(a[0].shape,s):new TQ(a[0].shape,s);return n.runWebGLProgram(i,a,r)}var CQ={kernelName:wi,backendName:"webgl",kernelFunc:Bh};function EQ(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{axis:s,keepDims:i}=a,o=r.shape.length,l=w.parseAxisParam(s,r.shape),u=l,p=C.getAxesPermutation(u,o),d=r;p!=null&&(d=Wt({inputs:{x:r},backend:n,attrs:{perm:p}}),u=C.getInnerMostAxes(u.length,o)),C.assertAxesAreInnerMostDims("all",u,o);let[c,h]=C.computeOutAndReduceShapes(d.shape,u),m=w.sizeFromShape(h),f=me({inputs:{x:d},backend:n,attrs:{shape:[-1,m]}}),g=To(f,f.dtype,"all",n),y;if(i){let b=C.expandShapeToKeepDim(c,l);y=me({inputs:{x:g},backend:n,attrs:{shape:b}})}else y=me({inputs:{x:g},backend:n,attrs:{shape:c}});return n.disposeIntermediateTensorInfo(f),n.disposeIntermediateTensorInfo(g),p!=null&&n.disposeIntermediateTensorInfo(d),y}var FQ={kernelName:Wl,backendName:"webgl",kernelFunc:EQ};function AQ(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{axis:s,keepDims:i}=a,o=r.shape.length,l=w.parseAxisParam(s,r.shape),u=l,p=C.getAxesPermutation(u,o),d=r;p!=null&&(d=Wt({inputs:{x:r},backend:n,attrs:{perm:p}}),u=C.getInnerMostAxes(u.length,o)),C.assertAxesAreInnerMostDims("any",u,o);let[c,h]=C.computeOutAndReduceShapes(d.shape,u),m=w.sizeFromShape(h),f=me({inputs:{x:d},backend:n,attrs:{shape:[-1,m]}}),g=To(f,f.dtype,"any",n),y;if(i){let b=C.expandShapeToKeepDim(c,l);y=me({inputs:{x:g},backend:n,attrs:{shape:b}})}else y=me({inputs:{x:g},backend:n,attrs:{shape:c}});return n.disposeIntermediateTensorInfo(f),n.disposeIntermediateTensorInfo(g),p!=null&&n.disposeIntermediateTensorInfo(d),y}var $Q={kernelName:Vl,backendName:"webgl",kernelFunc:AQ},DQ=class{constructor(e,t,n){this.variableNames=["A"];let{windowSize:a,batchSize:r,outSize:s}=e;n||this.variableNames.push("bestIndicesA"),this.outputShape=[r,s];let i=t==="max"?">":"<",o=n?"inOffset + i;":"round(getBestIndicesA(batch, inOffset + i));";this.userCode=`
void main() {
ivec2 coords = getOutputCoords();
int batch = coords[0];
int outIdx = coords[1];
int inOffset = outIdx * ${a};
int bestIndex = inOffset;
float bestValue = getA(batch, bestIndex);
for (int i = 0; i < ${a}; i++) {
int inIdx = ${o};
float candidate = getA(batch, inIdx);
if (candidate ${i} bestValue) {
bestValue = candidate;
bestIndex = inIdx;
}
}
setOutput(float(bestIndex));
}
`}},RQ=class{constructor(e,t,n,a){this.variableNames=["A"],this.packedInputs=!0,this.packedOutput=!0,w.assert(e.length>2,()=>`Packed arg${n.charAt(0).toUpperCase()+n.slice(1)} supports only inputs with rank above 2.`);let r=e[e.length-1],s=Math.ceil(r/t);this.outputShape=e.slice(0,-1),s>1&&this.outputShape.push(s),a||this.variableNames.push("bestIndicesA");let i=this.outputShape,o=i.length,l=mt(o),u=Nn("coords",o),p,d;if(s===1){d=o+1;let _=mt(d);p=`
${_} sourceLocR = ${_}(${u.join()}, 0);
++${u[o-1]};
${_} sourceLocG = ${_}(${u.join()}, 0);
++${u[o-2]};
${_} sourceLocA = ${_}(${u.join()}, 0);
--${u[o-1]};
${_} sourceLocB = ${_}(${u.join()}, 0);
--${u[o-2]};`}else d=o,p=`
${l} sourceLocR = coords;
++${u[o-1]};
${l} sourceLocG = coords;
++${u[o-2]};
${l} sourceLocA = coords;
--${u[o-1]};
${l} sourceLocB = coords;
--${u[o-2]};`;let c=["x","y","z","w","u","v"].slice(0,d),h="."+c[d-1],m=c.map(_=>"int "+_),f=Nn("sourceLocR",d-1).concat("inIdx.r"),g=Nn("sourceLocG",d-1).concat("inIdx.g"),y=Nn("sourceLocB",d-1).concat("inIdx.b"),b=Nn("sourceLocA",d-1).concat("inIdx.a"),x=n==="max"?"greaterThan":"lessThan",v=a?"":`
inIdx = round(vec4(getBestIndicesAChannel(${f.join()}),
getBestIndicesAChannel(${g.join()}),
getBestIndicesAChannel(${y.join()}),
getBestIndicesAChannel(${b.join()})));`,k=`vec4(
getAChannel(${f.join()}),
hasNextCol ? getAChannel(${g.join()}) : 0.,
hasNextRow ? getAChannel(${y.join()}) : 0.,
hasNextRow && hasNextCol ? getAChannel(${b.join()}) : 0.)`,T=a?"":`
float getBestIndicesAChannel(${m.join()}) {
return getChannel(getBestIndicesA(${c.join()}),
vec2(${c.slice(-2).join()}));
}`;this.userCode=`
float getAChannel(${m.join()}) {
return getChannel(getA(${c.join()}),
vec2(${c.slice(-2).join()}));
}
${T}
void main() {
${l} coords = getOutputCoords();
bool hasNextCol = ${u[o-1]} < ${i[o-1]-1};
bool hasNextRow = ${u[o-2]} < ${i[o-2]-1};
${p}
ivec4 srcIdx = ivec4(sourceLocR${h}, sourceLocG${h},
sourceLocB${h}, sourceLocA${h}) * ${t};
ivec4 inIdx = srcIdx;
vec4 bestIndex = vec4(inIdx);
vec4 bestValue = ${k};
for (int i = 0; i < ${t}; i++) {
inIdx = srcIdx;
${v}
vec4 candidate = ${k};
bvec4 nan = isnan(candidate);
bvec4 replace = bvec4(
vec4(${x}(candidate, bestValue)) * (vec4(1.0) - vec4(nan)));
bestValue = vec4(replace.x ? candidate.x : bestValue.x,
replace.y ? candidate.y : bestValue.y,
replace.z ? candidate.z : bestValue.z,
replace.w ? candidate.w : bestValue.w);
bestIndex = mix(bestIndex, vec4(inIdx), vec4(replace));
srcIdx++;
}
setOutput(bestIndex);
}
`}};function OC(e,t,n,a=null){let r=t.shape[0],s=t.shape[1];a!=null&&(r=a.shape[0],s=a.shape[1]);let i=C.computeOptimalWindowSize(s),o={windowSize:i,inSize:s,batchSize:r,outSize:Math.ceil(s/i)},l=new DQ(o,n,a==null),u=[t];a!=null&&u.push(a);let p=e.runWebGLProgram(l,u,"int32");if(p.shape[1]===1)return p;let d=OC(e,t,n,p);return e.disposeIntermediateTensorInfo(p),d}function LC(e,t,n,a=null){let r=a!=null?a.shape:t.shape,s=r[r.length-1],i=C.computeOptimalWindowSize(s),o=new RQ(r,i,n,a==null),l=a==null?[t]:[t,a],u=e.runWebGLProgram(o,l,"int32");if(u.shape.length===t.shape.length){let p=LC(e,t,n,u);return e.disposeIntermediateTensorInfo(u),p}return u}function zC(e,t,n,a){let r=[n];if(C.assertAxesAreInnerMostDims("arg"+a.charAt(0).toUpperCase()+a.slice(1),r,t.shape.length),!X().getBool("WEBGL_PACK_REDUCE")||t.shape.length<=2){let s=[],i=e.texData.get(t.dataId),o=i!==null&&i.isPacked,l=t;o&&(l=e.unpackTensor(t),s.push(l));let[u,p]=C.computeOutAndReduceShapes(l.shape,r),d=w.sizeFromShape(p),c=me({inputs:{x:l},backend:e,attrs:{shape:[-1,d]}});s.push(c);let h=OC(e,c,a);s.push(h);let m=me({inputs:{x:h},backend:e,attrs:{shape:u}});return s.forEach(f=>e.disposeIntermediateTensorInfo(f)),m}return LC(e,t,a)}function MQ(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{axis:s}=a,i=w.parseAxisParam(s,r.shape),o=C.getAxesPermutation(i,r.shape.length),l=r,u=[];o!=null&&(l=Wt({inputs:{x:r},backend:n,attrs:{perm:o}}),u.push(l),i=C.getInnerMostAxes(i.length,l.shape.length)),C.assertAxesAreInnerMostDims("argMax",[i[0]],l.shape.length);let p=zC(n,l,i[0],"max");return u.forEach(d=>n.disposeIntermediateTensorInfo(d)),p}var PQ={kernelName:ki,backendName:"webgl",kernelFunc:MQ};function OQ(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{axis:s}=a,i=w.parseAxisParam(s,r.shape),o=C.getAxesPermutation(i,r.shape.length),l=r,u=[];o!=null&&(l=Wt({inputs:{x:r},backend:n,attrs:{perm:o}}),u.push(l),i=C.getInnerMostAxes(i.length,l.shape.length)),C.assertAxesAreInnerMostDims("argMin",[i[0]],l.shape.length);let p=zC(n,l,i[0],"min");return u.forEach(d=>n.disposeIntermediateTensorInfo(d)),p}var LQ={kernelName:yc,backendName:"webgl",kernelFunc:OQ},zQ=Aa+`
if (abs(x) > 1.) {
return NAN;
}
return asin(x);
`,BQ=Qe({opSnippet:zQ}),WQ={kernelName:Ul,backendName:"webgl",kernelFunc:BQ},VQ=Aa+"return log(x + sqrt(x * x + 1.0));",UQ=Qe({opSnippet:VQ}),GQ={kernelName:Gl,backendName:"webgl",kernelFunc:UQ},HQ=Aa+`
return atan(x);
`,jQ=Qe({opSnippet:HQ}),qQ={kernelName:Hl,backendName:"webgl",kernelFunc:jQ},KQ=tQ+`
return atan(a, b);
`,XQ=`
vec4 result = atan(a, b);
vec4 isNaN = min(vec4(isnan(a)) + vec4(isnan(b)), vec4(1.0));
`+nQ+`
return result;
`,YQ=cn({opSnippet:KQ,packedOpSnippet:XQ}),QQ={kernelName:ql,backendName:"webgl",kernelFunc:YQ},JQ=Aa+`
if ((x < -1.0) || (x > 1.0)) return NAN;
return (log(1.0 + x) - log(1.0 - x)) / 2.0;`,ZQ=Qe({opSnippet:JQ}),eJ={kernelName:jl,backendName:"webgl",kernelFunc:ZQ},hc=class{constructor(e,t,n,a=!1,r=!1){if(this.variableNames=["x"],t==="avg"&&n)throw new Error("Cannot compute positions for average pool.");let s=e.filterWidth,i=e.strideHeight,o=e.strideWidth,l=e.dilationHeight,u=e.dilationWidth,p=e.effectiveFilterHeight,d=e.effectiveFilterWidth,c=e.padInfo.top,h=e.padInfo.left;this.outputShape=e.outShape;let m=t==="avg",f=`((batch * ${e.inHeight} + xR) * ${e.inWidth} + xC) * ${e.inChannels} + d`,g=`(xR * ${e.inWidth} + xC) * ${e.inChannels} + d`,y="0.0";if(m||(y="-1.0 / 1e-20"),n){let _=">=";this.userCode=`
const ivec2 strides = ivec2(${i}, ${o});
const ivec2 pads = ivec2(${c}, ${h});
void main() {
ivec4 coords = getOutputCoords();
int batch = coords[0];
int d = coords[3];
ivec2 xRCCorner = coords.yz * strides - pads;
int xRCorner = xRCCorner.x;
int xCCorner = xRCCorner.y;
// max/min x(?, ?, d) to get y(yR, yC, d).
// ? = to be determined
float minMaxValue = 0.0;
float minMaxValueFound = 0.0;
int minMaxPosition = 0;
float avgValue = 0.0;
for (int wR = 0; wR < ${p};
wR += ${l}) {
int xR = xRCorner + wR;
if (xR < 0 || xR >= ${e.inHeight}) {
continue;
}
for (int wC = 0; wC < ${d};
wC += ${u}) {
int xC = xCCorner + wC;
if (xC < 0 || xC >= ${e.inWidth}) {
continue;
}
float value = getX(batch, xR, xC, d);
// If a min / max value has already been found, use it. If not,
// use the current value.
float currMinMaxValue = mix(
value, minMaxValue, minMaxValueFound);
if (value ${_} currMinMaxValue) {
minMaxValue = value;
minMaxValueFound = 1.0;
minMaxPosition = ${a?r?f:g:`wR * ${d} + wC`};
}
}
}
setOutput(float(minMaxPosition));
}
`;return}let b="max",x=`${t}(${t}(${t}(minMaxValue[0], minMaxValue[1]), minMaxValue[2]), minMaxValue[3])`;t==="avg"&&(x="avgValue / count");let v=Math.floor(s/4)*4,k=s%4,T=`
if (${m}) {
avgValue += dot(values, ones);
} else {
minMaxValue = ${b}(values, minMaxValue);
}
`;this.userCode=`
const ivec2 strides = ivec2(${i}, ${o});
const ivec2 pads = ivec2(${c}, ${h});
const float initializationValue = ${y};
const vec4 ones = vec4(1.0, 1.0, 1.0, 1.0);
float count = 0.0;
float getValue(int batch, int xR, int xC, int d) {
if (xC < 0 || xC >= ${e.inWidth}) {
return initializationValue;
}
count += 1.0;
return getX(batch, xR, xC, d);
}
void main() {
ivec4 coords = getOutputCoords();
int batch = coords[0];
int d = coords[3];
ivec2 xRCCorner = coords.yz * strides - pads;
int xRCorner = xRCCorner.x;
int xCCorner = xRCCorner.y;
// max/min x(?, ?, d) to get y(yR, yC, d).
// ? = to be determined
vec4 minMaxValue = vec4(${y});
float avgValue = 0.0;
count = 0.0;
for (int wR = 0; wR < ${p};
wR += ${l}) {
int xR = xRCorner + wR;
if (xR < 0 || xR >= ${e.inHeight}) {
continue;
}
for (int wC = 0; wC < ${v}; wC += 4) {
int xC = xCCorner + wC * ${u};
vec4 values = vec4(
getValue(batch, xR, xC, d),
getValue(batch, xR, xC + ${u}, d),
getValue(batch, xR, xC + 2 * ${u}, d),
getValue(batch, xR, xC + 3 * ${u}, d)
);
${T}
}
int xC = xCCorner + ${v};
if (${k===1}) {
vec4 values = vec4(
getValue(batch, xR, xC, d),
initializationValue,
initializationValue,
initializationValue
);
${T}
} else if (${k===2}) {
vec4 values = vec4(
getValue(batch, xR, xC, d),
getValue(batch, xR, xC + ${u}, d),
initializationValue,
initializationValue
);
${T}
} else if (${k===3}) {
vec4 values = vec4(
getValue(batch, xR, xC, d),
getValue(batch, xR, xC + ${u}, d),
getValue(batch, xR, xC + 2 * ${u}, d),
initializationValue
);
${T}
}
}
setOutput(${x});
}
`}},U0=class{constructor(e,t,n,a=!1,r=!1){if(this.variableNames=["x"],t==="avg"&&n)throw new Error("Cannot compute positions for average pool.");let s=e.filterWidth,i=e.strideDepth,o=e.strideHeight,l=e.strideWidth,u=e.dilationDepth,p=e.dilationHeight,d=e.dilationWidth,c=e.effectiveFilterDepth,h=e.effectiveFilterHeight,m=e.effectiveFilterWidth,f=e.padInfo.front,g=e.padInfo.top,y=e.padInfo.left;this.outputShape=e.outShape;let b=t==="avg",x="0.0";if(b||(x="-1.0 / 1e-20"),n){let A=">=";this.userCode=`
const ivec3 strides =
ivec3(${i}, ${o}, ${l});
const ivec3 pads = ivec3(${f}, ${g}, ${y});
void main() {
ivec5 coords = getOutputCoords();
int batch = coords.x;
int ch = coords.u;
ivec3 xCorner = ivec3(coords.y, coords.z, coords.w) * strides - pads;
int xDCorner = xCorner.x;
int xRCorner = xCorner.y;
int xCCorner = xCorner.z;
// max/min x(?, ?, ?, ch) to get y(yD, yR, yC, ch).
// ? = to be determined
float minMaxValue = 0.0;
float minMaxValueFound = 0.0;
int minMaxPosition = 0;
for (int wD = 0; wD < ${c};
wD += ${u}) {
int xD = xDCorner + wD;
if (xD < 0 || xD >= ${e.inDepth}) {
continue;
}
for (int wR = 0; wR < ${h};
wR += ${p}) {
int xR = xRCorner + wR;
if (xR < 0 || xR >= ${e.inHeight}) {
continue;
}
for (int wC = 0; wC < ${m};
wC += ${d}) {
int xC = xCCorner + wC;
if (xC < 0 || xC >= ${e.inWidth}) {
continue;
}
float value = getX(batch, xD, xR, xC, ch);
// If a min / max value has already been found, use it. If not,
// use the current value.
float currMinMaxValue = mix(
value, minMaxValue, minMaxValueFound);
if (value ${A} currMinMaxValue) {
minMaxValue = value;
minMaxValueFound = 1.0;
minMaxPosition = ${a?r?`(((batch * ${e.inDepth} + xD) * ${e.inHeight} + xR) * ${e.inWidth} + xC) * ${e.inChannels} + ch`:`((xD * ${e.inHeight} + xR) * ${e.inWidth} + xC) * ${e.inChannels} + ch`:`wD * ${h} * ${m} +
wR * ${m} + wC`};
}
}
}
}
setOutput(float(minMaxPosition));
}
`;return}let v="max",k=`${t}(${t}(${t}(minMaxValue[0], minMaxValue[1]), minMaxValue[2]), minMaxValue[3])`;t==="avg"&&(k="avgValue / count");let T=Math.floor(s/4)*4,_=s%4,E=`
if (${b}) {
avgValue += dot(values, ones);
} else {
minMaxValue = ${v}(values, minMaxValue);
}
`;this.userCode=`
const ivec3 strides =
ivec3(${i}, ${o}, ${l});
const ivec3 pads = ivec3(${f}, ${g}, ${y});
const float initializationValue = ${x};
const vec4 ones = vec4(1.0, 1.0, 1.0, 1.0);
float count = 0.0;
float getValue(int batch, int xD, int xR, int xC, int ch) {
if (xC < 0 || xC >= ${e.inWidth}) {
return initializationValue;
}
count += 1.0;
return getX(batch, xD, xR, xC, ch);
}
void main() {
ivec5 coords = getOutputCoords();
int batch = coords.x;
int ch = coords.u;
ivec3 xCorner = ivec3(coords.y, coords.z, coords.w) * strides - pads;
int xDCorner = xCorner.x;
int xRCorner = xCorner.y;
int xCCorner = xCorner.z;
// max/min x(?, ?, ?, d) to get y(yD, yR, yC, ch).
// ? = to be determined
vec4 minMaxValue = vec4(${x});
float avgValue = 0.0;
count = 0.0;
for (int wD = 0; wD < ${c};
wD += ${u}) {
int xD = xDCorner + wD;
if (xD < 0 || xD >= ${e.inDepth}) {
continue;
}
for (int wR = 0; wR < ${h};
wR += ${p}) {
int xR = xRCorner + wR;
if (xR < 0 || xR >= ${e.inHeight}) {
continue;
}
for (int wC = 0; wC < ${T}; wC += 4) {
int xC = xCCorner + wC * ${d};
vec4 values = vec4(
getValue(batch, xD, xR, xC, ch),
getValue(batch, xD, xR, xC + ${d}, ch),
getValue(batch, xD, xR, xC + 2 * ${d}, ch),
getValue(batch, xD, xR, xC + 3 * ${d}, ch)
);
${E}
}
int xC = xCCorner + ${T};
if (${_===1}) {
vec4 values = vec4(
getValue(batch, xD, xR, xC, ch),
initializationValue,
initializationValue,
initializationValue
);
${E}
} else if (${_===2}) {
vec4 values = vec4(
getValue(batch, xD, xR, xC, ch),
getValue(batch, xD, xR, xC + ${d}, ch),
initializationValue,
initializationValue
);
${E}
} else if (${_===3}) {
vec4 values = vec4(
getValue(batch, xD, xR, xC, ch),
getValue(batch, xD, xR, xC + ${d}, ch),
getValue(batch, xD, xR, xC + 2 * ${d}, ch),
initializationValue
);
${E}
}
}
setOutput(${k});
}
}
`}};function tJ(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t;Zu(r,"avgPool");let{filterSize:s,strides:i,pad:o,dimRoundingMode:l}=a,u=1;w.assert(C.eitherStridesOrDilationsAreOne(i,u),()=>`Error in avgPool: Either strides or dilations must be 1. Got strides ${i} and dilations '${u}'`);let p=C.computePool2DInfo(r.shape,s,i,u,o,l);if(p.filterWidth===1&&p.filterHeight===1&&w.arraysEqual(p.inShape,p.outShape))return aa({inputs:{x:r},backend:n});let d=new hc(p,"avg",!1);return n.runWebGLProgram(d,[r],"float32")}var nJ={kernelName:Ii,backendName:"webgl",kernelFunc:tJ};function aJ(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{filterSize:s,strides:i,pad:o,dimRoundingMode:l,dataFormat:u}=a,p=[1,1,1],d=C.computePool3DInfo(r.shape,s,i,p,o,l,u),c=new U0(d,"avg",!1);return n.runWebGLProgram(c,[r],"float32")}var rJ={kernelName:bc,backendName:"webgl",kernelFunc:aJ},sJ=class{constructor(e){this.variableNames=["dy"],this.outputShape=e.inShape;let t=e.filterHeight,n=e.filterWidth,a=e.strideHeight,r=e.strideWidth,s=e.dilationHeight,i=e.dilationWidth,o=e.effectiveFilterHeight,l=e.effectiveFilterWidth,u=o-1-e.padInfo.top,p=l-1-e.padInfo.left,d=1/(t*n);this.userCode=`
const ivec2 pads = ivec2(${u}, ${p});
const float avgMultiplier = float(${d});
void main() {
ivec4 coords = getOutputCoords();
int b = coords[0];
int d = coords[3];
ivec2 dyRCCorner = coords.yz - pads;
int dyRCorner = dyRCCorner.x;
int dyCCorner = dyRCCorner.y;
// Convolve dy(?, ?, d) with pos mask(:, :, d) to get dx(xR, xC, d).
// ? = to be determined. : = across all values in that axis.
float dotProd = 0.0;
for (int wR = 0; wR < ${o};
wR += ${s}) {
float dyR = float(dyRCorner + wR) / ${a}.0;
if (dyR < 0.0 || dyR >= ${e.outHeight}.0 || fract(dyR) > 0.0) {
continue;
}
int idyR = int(dyR);
for (int wC = 0; wC < ${l};
wC+= ${i}) {
float dyC = float(dyCCorner + wC) / ${r}.0;
if (dyC < 0.0 || dyC >= ${e.outWidth}.0 ||
fract(dyC) > 0.0) {
continue;
}
int idyC = int(dyC);
float dyValue = getDy(b, idyR, idyC, d);
dotProd += dyValue * avgMultiplier;
}
}
setOutput(dotProd);
}
`}},iJ=class{constructor(e){this.variableNames=["dy"],this.outputShape=e.inShape;let t=e.filterDepth,n=e.filterHeight,a=e.filterWidth,r=e.strideDepth,s=e.strideHeight,i=e.strideWidth,o=e.dilationDepth,l=e.dilationHeight,u=e.dilationWidth,p=e.effectiveFilterDepth,d=e.effectiveFilterHeight,c=e.effectiveFilterWidth,h=p-1-e.padInfo.front,m=d-1-e.padInfo.top,f=c-1-e.padInfo.left,g=1/(t*n*a);this.userCode=`
const ivec3 pads = ivec3(${h}, ${m}, ${f});
const float avgMultiplier = float(${g});
void main() {
ivec5 coords = getOutputCoords();
int batch = coords.x;
int ch = coords.u;
ivec3 dyCorner = ivec3(coords.y, coords.z, coords.w) - pads;
int dyDCorner = dyCorner.x;
int dyRCorner = dyCorner.y;
int dyCCorner = dyCorner.z;
// Convolve dy(?, ?, ?, d) with pos mask(:, :, :, ch) to get
// dx(xD, xR, xC, ch).
// ? = to be determined. : = across all values in that axis.
float dotProd = 0.0;
for (int wD = 0; wD < ${p};
wD += ${o}) {
float dyD = float(dyDCorner + wD) / ${r}.0;
if (dyD < 0.0 || dyD >= ${e.outDepth}.0 || fract(dyD) > 0.0) {
continue;
}
int idyD = int(dyD);
for (int wR = 0; wR < ${d};
wR += ${l}) {
float dyR = float(dyRCorner + wR) / ${s}.0;
if (dyR < 0.0 || dyR >= ${e.outHeight}.0 ||
fract(dyR) > 0.0) {
continue;
}
int idyR = int(dyR);
for (int wC = 0; wC < ${c};
wC += ${u}) {
float dyC = float(dyCCorner + wC) / ${i}.0;
if (dyC < 0.0 || dyC >= ${e.outWidth}.0 ||
fract(dyC) > 0.0) {
continue;
}
int idyC = int(dyC);
float dyValue = getDy(batch, idyD, idyR, idyC, ch);
dotProd += dyValue * avgMultiplier;
}
}
}
setOutput(dotProd);
}
`}};function oJ(e){let{inputs:t,backend:n,attrs:a}=e,{dy:r,input:s}=t,i=s,{filterSize:o,strides:l,pad:u,dimRoundingMode:p}=a,d=[1,1,1],c=C.computePool3DInfo(i.shape,o,l,d,u,p),h=new iJ(c);return n.runWebGLProgram(h,[r],i.dtype)}var lJ={kernelName:Im,backendName:"webgl",kernelFunc:oJ};function uJ(e){let{inputs:t,backend:n,attrs:a}=e,{dy:r,input:s}=t,i=s;Zu([r,s],"avgPoolGrad");let{filterSize:o,strides:l,pad:u}=a,p=C.computePool2DInfo(i.shape,o,l,1,u),d=new sJ(p);return n.runWebGLProgram(d,[r],i.dtype)}var pJ={kernelName:km,backendName:"webgl",kernelFunc:uJ};function cJ(e){let{inputs:t,backend:n,attrs:a}=e,{a:r,b:s}=t,{transposeA:i,transposeB:o}=a;return gm({a:r,b:s,transposeA:i,transposeB:o,backend:n})}var dJ={kernelName:Si,backendName:"webgl",kernelFunc:cJ},hJ=class{constructor(e,t,n,a,r,s){this.outputShape=[],this.variableNames=["x","mean","variance"],C.assertAndGetBroadcastShape(e,t),C.assertAndGetBroadcastShape(e,n);let i="0.0";a!=null&&(C.assertAndGetBroadcastShape(e,a),this.variableNames.push("offset"),i="getOffsetAtOutCoords()");let o="1.0";r!=null&&(C.assertAndGetBroadcastShape(e,r),this.variableNames.push("scale"),o="getScaleAtOutCoords()"),this.outputShape=e,this.userCode=`
void main() {
float x = getXAtOutCoords();
float mean = getMeanAtOutCoords();
float variance = getVarianceAtOutCoords();
float offset = ${i};
float scale = ${o};
float inv = scale * inversesqrt(variance + float(${s}));
setOutput(dot(vec3(x, -mean, offset), vec3(inv, inv, 1)));
}
`}},mJ=class{constructor(e,t,n,a,r,s){this.packedInputs=!0,this.packedOutput=!0,this.variableNames=["x","mean","variance"],C.assertAndGetBroadcastShape(e,t),C.assertAndGetBroadcastShape(e,n);let i="vec4(0.0)";a!=null&&(C.assertAndGetBroadcastShape(e,a),this.variableNames.push("offset"),i="getOffsetAtOutCoords()");let o="vec4(1.0)";r!=null&&(C.assertAndGetBroadcastShape(e,r),this.variableNames.push("scale"),o="getScaleAtOutCoords()"),this.outputShape=e,this.userCode=`
void main() {
vec4 offset = ${i};
vec4 scale = ${o};
vec4 x = getXAtOutCoords();
vec4 mean = getMeanAtOutCoords();
vec4 variance = getVarianceAtOutCoords();
vec4 inv = scale * inversesqrt(variance + vec4(${s}));
setOutput((x - mean) * inv + offset);
}
`}},fJ=({inputs:e,backend:t,attrs:n})=>{let{x:a,mean:r,variance:s,offset:i,scale:o}=e;w.assert(r.shape.length===s.shape.length,()=>"Batch normalization gradient requires mean and variance to have equal ranks."),w.assert(i==null||r.shape.length===i.shape.length,()=>"Batch normalization gradient requires mean and offset to have equal ranks."),w.assert(o==null||r.shape.length===o.shape.length,()=>"Batch normalization gradient requires mean and scale to have equal ranks.");let{varianceEpsilon:l}=n;l==null&&(l=.001);let u=[a,r,s],p=null;i!=null&&(p=i.shape,u.push(i));let d=null;o!=null&&(d=o.shape,u.push(o));let c=X().getBool("WEBGL_PACK_NORMALIZATION")?new mJ(a.shape,r.shape,s.shape,p,d,l):new hJ(a.shape,r.shape,s.shape,p,d,l);return t.runWebGLProgram(c,u,u[0].dtype)},gJ={kernelName:Li,backendName:"webgl",kernelFunc:fJ},yJ=class{constructor(e){this.variableNames=["source"],this.outputShape=e,this.rank=e.length;let t=mt(this.rank);this.customUniforms=[{name:"start",arrayIndex:this.rank,type:"int"}];let n=bJ(this.rank),a,r=e.map((s,i)=>`sourceLoc.${wx[i]} = start[${i}] + coords.${wx[i]};`);a=`
${t} sourceLoc;
${t} coords = getOutputCoords();
${r.join(`
`)}
`,this.userCode=`
void main() {
${a}
setOutput(getSource(${n}));
}
`}},wx=["x","y","z","w","u","v"];function bJ(e){if(e===1)return"sourceLoc";if(e<=6)return wx.slice(0,e).map(t=>"sourceLoc."+t).join(",");throw Error(`Slicing for rank ${e} is not yet supported`)}var xJ=class{constructor(e){this.variableNames=["source"],this.packedInputs=!0,this.packedOutput=!0,this.outputShape=e,this.rank=e.length,this.customUniforms=[{name:"start",arrayIndex:this.rank,type:"int"}];let t=mt(this.rank),n=Nn("coords",this.rank),a=Nn("sourceLoc",this.rank),r=this.rank===1?"sourceLoc":`vec2(${a.slice(-2).join()})`,s=`getChannel(getSource(${a.join()}), ${r})`,i=`
result.x = ${s};
if (++${n[this.rank-1]} < ${e[this.rank-1]}) {
++${a[this.rank-1]};
result.y = ${s};
--${a[this.rank-1]};
}
`,o=this.rank===1?"":`
--${n[this.rank-1]};
if (++${n[this.rank-2]} < ${e[this.rank-2]}) {
++${a[this.rank-2]};
result.z = ${s};
if (++${n[this.rank-1]} < ${e[this.rank-1]}) {
++${a[this.rank-1]};
result.w = ${s};
}
}
`,l=this.rank<=4?`sourceLoc = coords +
${t}(${e.map((u,p)=>`start[${p}]`).join()});`:e.map((u,p)=>`${a[p]} = ${n[p]} + start[${p}];`).join(`
`);this.userCode=`
void main() {
${t} coords = getOutputCoords();
${t} sourceLoc;
${l}
vec4 result = vec4(0.);
${i}
${o}
setOutput(result);
}
`}};function vJ(e,t,n,a){let r=a.texData.get(e.dataId),s=a.makeTensorInfo(n,e.dtype),i=a.texData.get(s.dataId);Object.assign(i,r),i.refCount=1,i.shape=n,i.dtype=e.dtype;let o=Kt.computeFlatOffset(t,w.computeStrides(e.shape));r.slice&&(o+=r.slice.flatOffset),i.slice={flatOffset:o,origDataId:r.slice&&r.slice.origDataId||e.dataId};let l=a.dataRefCount.get(i.slice.origDataId)||1;return a.dataRefCount.set(i.slice.origDataId,l+1),s}function sp(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{begin:s,size:i}=a,[o,l]=Kt.parseSliceParams(r,s,i);if(Kt.assertParamsValid(r,o,l),w.sizeFromShape(l)===0)return n.makeTensorInfo(l,r.dtype,[]);if(n.shouldExecuteOnCPU([r])||r.dtype==="string"){let d=n.texData.get(r.dataId),c=u9(d.values,o,l,r.shape,r.dtype);return n.makeTensorInfo(l,r.dtype,c)}let{isPacked:u}=n.texData.get(r.dataId),p=Kt.isSliceContinous(r.shape,o,l);if(u||!p){let d=X().getBool("WEBGL_PACK_ARRAY_OPERATIONS")?new xJ(l):new yJ(l),c=[o];return n.runWebGLProgram(d,[r],r.dtype,c)}return n.uploadToGPU(r.dataId),vJ(r,o,l,n)}var wJ={kernelName:_u,backendName:"webgl",kernelFunc:sp},kJ=e=>{let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{blockShape:s,crops:i}=a;w.assert(r.shape.length<=4,()=>"batchToSpaceND for rank > 4 with a WebGL backend not implemented yet");let o=s.reduce((b,x)=>b*x),l=C.getReshaped(r.shape,s,o),u=C.getPermuted(l.length,s.length),p=C.getReshapedPermuted(r.shape,s,o),d=C.getSliceBeginCoords(i,s.length),c=C.getSliceSize(p,i,s.length),h=[],m=me({inputs:{x:r},backend:n,attrs:{shape:l}}),f=Wt({inputs:{x:m},backend:n,attrs:{perm:u}}),g=me({inputs:{x:f},backend:n,attrs:{shape:p}}),y=sp({inputs:{x:g},backend:n,attrs:{begin:d,size:c}});return h.push(m),h.push(f),h.push(g),h.forEach(b=>n.disposeIntermediateTensorInfo(b)),y},IJ={kernelName:Kl,backendName:"webgl",kernelFunc:kJ};function SJ(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,weights:s}=t,{size:i}=a,o=n.readSync(r.dataId),l=n.readSync(s.dataId),u=SC(o,l,s.dtype,s.shape,i);return n.makeTensorInfo([i],s.dtype,u)}var NJ={kernelName:Sm,backendName:"webgl",kernelFunc:SJ};function TJ(e){let{inputs:t,backend:n}=e,{s0:a,s1:r}=t,s=n.readSync(a.dataId),i=n.readSync(r.dataId),o=C.assertAndGetBroadcastShape(Array.from(s),Array.from(i));return n.makeTensorInfo([o.length],"int32",Int32Array.from(o))}var _J={kernelName:Nm,backendName:"webgl",kernelFunc:TJ},CJ="return float(a != b);",BC=cn({opSnippet:CJ,cpuKernelImpl:a9,dtype:"bool"}),EJ={kernelName:gu,backendName:"webgl",kernelFunc:BC};function hd(e){let{inputs:t,backend:n}=e,{input:a}=t,r=n.texData.get(a.dataId);return aa({inputs:{x:r.complexTensorInfos.real},backend:n})}var FJ={kernelName:Hm,backendName:"webgl",kernelFunc:hd},AJ="return float(int(x));";function $J(e,t){let n=new _r(e.shape,AJ),a=t.runWebGLProgram(n,[e],"int32");return{dataId:a.dataId,shape:a.shape,dtype:a.dtype}}function kx(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{dtype:s}=a;if(s==="complex64"){if(r.dtype==="complex64")return aa({inputs:{x:r},backend:n});let i=St(r.shape),o=kx({inputs:{x:r},backend:n,attrs:{dtype:"float32"}}),l=Cs({inputs:{real:o,imag:i},backend:n});return i.dispose(),n.disposeIntermediateTensorInfo(o),l}if(r.dtype==="complex64"){let i=hd({inputs:{input:r},backend:n}),o=kx({inputs:{x:i},backend:n,attrs:{dtype:s}});return n.disposeIntermediateTensorInfo(i),o}if(!w.hasEncodingLoss(r.dtype,s)){let i=aa({inputs:{x:r},backend:n});return{dataId:i.dataId,shape:i.shape,dtype:s}}if(s==="int32")return $J(r,n);if(s==="bool"){let i=n.makeTensorInfo([],"bool",w.getTypedArrayFromDType("bool",1)),o=BC({inputs:{a:r,b:i},backend:n});return n.disposeIntermediateTensorInfo(i),o}throw new Error(`Error in Cast: failed to cast ${r.dtype} to ${s}`)}var DJ={kernelName:Ni,backendName:"webgl",kernelFunc:kx},aI="return ceil(x);",RJ=Qe({opSnippet:aI,packedOpSnippet:aI,cpuKernelImpl:LY}),MJ={kernelName:Ti,backendName:"webgl",kernelFunc:RJ},PJ=class{constructor(e){this.variableNames=["A"],this.customUniforms=[{name:"minVal",type:"float"},{name:"maxVal",type:"float"}],this.outputShape=e,this.userCode=`
void main() {
float value = getAAtOutCoords();
if (isnan(value)) {
setOutput(value);
return;
}
setOutput(clamp(value, minVal, maxVal));
}
`}},OJ=class{constructor(e){this.variableNames=["A"],this.packedInputs=!0,this.packedOutput=!0,this.customUniforms=[{name:"minVal",type:"float"},{name:"maxVal",type:"float"}],this.outputShape=e,this.userCode=`
void main() {
vec4 value = getAAtOutCoords();
if (any(isnan(value))) {
setOutput(value);
return;
}
setOutput(clamp(value, vec4(minVal), vec4(maxVal)));
}
`}};function LJ(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{clipValueMin:s,clipValueMax:i}=a,o;X().getBool("WEBGL_PACK_CLIP")?o=new OJ(r.shape):o=new PJ(r.shape);let l=[[s],[i]];return n.runWebGLProgram(o,[r],r.dtype,l)}var zJ={kernelName:xs,backendName:"webgl",kernelFunc:LJ},BJ=class{constructor(e){this.variableNames=["real","imag"],this.outputShape=e,this.userCode=`
void main() {
float re = abs(getRealAtOutCoords());
float im = abs(getImagAtOutCoords());
float mx = max(re, im);
// sadly the length function in glsl is not underflow-safe
// (at least not on Intel GPUs). So the safe solution is
// to ensure underflow-safety in all cases.
setOutput(
mx == 0.0 ? 0.0 : mx * length(vec2(1, min(re, im)/mx))
);
}
`}};function rI(e,t){return{dataId:t.dataId,dtype:t.dtype,shape:e.shape}}function WJ(e){let{inputs:t,backend:n}=e,{x:a}=t,r=n.texData.get(a.dataId),s=new BJ(a.shape),i=[rI(a,r.complexTensorInfos.real),rI(a,r.complexTensorInfos.imag)];return n.runWebGLProgram(s,i,i[0].dtype)}var VJ={kernelName:xc,backendName:"webgl",kernelFunc:WJ},UJ=class{constructor(e){this.outputShape=[],this.outputShape=C.computeOutShape(e,1),this.variableNames=e.map((s,i)=>`T${i}`);let t=new Array(e.length-1);t[0]=e[0][1];for(let s=1;s<t.length;s++)t[s]=t[s-1]+e[s][1];let n=[`if (yC < ${t[0]}) setOutput(getT0(yR, yC));`];for(let s=1;s<t.length;s++){let i=t[s-1];n.push(`else if (yC < ${t[s]}) setOutput(getT${s}(yR, yC-${i}));`)}let a=t.length,r=t[t.length-1];n.push(`else setOutput(getT${a}(yR, yC-${r}));`),this.userCode=`
void main() {
ivec2 coords = getOutputCoords();
int yR = coords.x;
int yC = coords.y;
${n.join(`
`)}
}
`}},GJ=class{constructor(e,t){this.packedInputs=!0,this.packedOutput=!0,this.outputShape=[],this.outputShape=C.computeOutShape(e,t);let n=this.outputShape,a=n.length,r=mt(a),s=Nn("coords",a),i=["x","y","z","w","u","v"].slice(0,a);this.variableNames=e.map((m,f)=>`T${f}`);let o=new Array(e.length-1);o[0]=e[0][t];for(let m=1;m<o.length;m++)o[m]=o[m-1]+e[m][t];let l=i[t],u=i.slice(-2),p=i.join(),d=`if (${l} < ${o[0]}) {
return getChannel(
getT0(${p}), vec2(${u.join()}));
}`;for(let m=1;m<o.length;m++){let f=o[m-1];d+=`
if (${l} < ${o[m]} && ${l} >= ${o[m-1]}) {
return getChannel(
getT${m}(${Eh(i,l,f)}),
vec2(${Eh(u,l,f)}));
}`}let c=o.length,h=o[o.length-1];d+=`
return getChannel(
getT${c}(${Eh(i,l,h)}),
vec2(${Eh(u,l,h)}));`,this.userCode=`
float getValue(${i.map(m=>"int "+m)}) {
${d}
}
void main() {
${r} coords = getOutputCoords();
vec4 result = vec4(getValue(${s}), 0., 0., 0.);
${s[a-1]} = ${s[a-1]} + 1;
if (${s[a-1]} < ${n[a-1]}) {
result.g = getValue(${s});
}
${s[a-2]} = ${s[a-2]} + 1;
if (${s[a-2]} < ${n[a-2]}) {
result.a = getValue(${s});
}
${s[a-1]} = ${s[a-1]} - 1;
if (${s[a-2]} < ${n[a-2]} &&
${s[a-1]} < ${n[a-1]}) {
result.b = getValue(${s});
}
setOutput(result);
}
`}};function Eh(e,t,n){let a=e.indexOf(t);return e.map((r,s)=>s===a?`${r} - ${n}`:r).join()}function pg(e){let{inputs:t,backend:n}=e,{input:a}=t,r=n.texData.get(a.dataId);return aa({inputs:{x:r.complexTensorInfos.imag},backend:n})}var HJ={kernelName:Lm,backendName:"webgl",kernelFunc:pg};function ml(e,t,n){let a=e[0].dtype;if(a==="complex64"){let p=e.map(f=>hd({inputs:{input:f},backend:n})),d=e.map(f=>pg({inputs:{input:f},backend:n})),c=ml(p,t,n),h=ml(d,t,n),m=Cs({inputs:{real:c,imag:h},backend:n});return p.forEach(f=>n.disposeIntermediateTensorInfo(f)),d.forEach(f=>n.disposeIntermediateTensorInfo(f)),n.disposeIntermediateTensorInfo(c),n.disposeIntermediateTensorInfo(h),m}let r=n.shouldExecuteOnCPU(e);if(a==="string"&&(r=!0),r){let p=e.map(y=>{let b=w.sizeFromShape(y.shape.slice(t));return me({inputs:{x:y},backend:n,attrs:{shape:[-1,b]}})}),d=p.map(y=>({vals:n.readSync(y.dataId),shape:y.shape})),c=C.computeOutShape(p.map(y=>y.shape),1),h=p[0].shape[0]===1,m=zY(d,c,a,h),f=C.computeOutShape(e.map(y=>y.shape),t),g=n.makeTensorInfo(f,a,m);return p.forEach(y=>n.disposeIntermediateTensorInfo(y)),g}if(e.length>X().getNumber("WEBGL_MAX_TEXTURES_IN_SHADER")){let p=Math.floor(e.length/2),d=ml(e.slice(0,p),t,n),c=ml(e.slice(p),t,n),h=ml([d,c],t,n);return n.disposeIntermediateTensorInfo(d),n.disposeIntermediateTensorInfo(c),h}if(X().getBool("WEBGL_PACK_ARRAY_OPERATIONS")&&e[0].shape.length>1){let p=new GJ(e.map(d=>d.shape),t);return n.runWebGLProgram(p,e,a)}let{tensors2D:s,outShape:i}=jJ(e,t,n),o=new UJ(s.map(p=>p.shape)),l=n.runWebGLProgram(o,s,a);s.forEach(p=>n.disposeIntermediateTensorInfo(p));let u=me({inputs:{x:l},attrs:{shape:i},backend:n});return n.disposeIntermediateTensorInfo(l),u}function jJ(e,t,n){let a=C.computeOutShape(e.map(r=>r.shape),t);return{tensors2D:e.map(r=>me({inputs:{x:r},attrs:{shape:[-1,w.sizeFromShape(r.shape.slice(t))]},backend:n})),outShape:a}}function WC(e){let{inputs:t,backend:n,attrs:a}=e,{axis:r}=a,s=w.parseAxisParam(r,t[0].shape)[0],i=C.computeOutShape(t.map(u=>u.shape),s);if(w.sizeFromShape(i)===0)return n.makeTensorInfo(i,t[0].dtype,[]);let o=t.filter(u=>w.sizeFromShape(u.shape)>0);if(o.length===1)return aa({inputs:{x:o[0]},backend:n});let l=o.map(u=>u.shape);return C.assertParamsConsistent(l,s),ml(o,s,n)}var qJ={kernelName:Xl,backendName:"webgl",kernelFunc:WC},VC=class{constructor(e,t=!1,n=null,a=!1,r=!1){this.variableNames=["x","W"],this.outputShape=e.outShape;let s=e.padInfo.top,i=e.padInfo.left,o=e.strideHeight,l=e.strideWidth,u=e.dilationHeight,p=e.dilationWidth,d=e.filterHeight,c=e.filterWidth,h=Math.floor(e.inChannels/4)*4,m=e.inChannels%4,f=e.dataFormat==="channelsLast",g=f?1:2,y=f?2:3,b=f?3:1,x="",v="";n&&(a?x=`float activation(float a) {
float b = getPreluActivationWeightsAtOutCoords();
${n}
}`:r?x=`float activation(float a) {
float b = getLeakyreluAlphaAtOutCoords();
${n}
}`:x=`
float activation(float x) {
${n}
}
`,v="result = activation(result);");let k=t?"result += getBiasAtOutCoords();":"";t&&this.variableNames.push("bias"),a&&this.variableNames.push("preluActivationWeights"),r&&this.variableNames.push("leakyreluAlpha"),this.userCode=`
${x}
const ivec2 strides = ivec2(${o}, ${l});
const ivec2 pads = ivec2(${s}, ${i});
void main() {
ivec4 coords = getOutputCoords();
int batch = coords[0];
int d2 = coords[${b}];
ivec2 xRCCorner =
ivec2(coords[${g}], coords[${y}]) * strides - pads;
int xRCorner = xRCCorner.x;
int xCCorner = xRCCorner.y;
// Convolve x(?, ?, d1) with w(:, :, d1, d2) to get y(yR, yC, d2).
// ? = to be determined. : = across all values in that axis.
float dotProd = 0.0;
for (int wR = 0; wR < ${d}; wR++) {
int xR = xRCorner + wR * ${u};
if (xR < 0 || xR >= ${e.inHeight}) {
continue;
}
for (int wC = 0; wC < ${c}; wC++) {
int xC = xCCorner + wC * ${p};
if (xC < 0 || xC >= ${e.inWidth}) {
continue;
}
for (int d1 = 0; d1 < ${h}; d1 += 4) {
vec4 wValues = vec4(
getW(wR, wC, d1, d2),
getW(wR, wC, d1 + 1, d2),
getW(wR, wC, d1 + 2, d2),
getW(wR, wC, d1 + 3, d2)
);
if (${f}) {
vec4 xValues = vec4(
getX(batch, xR, xC, d1),
getX(batch, xR, xC, d1 + 1),
getX(batch, xR, xC, d1 + 2),
getX(batch, xR, xC, d1 + 3)
);
dotProd += dot(xValues, wValues);
} else {
vec4 xValues = vec4(
getX(batch, d1, xR, xC),
getX(batch, d1 + 1, xR, xC),
getX(batch, d1 + 2, xR, xC),
getX(batch, d1 + 3, xR, xC)
);
dotProd += dot(xValues, wValues);
}
}
if (${m===1}) {
if (${f}) {
dotProd +=
getX(batch, xR, xC, ${h}) *
getW(wR, wC, ${h}, d2);
} else {
dotProd +=
getX(batch, ${h}, xR, xC) *
getW(wR, wC, ${h}, d2);
}
} else if (${m===2}) {
vec2 wValues = vec2(
getW(wR, wC, ${h}, d2),
getW(wR, wC, ${h} + 1, d2)
);
if (${f}) {
vec2 xValues = vec2(
getX(batch, xR, xC, ${h}),
getX(batch, xR, xC, ${h} + 1)
);
dotProd += dot(xValues, wValues);
} else {
vec2 xValues = vec2(
getX(batch, ${h}, xR, xC),
getX(batch, ${h} + 1, xR, xC)
);
dotProd += dot(xValues, wValues);
}
} else if (${m===3}) {
vec3 wValues = vec3(
getW(wR, wC, ${h}, d2),
getW(wR, wC, ${h} + 1, d2),
getW(wR, wC, ${h} + 2, d2)
);
if (${f}) {
vec3 xValues = vec3(
getX(batch, xR, xC, ${h}),
getX(batch, xR, xC, ${h} + 1),
getX(batch, xR, xC, ${h} + 2)
);
dotProd += dot(xValues, wValues);
} else {
vec3 xValues = vec3(
getX(batch, ${h}, xR, xC),
getX(batch, ${h} + 1, xR, xC),
getX(batch, ${h} + 2, xR, xC)
);
dotProd += dot(xValues, wValues);
}
}
}
}
float result = dotProd;
${k}
${v}
setOutput(result);
}
`}},KJ=class{constructor(e){this.variableNames=["x","W"],this.outputShape=e.outShape;let t=e.padInfo.front,n=e.padInfo.top,a=e.padInfo.left,r=e.strideDepth,s=e.strideHeight,i=e.strideWidth,o=e.dilationDepth,l=e.dilationHeight,u=e.dilationWidth,p=e.filterDepth,d=e.filterHeight,c=e.filterWidth,h=Math.floor(e.inChannels/4)*4,m=e.inChannels%4;this.userCode=`
const ivec3 strides = ivec3(${r}, ${s}, ${i});
const ivec3 pads = ivec3(${t}, ${n}, ${a});
void main() {
ivec5 coords = getOutputCoords();
int batch = coords.x;
int d2 = coords.u;
ivec3 xFRCCorner = ivec3(coords.y, coords.z, coords.w) * strides - pads;
int xFCorner = xFRCCorner.x;
int xRCorner = xFRCCorner.y;
int xCCorner = xFRCCorner.z;
// Convolve x(?, ?, ?, d1) with w(:, :, :, d1, d2) to get
// y(yF, yR, yC, d2). ? = to be determined. : = across all
// values in that axis.
float dotProd = 0.0;
for (int wF = 0; wF < ${p}; wF++) {
int xF = xFCorner + wF * ${o};
if (xF < 0 || xF >= ${e.inDepth}) {
continue;
}
for (int wR = 0; wR < ${d}; wR++) {
int xR = xRCorner + wR * ${l};
if (xR < 0 || xR >= ${e.inHeight}) {
continue;
}
for (int wC = 0; wC < ${c}; wC++) {
int xC = xCCorner + wC * ${u};
if (xC < 0 || xC >= ${e.inWidth}) {
continue;
}
for (int d1 = 0; d1 < ${h}; d1 += 4) {
vec4 xValues = vec4(
getX(batch, xF, xR, xC, d1),
getX(batch, xF, xR, xC, d1 + 1),
getX(batch, xF, xR, xC, d1 + 2),
getX(batch, xF, xR, xC, d1 + 3)
);
vec4 wValues = vec4(
getW(wF, wR, wC, d1, d2),
getW(wF, wR, wC, d1 + 1, d2),
getW(wF, wR, wC, d1 + 2, d2),
getW(wF, wR, wC, d1 + 3, d2)
);
dotProd += dot(xValues, wValues);
}
if (${m===1}) {
dotProd +=
getX(batch, xF, xR, xC, ${h}) *
getW(wF, wR, wC, ${h}, d2);
} else if (${m===2}) {
vec2 xValues = vec2(
getX(batch, xF, xR, xC, ${h}),
getX(batch, xF, xR, xC, ${h} + 1)
);
vec2 wValues = vec2(
getW(wF, wR, wC, ${h}, d2),
getW(wF, wR, wC, ${h} + 1, d2)
);
dotProd += dot(xValues, wValues);
} else if (${m===3}) {
vec3 xValues = vec3(
getX(batch, xF, xR, xC, ${h}),
getX(batch, xF, xR, xC, ${h} + 1),
getX(batch, xF, xR, xC, ${h} + 2)
);
vec3 wValues = vec3(
getW(wF, wR, wC, ${h}, d2),
getW(wF, wR, wC, ${h} + 1, d2),
getW(wF, wR, wC, ${h} + 2, d2)
);
dotProd += dot(xValues, wValues);
}
}
}
}
setOutput(dotProd);
}
`}},XJ=class{constructor(e,t){this.variableNames=["A"],this.packedInputs=!0,this.packedOutput=!0,this.customUniforms=[{name:"inputShape",type:"ivec3"},{name:"pad",type:"ivec2"},{name:"stride",type:"ivec2"},{name:"dilation",type:"ivec2"},{name:"inChannels",type:"int"},{name:"itemsPerBlockRow",type:"int"},{name:"outWidth",type:"int"}],this.outputShape=e,this.enableShapeUniforms=Hn(this.outputShape.length);let{dataFormat:n}=t,a=En(),r=n==="channelsLast",s=r?0:1,i=r?1:2,o=this.enableShapeUniforms?"if(blockIndex < outShape[1] && pos < outShape[0]) {":`if(blockIndex < ${e[1]} && pos < ${e[0]}) {`,l="";for(let u=0;u<=1;u++)for(let p=0;p<=1;p++)l+=`
blockIndex = rc.y + ${p};
pos = rc.x + ${u};
${o}
offsetY = int(blockIndex / outWidth) * stride[0] - pad[0];
d0 = offsetY + dilation[0] * (pos / itemsPerBlockRow);
if(d0 < inputShape[${s}] && d0 >= 0) {
// Use custom imod instead mod. On Intel GPU, mod may generate
// unexpected value.
// https://github.com/tensorflow/tfjs/issues/5447
offsetX = imod(blockIndex, outWidth) * stride[1] - pad[1];
d1 = offsetX + dilation[1] * (imod(pos, itemsPerBlockRow) /
inChannels);
if(d1 < inputShape[${i}] && d1 >= 0) {
ch = imod(pos, inChannels);
if (${r}) {
innerDims = vec2(d1, ch);
result[${u*2+p}] = getChannel(
getA(d0, int(innerDims.x),
int(innerDims.y)), innerDims);
} else {
innerDims = vec2(d0, d1);
result[${u*2+p}] = getChannel(
getA(ch, int(innerDims.x),
int(innerDims.y)), innerDims);
}
}
}
}
`;this.userCode=`
void main() {
ivec2 rc = getOutputCoords();
vec4 result = vec4(0);
int blockIndex, pos, offsetY, d0, offsetX, d1, ch;
vec2 innerDims;
${l}
${a.output} = result;
}
`}};function UC({x:e,filter:t,convInfo:n,backend:a,bias:r=null,preluActivationWeights:s=null,leakyreluAlpha:i=0,activation:o=null}){let l=e.shape,u=a.texData.get(e.dataId),p=n.inChannels,d=l[0]*l[1]*l[2],c=n.outChannels,h=n.dataFormat==="channelsLast",m=!1,f=!1,g,y=[];if(s!=null&&!h&&s.shape.length===3){let b=Wt({inputs:{x:s},backend:a,attrs:{perm:[1,2,0]}});y.push(b),s=b}if(!((d===1||c===1)&&p>PC)&&u.isPacked&&h&&u.texture!=null&&l[2]%2!==0&&w.arraysEqual(u.shape.slice(-3),l.slice(-3))){let b=l[0]*l[1]*(l[2]+1),x={dataId:e.dataId,shape:[1,b,n.inChannels],dtype:e.dtype},v=u.shape;u.shape=u.shape.slice(),u.shape[u.shape.length-2]++,w.assert(dc(u.shape,x.shape),()=>`packed reshape ${u.shape} to ${x.shape} isn't free`);let k=me({inputs:{x:t},backend:a,attrs:{shape:[1,n.inChannels,n.outChannels]}});y.push(k);let T=gm({a:x,b:k,backend:a,transposeA:m,transposeB:f,bias:r,activation:o,preluActivationWeights:s,leakyreluAlpha:i}),_=a.texData.get(T.dataId);w.assert(_.isPacked,()=>"batchMatMul result is expected to be packed"),u.shape=v,_.shape=n.outShape,g=aa({inputs:{x:T},backend:a}),g.shape=n.outShape,y.push(T)}else{let b=h?e:Wt({inputs:{x:e},backend:a,attrs:{perm:[0,2,3,1]}}),x=b.shape,v=x[0]*x[1]*x[2],k=me({inputs:{x:b},backend:a,attrs:{shape:[1,v,n.inChannels]}}),T=me({inputs:{x:t},backend:a,attrs:{shape:[1,n.inChannels,n.outChannels]}}),_=gm({a:k,b:T,transposeA:m,transposeB:f,backend:a,bias:r,activation:o,preluActivationWeights:s,leakyreluAlpha:i}),E=[n.batchSize,n.outHeight,n.outWidth,n.outChannels],A=me({inputs:{x:_},backend:a,attrs:{shape:E}});g=h?A:Wt({inputs:{x:A},backend:a,attrs:{perm:[0,3,1,2]}}),h||(y.push(b),y.push(A)),y.push(k),y.push(T),y.push(_)}for(let b of y)a.disposeIntermediateTensorInfo(b);return g}function GC({x:e,filter:t,convInfo:n,backend:a,bias:r=null,preluActivationWeights:s=null,leakyreluAlpha:i=0,activation:o=null}){let{filterWidth:l,filterHeight:u,inChannels:p,outWidth:d,outHeight:c,dataFormat:h}=n,m=h==="channelsLast",f=l*u*p,g=c*d,y=[f,g],b=!0,x=!1,v=[];if(s!=null&&!m&&s.shape.length===3){let Y=Wt({inputs:{x:s},backend:a,attrs:{perm:[1,2,0]}});v.push(Y),s=Y}let k=me({inputs:{x:e},backend:a,attrs:{shape:e.shape.slice(1)}}),T=me({inputs:{x:t},backend:a,attrs:{shape:[1,f,w.sizeFromShape(t.shape)/f]}});v.push(k),v.push(T);let _=new XJ(y,n),E=[k.shape,[n.padInfo.top,n.padInfo.left],[n.strideHeight,n.strideWidth],[n.dilationHeight,n.dilationWidth],[n.inChannels],[n.filterWidth*n.inChannels],[n.outWidth]],A=a.runWebGLProgram(_,[k],"float32",E),M=me({inputs:{x:A},backend:a,attrs:{shape:[1,y[0],y[1]]}});v.push(A),v.push(M);let $=r!=null,S=s!=null,P=o==="leakyrelu",V=o?og(o,!0):null,j=new MC(M.shape,T.shape,[1,g,n.outChannels],b,x,$,V,S,P),q=[M,T];if(r&&q.push(r),S&&q.push(s),P){let Y=a.makeTensorInfo([],"float32",w.createScalarValue(i,"float32"));q.push(Y),v.push(Y)}let K=a.runWebGLProgram(j,q,"float32"),Z=[1,c,d,n.outChannels],ee=me({inputs:{x:K},backend:a,attrs:{shape:Z}}),re=m?ee:Wt({inputs:{x:ee},backend:a,attrs:{perm:[0,3,1,2]}});m||v.push(ee),v.push(K);for(let Y of v)a.disposeIntermediateTensorInfo(Y);return re}function YJ(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,filter:s}=t,{strides:i,pad:o,dataFormat:l,dilations:u,dimRoundingMode:p}=a,d=C.convertConv2DDataFormat(l),c=C.computeConv2DInfo(r.shape,s.shape,i,u,o,p,!1,d),h;if(c.filterHeight===1&&c.filterWidth===1&&c.dilationHeight===1&&c.dilationWidth===1&&c.strideHeight===1&&c.strideWidth===1&&(c.padInfo.type==="SAME"||c.padInfo.type==="VALID"))h=UC({x:r,filter:s,convInfo:c,backend:n});else if(X().getBool("WEBGL_CONV_IM2COL")&&r.shape[0]===1)h=GC({x:r,filter:s,convInfo:c,backend:n});else{let f=new VC(c);h=n.runWebGLProgram(f,[r,s],"float32")}let m=me({inputs:{x:h},backend:n,attrs:{shape:c.outShape}});return n.disposeIntermediateTensorInfo(h),m}var QJ={kernelName:_i,backendName:"webgl",kernelFunc:YJ},JJ=class{constructor(e){this.variableNames=["x","dy"],this.outputShape=e.filterShape;let t=e.strideHeight,n=e.strideWidth,a=e.padInfo.top,r=e.padInfo.left,s=e.dataFormat==="channelsLast";this.userCode=`
void main() {
ivec4 coords = getOutputCoords();
int wR = coords.x;
int wC = coords.y;
int d1 = coords.z;
int d2 = coords.w;
// Convolve x(?, ?, d1) with dy(:, :, d2) to get dw(wR, wC, d1, d2).
// ? = to be determined. : = across all values in that axis.
float dotProd = 0.0;
for (int b = 0; b < ${e.batchSize}; b++) {
for (int yR = 0; yR < ${e.outHeight}; yR++) {
int xR = wR + yR * ${t} - ${a};
if (xR < 0 || xR >= ${e.inHeight}) {
continue;
}
for (int yC = 0; yC < ${e.outWidth}; yC++) {
int xC = wC + yC * ${n} - ${r};
if (xC < 0 || xC >= ${e.inWidth}) {
continue;
}
if (${s}) {
float dyValue = getDy(b, yR, yC, d2);
float xValue = getX(b, xR, xC, d1);
dotProd += (xValue * dyValue);
} else {
float dyValue = getDy(b, d2, yR, yC);
float xValue = getX(b, d1, xR, xC);
dotProd += (xValue * dyValue);
}
}
}
}
setOutput(dotProd);
}
`}},ZJ=class{constructor(e){this.variableNames=["dy","W"],this.outputShape=e.inShape;let t=e.filterHeight,n=e.filterWidth,a=e.strideHeight,r=e.strideWidth,s=e.dataFormat==="channelsLast",i=t-1-e.padInfo.top,o=n-1-e.padInfo.left,l=s?1:2,u=s?2:3,p=s?3:1;this.userCode=`
const ivec2 pads = ivec2(${i}, ${o});
void main() {
ivec4 coords = getOutputCoords();
int batch = coords[0];
int d1 = coords[${p}];
ivec2 dyCorner = ivec2(coords[${l}], coords[${u}]) - pads;
int dyRCorner = dyCorner.x;
int dyCCorner = dyCorner.y;
// Convolve dy(?, ?, d2) with w(:, :, d1, d2) to compute dx(xR, xC, d1).
// ? = to be determined. : = across all values in that axis.
float dotProd = 0.0;
for (int wR = 0; wR < ${t}; wR++) {
float dyR = float(dyRCorner + wR) / ${a}.0;
if (dyR < 0.0 || dyR >= ${e.outHeight}.0 || fract(dyR) > 0.0) {
continue;
}
int idyR = int(dyR);
int wRPerm = ${t} - 1 - wR;
for (int wC = 0; wC < ${n}; wC++) {
float dyC = float(dyCCorner + wC) / ${r}.0;
if (dyC < 0.0 || dyC >= ${e.outWidth}.0 ||
fract(dyC) > 0.0) {
continue;
}
int idyC = int(dyC);
int wCPerm = ${n} - 1 - wC;
for (int d2 = 0; d2 < ${e.outChannels}; d2++) {
if (${s}) {
float xValue = getDy(batch, idyR, idyC, d2);
float wValue = getW(wRPerm, wCPerm, d1, d2);
dotProd += xValue * wValue;
} else {
float xValue = getDy(batch, d2, idyR, idyC);
float wValue = getW(wRPerm, wCPerm, d1, d2);
dotProd += xValue * wValue;
}
}
}
}
setOutput(dotProd);
}
`}},eZ=class{constructor(e){this.variableNames=["x","dy"],this.outputShape=e.filterShape;let t=e.strideDepth,n=e.strideHeight,a=e.strideWidth,r=e.padInfo.front,s=e.padInfo.top,i=e.padInfo.left;this.userCode=`
void main() {
ivec5 coords = getOutputCoords();
int wF = coords.x;
int wR = coords.y;
int wC = coords.z;
int d1 = coords.w;
int d2 = coords.u;
float dotProd = 0.0;
for (int b = 0; b < ${e.batchSize}; b++) {
for (int yF = 0; yF < ${e.outDepth}; yF++) {
int xF = wF + yF * ${t} - ${r};
if (xF < 0 || xF >= ${e.inDepth}) {
continue;
}
for (int yR = 0; yR < ${e.outHeight}; yR++) {
int xR = wR + yR * ${n} - ${s};
if (xR < 0 || xR >= ${e.inHeight}) {
continue;
}
for (int yC = 0; yC < ${e.outWidth}; yC++) {
int xC = wC + yC * ${a} - ${i};
if (xC < 0 || xC >= ${e.inWidth}) {
continue;
}
float dyValue = getDy(b, yF, yR, yC, d2);
float xValue = getX(b, xF, xR, xC, d1);
dotProd += (xValue * dyValue);
}
}
}
}
setOutput(dotProd);
}
`}},tZ=class{constructor(e){this.variableNames=["dy","W"],this.outputShape=e.inShape;let t=e.filterDepth,n=e.filterHeight,a=e.filterWidth,r=e.strideDepth,s=e.strideHeight,i=e.strideWidth,o=t-1-e.padInfo.front,l=n-1-e.padInfo.top,u=a-1-e.padInfo.left;this.userCode=`
const ivec3 pads = ivec3(${o}, ${l}, ${u});
void main() {
ivec5 coords = getOutputCoords();
int batch = coords.x;
int d1 = coords.u;
ivec3 dyCorner = ivec3(coords.y, coords.z, coords.w) - pads;
int dyFCorner = dyCorner.x;
int dyRCorner = dyCorner.y;
int dyCCorner = dyCorner.z;
float dotProd = 0.0;
for (int wF = 0; wF < ${t}; wF++) {
float dyF = float(dyFCorner + wF) / ${r}.0;
if (dyF < 0.0 || dyF >= ${e.outDepth}.0 || fract(dyF) > 0.0) {
continue;
}
int idyF = int(dyF);
int wFPerm = ${t} - 1 - wF;
for (int wR = 0; wR < ${n}; wR++) {
float dyR = float(dyRCorner + wR) / ${s}.0;
if (dyR < 0.0 || dyR >= ${e.outHeight}.0 ||
fract(dyR) > 0.0) {
continue;
}
int idyR = int(dyR);
int wRPerm = ${n} - 1 - wR;
for (int wC = 0; wC < ${a}; wC++) {
float dyC = float(dyCCorner + wC) / ${i}.0;
if (dyC < 0.0 || dyC >= ${e.outWidth}.0 ||
fract(dyC) > 0.0) {
continue;
}
int idyC = int(dyC);
int wCPerm = ${a} - 1 - wC;
for (int d2 = 0; d2 < ${e.outChannels}; d2++) {
float xValue = getDy(batch, idyF, idyR, idyC, d2);
float wValue = getW(wFPerm, wRPerm, wCPerm, d1, d2);
dotProd += xValue * wValue;
}
}
}
}
setOutput(dotProd);
}
`}};function nZ(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,dy:s}=t,{strides:i,pad:o,dataFormat:l,dimRoundingMode:u,filterShape:p}=a,d=C.convertConv2DDataFormat(l),c=C.computeConv2DInfo(r.shape,p,i,1,o,u,!1,d),h=new JJ(c);return n.runWebGLProgram(h,[r,s],"float32")}var aZ={kernelName:_m,backendName:"webgl",kernelFunc:nZ};function rZ(e){let{inputs:t,backend:n,attrs:a}=e,{dy:r,filter:s}=t,{inputShape:i,strides:o,pad:l,dataFormat:u,dimRoundingMode:p}=a,d=C.convertConv2DDataFormat(u),c=C.computeConv2DInfo(i,s.shape,o,1,l,p,!1,d),h=new ZJ(c);return n.runWebGLProgram(h,[r,s],"float32")}var sZ={kernelName:Ci,backendName:"webgl",kernelFunc:rZ};function iZ(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,filter:s}=t,{strides:i,pad:o,dilations:l}=a,u=C.computeConv3DInfo(r.shape,s.shape,i,l,o),p=new KJ(u);return n.runWebGLProgram(p,[r,s],"float32")}var oZ={kernelName:vc,backendName:"webgl",kernelFunc:iZ};function lZ(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,dy:s}=t,{strides:i,pad:o,filterShape:l}=a,u=C.computeConv3DInfo(r.shape,l,i,1,o),p=new eZ(u);return n.runWebGLProgram(p,[r,s],"float32")}var uZ={kernelName:Cm,backendName:"webgl",kernelFunc:lZ};function pZ(e){let{inputs:t,backend:n,attrs:a}=e,{dy:r,filter:s}=t,{pad:i,strides:o,inputShape:l}=a,u=C.computeConv3DInfo(l,s.shape,o,1,i),p=new tZ(u);return n.runWebGLProgram(p,[r,s],"float32")}var cZ={kernelName:Em,backendName:"webgl",kernelFunc:pZ},dZ=rp+`
return cos(x);
`,hZ=Qe({opSnippet:dZ}),mZ={kernelName:Ei,backendName:"webgl",kernelFunc:hZ},fZ=`
float e2x = exp(-x);
return (e2x + 1.0 / e2x) / 2.0;
`,gZ=Qe({opSnippet:fZ}),yZ={kernelName:Fi,backendName:"webgl",kernelFunc:gZ},bZ=class{constructor(e,t,n,a,r){this.variableNames=["Image","Boxes","BoxInd"],this.outputShape=[];let[s,i,o,l]=e,[u]=t,[p,d]=n;this.outputShape=[u,p,d,l];let c=a==="bilinear"?1:0,[h,m]=[`${i-1}.0`,`${o-1}.0`],[f,g,y]=p>1?[`${(i-1)/(p-1)}`,"(y2-y1) * height_ratio",`y1*${h} + float(y)*(height_scale)`]:["0.0","0.0",`0.5 * (y1+y2) * ${h}`],[b,x,v]=d>1?[`${(o-1)/(d-1)}`,"(x2-x1) * width_ratio",`x1*${m} + float(x)*(width_scale)`]:["0.0","0.0",`0.5 * (x1+x2) * ${m}`];this.userCode=`
const float height_ratio = float(${f});
const float width_ratio = float(${b});
void main() {
ivec4 coords = getOutputCoords();
int b = coords[0];
int y = coords[1];
int x = coords[2];
int d = coords[3];
// get box vals
float y1 = getBoxes(b,0);
float x1 = getBoxes(b,1);
float y2 = getBoxes(b,2);
float x2 = getBoxes(b,3);
// get image in batch index
int bInd = round(getBoxInd(b));
if(bInd < 0 || bInd >= ${s}) {
return;
}
float height_scale = ${g};
float width_scale = ${x};
float in_y = ${y};
if( in_y < 0.0 || in_y > ${h} ) {
setOutput(float(${r}));
return;
}
float in_x = ${v};
if( in_x < 0.0 || in_x > ${m} ) {
setOutput(float(${r}));
return;
}
vec2 sourceFracIndexCR = vec2(in_x,in_y);
if(${c} == 1) {
// Compute the four integer indices.
ivec2 sourceFloorCR = ivec2(sourceFracIndexCR);
ivec2 sourceCeilCR = ivec2(ceil(sourceFracIndexCR));
float topLeft = getImage(b, sourceFloorCR.y, sourceFloorCR.x, d);
float bottomLeft = getImage(b, sourceCeilCR.y, sourceFloorCR.x, d);
float topRight = getImage(b, sourceFloorCR.y, sourceCeilCR.x, d);
float bottomRight = getImage(b, sourceCeilCR.y, sourceCeilCR.x, d);
vec2 fracCR = sourceFracIndexCR - vec2(sourceFloorCR);
float top = topLeft + (topRight - topLeft) * fracCR.x;
float bottom = bottomLeft + (bottomRight - bottomLeft) * fracCR.x;
float newValue = top + (bottom - top) * fracCR.y;
setOutput(newValue);
} else {
// Compute the coordinators of nearest neighbor point.
ivec2 sourceNearestCR = ivec2(floor(
sourceFracIndexCR + vec2(0.5,0.5)));
float newValue = getImage(b, sourceNearestCR.y, sourceNearestCR.x, d);
setOutput(newValue);
}
}
`}},xZ=e=>{let{inputs:t,backend:n,attrs:a}=e,{image:r,boxes:s,boxInd:i}=t,{cropSize:o,method:l,extrapolationValue:u}=a,p=new bZ(r.shape,s.shape,o,l,u);return n.runWebGLProgram(p,[r,s,i],"float32")},vZ={kernelName:Ql,backendName:"webgl",kernelFunc:xZ},mc;(function(e){e.Prod="*",e.Sum="+"})(mc||(mc={}));var sI=class{constructor(e,t,n,a){this.op=e,this.outputShape=t,this.variableNames=["x"],this.customUniforms=[{name:"index",type:"float"}];let r=this.outputShape.length,s=this.op===mc.Prod?"1.0":"0.0",i=n?s:`getX(${iI(r,"coords",this.op)})`,o=this.outputShape[this.outputShape.length-1],l="",u="";n?(l=a?`end != ${o-1}`:"end != 0",u=a?"end + 1":"end - 1"):(l=a?`end + pow2 < ${o}`:"end >= pow2",u=a?"end + pow2":"end - pow2"),this.userCode=`
void main() {
${mt(r)} coords = getOutputCoords();
int end = ${oI(r,"coords",this.op)};
float val = ${i};
int pow2 = int(pow(2.0, index));
if (${l}) {
int idx = ${u};
${oI(r,"coords",this.op)} = idx;
val ${this.op}= getX(${iI(r,"coords",this.op)});
}
setOutput(val);
}
`}};function iI(e,t,n){if(e===1)return`${t}`;if(e===2)return`${t}.x, ${t}.y`;if(e===3)return`${t}.x, ${t}.y, ${t}.z`;if(e===4)return`${t}.x, ${t}.y, ${t}.z, ${t}.w`;throw new Error(`Cumulative ${n} for rank ${e} is not yet supported`)}function oI(e,t,n){if(e===1)return`${t}`;if(e===2)return`${t}.y`;if(e===3)return`${t}.z`;if(e===4)return`${t}.w`;throw new Error(`Cumulative ${n} for rank ${e} is not yet supported`)}function HC(e,t,n,a,r,s){let i=t.shape.length,o=C.getAxesPermutation([a],i),l=t;o!=null&&(l=Wt({inputs:{x:t},backend:n,attrs:{perm:o}}));let u=C.getInnerMostAxes(1,i)[0];if(u!==i-1)throw new Error(`WebGL cumprod shader expects an inner-most axis=${t.shape.length-1} but got axis=${a}`);let p=l.shape[u],d=aa({inputs:{x:l},backend:n});for(let c=0;c<=Math.ceil(Math.log2(p))-1;c++){let h=new sI(e,l.shape,!1,s),m=[[c]],f=d;d=n.runWebGLProgram(h,[d],d.dtype,m),n.disposeIntermediateTensorInfo(f)}if(r){let c=new sI(e,l.shape,r,s),h=d;d=n.runWebGLProgram(c,[d],d.dtype),n.disposeIntermediateTensorInfo(h)}if(o!=null){let c=C.getUndoAxesPermutation(o),h=Wt({inputs:{x:d},backend:n,attrs:{perm:c}});return n.disposeIntermediateTensorInfo(d),n.disposeIntermediateTensorInfo(l),h}return d}function wZ(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{axis:s,exclusive:i,reverse:o}=a;return HC(mc.Prod,r,n,s,i,o)}var kZ={kernelName:Yl,backendName:"webgl",kernelFunc:wZ};function IZ(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{axis:s,exclusive:i,reverse:o}=a;return HC(mc.Sum,r,n,s,i,o)}var SZ={kernelName:Ai,backendName:"webgl",kernelFunc:IZ};function NZ(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,weights:s}=t,{size:i,binaryOutput:o}=a;if(r.shape.length===1){let l=n.readSync(r.dataId),u=n.readSync(s.dataId),p=SC(l,u,s.dtype,s.shape,i);return n.makeTensorInfo([i],s.dtype,p)}else if(r.shape.length===2){let l=n.bufferSync(r),u=n.bufferSync(s),p=OY(l,u,i,o);return n.makeTensorInfo(p.shape,s.dtype,p.values)}throw new Error(`Error in denseBincount: input must be at most rank 2, but got rank${r.shape.length}.`)}var TZ={kernelName:Fm,backendName:"webgl",kernelFunc:NZ},_Z=class{constructor(e,t,n){this.variableNames=["x"],this.outputShape=[],this.outputShape=e,this.blockSize=t,this.dataFormat=n,this.userCode=`
void main() {
ivec4 coords = getOutputCoords();
int b = coords[0];
int h = ${this.getHeightCoordString()};
int w = ${this.getWidthCoordString()};
int d = ${this.getDepthCoordString()};
int in_h = h / ${t};
int offset_h = imod(h, ${t});
int in_w = w / ${t};
int offset_w = imod(w, ${t});
int offset_d = (offset_h * ${t} + offset_w) *
${this.getOutputDepthSize()};
int in_d = d + offset_d;
float result = ${this.getInputSamplingString()};
setOutput(result);
}
`}getHeightCoordString(){return this.dataFormat==="NHWC"?"coords[1]":"coords[2]"}getWidthCoordString(){return this.dataFormat==="NHWC"?"coords[2]":"coords[3]"}getDepthCoordString(){return this.dataFormat==="NHWC"?"coords[3]":"coords[1]"}getOutputDepthSize(){return this.dataFormat==="NHWC"?this.outputShape[3]:this.outputShape[1]}getInputSamplingString(){return this.dataFormat==="NHWC"?"getX(b, in_h, in_w, in_d)":"getX(b, in_d, in_h, in_w)"}};function CZ(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{blockSize:s,dataFormat:i}=a,o=r.shape[0],l=i==="NHWC"?r.shape[1]:r.shape[2],u=i==="NHWC"?r.shape[2]:r.shape[3],p=i==="NHWC"?r.shape[3]:r.shape[1],d=l*s,c=u*s,h=p/(s*s),m=i==="NHWC"?[o,d,c,h]:[o,h,d,c],f=new _Z(m,s,i);return n.runWebGLProgram(f,[r],r.dtype)}var EZ={kernelName:Jl,backendName:"webgl",kernelFunc:CZ},jC=class{constructor(e,t=!1,n=null,a=!1,r=!1){this.variableNames=["x","W"],this.customUniforms=[{name:"pads",type:"ivec2"},{name:"strides",type:"ivec2"},{name:"dilations",type:"ivec2"},{name:"inDims",type:"ivec2"}],this.outputShape=e.outShape,this.enableShapeUniforms=Hn(this.outputShape.length);let s=e.filterHeight,i=e.filterWidth,o=e.outChannels/e.inChannels,l="",u="";n&&(a?l=`float activation(float a) {
float b = getPreluActivationWeightsAtOutCoords();
${n}
}`:r?l=`float activation(float a) {
float b = getLeakyreluAlphaAtOutCoords();
${n}
}`:l=`
float activation(float x) {
${n}
}
`,u="result = activation(result);");let p=t?"result += getBiasAtOutCoords();":"";t&&this.variableNames.push("bias"),a&&this.variableNames.push("preluActivationWeights"),r&&this.variableNames.push("leakyreluAlpha"),this.userCode=`
${l}
void main() {
ivec4 coords = getOutputCoords();
int batch = coords.x;
ivec2 xRCCorner = coords.yz * strides - pads;
int d2 = coords.w;
int d1 = d2 / ${o};
int q = d2 - d1 * ${o};
int xRCorner = xRCCorner.x;
int xCCorner = xRCCorner.y;
// Convolve x(?, ?, d1) with w(:, :, d1, q) to get y(yR, yC, d2).
// ? = to be determined. : = across all values in that axis.
float dotProd = 0.0;
// TO DO(dsmilkov): Flatten the two for loops and vec4 the operations.
for (int wR = 0; wR < ${s}; wR++) {
int xR = xRCorner + wR * dilations[0];
if (xR < 0 || xR >= inDims[0]) {
continue;
}
for (int wC = 0; wC < ${i}; wC++) {
int xC = xCCorner + wC * dilations[1];
if (xC < 0 || xC >= inDims[1]) {
continue;
}
float xVal = getX(batch, xR, xC, d1);
float wVal = getW(wR, wC, d1, q);
dotProd += xVal * wVal;
}
}
float result = dotProd;
${p}
${u}
setOutput(result);
}
`}},qC=class{constructor(e,t=!1,n=null,a=!1,r=!1){this.variableNames=["x","W"],this.packedInputs=!0,this.packedOutput=!0,this.customUniforms=[{name:"pads",type:"ivec2"},{name:"strides",type:"ivec2"},{name:"dilations",type:"ivec2"},{name:"inDims",type:"ivec2"}],this.outputShape=e.outShape,this.enableShapeUniforms=Hn(this.outputShape.length);let s=e.outChannels/e.inChannels,i=e.padInfo.left,o=e.strideWidth,l=e.dilationWidth,u=e.filterHeight,p=e.filterWidth,d=p,c=`
int xR; int xC; int xCOffset;
vec4 wTexel; vec4 previous; vec4 final;`;for(let g=0;g<p;g++)c+=`
vec4 xTexelC${g*2};
int xTexelC${g*2}Ready;
vec4 xTexelC${g*2+1};
int xTexelC${g*2+1}Ready;
vec4 xC${g};`;c+=`
for (int r = 0; r < ${u}; r++) {
`;for(let g=0;g<p;g++)c+=`
xTexelC${g*2} = vec4(0.0);
xTexelC${g*2}Ready = 0;
xTexelC${g*2+1} = vec4(0.0);
xTexelC${g*2+1}Ready = 0;
xC${g} = vec4(0.0);`;c+=`
xR = xRCorner + r * dilations[0];
if (xR >=0 && xR < inDims[0]) {
`;for(let g=0;g<(d+1)/2;g++){let y=g*2;if(c+=`
xC = xCCorner + ${y*l};
`,o===1){if(y<p&&(i%2===1?(c+=`
xCOffset = xC + 1;
if (xCOffset >= 0 && xCOffset < inDims[1] && xTexelC${y}Ready == 0) {
xTexelC${y} = getX(batch, xR, xCOffset, d1);
// Need to manually clear unused channels in case
// we're reading from recycled texture.
if (xCOffset + 1 >= inDims[1]) {
xTexelC${y}.zw = vec2(0.0);
}
xTexelC${y}Ready = 1;
}
`,l===1&&y>0?c+=`
xC${y} = vec4(xTexelC${y-2}.zw, xTexelC${y}.xy);
`:c+=`
xCOffset = xC + 1 - 2;
if (xCOffset >= 0 && xCOffset < inDims[1]) {
previous = getX(batch, xR, xCOffset, d1);
// Need to manually clear unused channels in case
// we're reading from recycled texture.
if (xCOffset + 1 >= inDims[1]) {
previous.zw = vec2(0.0);
}
xC${y} = vec4(previous.zw, xTexelC${y}.xy);
} else {
xC${y} = vec4(0.0, 0.0, xTexelC${y}.xy);
}
`):c+=`
if (xC >= 0 && xC < inDims[1] && xTexelC${y}Ready == 0) {
xTexelC${y} = getX(batch, xR, xC, d1);
if (xC + 1 >= inDims[1]) {
xTexelC${y}.zw = vec2(0.0);
}
xTexelC${y}Ready = 1;
}
xC${y} = xTexelC${y};
`,y+1<p)){let b=i%2===0?w.nearestLargerEven(l):l;l%2===0&&i%2===1||l%2!==0&&i%2!==1?(c+=`
xCOffset = xC + imod(pads[1], 2) + ${b};
if (xCOffset >= 0 && xCOffset < inDims[1] && xTexelC${y+1}Ready == 0) {
xTexelC${y+1} = getX(batch, xR, xCOffset, d1);
// Need to manually clear unused channels in case
// we're reading from recycled texture.
if (xCOffset + 1 >= inDims[1]) {
xTexelC${y+1}.zw = vec2(0.0);
}
xTexelC${y+1}Ready = 1;
}
`,l>1&&(c+=`
xCOffset -= 2;
if (xCOffset >= 0 && xCOffset < inDims[1] && xTexelC${y}Ready == 0) {
xTexelC${y} = getX(batch, xR, xCOffset, d1);
xTexelC${y}Ready = 1;
}
`),c+=`
xC${y+1} = vec4(xTexelC${y}.zw, xTexelC${y+1}.xy);
`):b===1?c+=`
xC${y+1} = xTexelC${y};
`:c+=`
xCOffset = xC + ${b};
if (xCOffset >= 0 && xCOffset < inDims[1] && xTexelC${y+1}Ready == 0) {
xTexelC${y+1} = getX(batch, xR, xCOffset, d1);
if (xCOffset + 1 >= inDims[1]) {
xTexelC${y+1}.zw = vec2(0.0);
}
xTexelC${y+1}Ready = 1;
}
xC${y+1} = xTexelC${y+1};
`}}else y<p&&(i%2===1?(c+=`
xCOffset = xC + 1 - strides[1];
if(xCOffset >= 0 && xCOffset < inDims[1] && xTexelC${y}Ready == 0) {
xTexelC${y} = getX(batch, xR, xCOffset, d1);
// Need to manually clear unused channels in case
// we're reading from recycled texture.
if (xCOffset + 1 >= inDims[1]) {
xTexelC${y}.zw = vec2(0.0);
}
xTexelC${y}Ready = 1;
}
if(xC + 1 >= 0 && xC + 1 < inDims[1] && xTexelC${y+1}Ready == 0) {
xTexelC${y+1} = getX(batch, xR, xC + 1, d1);
// Need to manually clear unused channels in case
// we're reading from recycled texture.
if (xC + 2 >= inDims[1]) {
xTexelC${y+1}.zw = vec2(0.0);
}
xTexelC${y+1}Ready = 1;
}
xC${y} = vec4(xTexelC${y}.zw, xTexelC${y+1}.zw);
`,y+1<p&&(c+=`
final = vec4(0.0);
xCOffset = xC + 1 + strides[1];
if(xCOffset >= 0 && xCOffset < inDims[1]) {
final = getX(batch, xR, xCOffset, d1);
}
xC${y+1} = vec4(xTexelC${y+1}.xy, final.xy);
`)):(c+=`
if(xC >= 0 && xC < inDims[1] && xTexelC${y}Ready == 0) {
xTexelC${y} = getX(batch, xR, xC, d1);
if (xC + 1 >= inDims[1]) {
xTexelC${y}.zw = vec2(0.0);
}
xTexelC${y}Ready = 1;
}
xCOffset = xC + strides[1];
if(xCOffset >= 0 && xCOffset < inDims[1] && xTexelC${y+1}Ready == 0) {
xTexelC${y+1} = getX(batch, xR, xCOffset, d1);
if (xCOffset + 1 >= inDims[1]) {
xTexelC${y+1}.zw = vec2(0.);
}
xTexelC${y+1}Ready = 1;
}
xC${y} = vec4(
xTexelC${y}.xy, xTexelC${y+1}.xy);
`,y+1<p&&(c+=`
xC${y+1} = vec4(xTexelC${y}.zw, xTexelC${y+1}.zw);
`)));y<p&&(c+=`
wTexel = getW(r, ${y}, d1, q);
dotProd += xC${y} * vec4(wTexel.xz, wTexel.xz);
`,y+1<p&&(c+=`
wTexel = getW(r, ${y+1}, d1, q);
dotProd += xC${y+1} * vec4(wTexel.xz, wTexel.xz);
`))}c+=`
}
`,c+=`
}
`;let h="",m="";n&&(a?h=`vec4 activation(vec4 a) {
vec4 b = getPreluActivationWeightsAtOutCoords();
${n}
}`:r?h=`vec4 activation(vec4 a) {
vec4 b = getLeakyreluAlphaAtOutCoords();
${n}
}`:h=`vec4 activation(vec4 x) {
${n}
}`,m="result = activation(result);");let f=t?"result += getBiasAtOutCoords();":"";t&&this.variableNames.push("bias"),a&&this.variableNames.push("preluActivationWeights"),r&&this.variableNames.push("leakyreluAlpha"),this.userCode=`
${h}
void main() {
ivec4 coords = getOutputCoords();
int batch = coords.x;
ivec2 xRCCorner = coords.yz * strides - pads;
int d2 = coords.w;
int d1 = d2 / ${s};
int q = d2 - d1 * ${s};
int xRCorner = xRCCorner.x;
int xCCorner = xRCCorner.y;
//intialize dotProd with a small epsilon seems to reduce GPU accuracy loss.
vec4 dotProd = vec4(0.000000000000001);
${c}
vec4 result = dotProd - vec4(0.000000000000001);
${f}
${m}
setOutput(result);
}
`}};function FZ(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,filter:s}=t,{strides:i,pad:o,dilations:l,dimRoundingMode:u}=a,p=l;p==null&&(p=[1,1]),w.assert(C.eitherStridesOrDilationsAreOne(i,p),()=>`Error in depthwiseConv2d: Either strides or dilations must be 1. Got strides ${i} and dilations '${p}'`);let d=C.computeConv2DInfo(r.shape,s.shape,i,p,o,u,!0),c;X().getBool("WEBGL_PACK_DEPTHWISECONV")&&d.strideWidth<=2&&d.outChannels/d.inChannels===1?c=new qC(d):c=new jC(d);let h=[[d.padInfo.top,d.padInfo.left],[d.strideHeight,d.strideWidth],[d.dilationHeight,d.dilationWidth],[d.inHeight,d.inWidth]];return n.runWebGLProgram(c,[r,s],"float32",h)}var AZ={kernelName:$i,backendName:"webgl",kernelFunc:FZ},$Z=class{constructor(e){this.variableNames=["x","dy"],this.outputShape=e.filterShape;let t=e.strideHeight,n=e.strideWidth,a=e.padInfo.top,r=e.padInfo.left,s=e.outChannels/e.inChannels;this.userCode=`
void main() {
ivec4 coords = getOutputCoords();
int wR = coords.x;
int wC = coords.y;
int d1 = coords.z;
int dm = coords.w;
int d2 = d1 * ${s} + dm;
float dotProd = 0.0;
// TO DO: Vec4 over the batch size
for (int b = 0; b < ${e.batchSize}; b++) {
for (int yR = 0; yR < ${e.outHeight}; yR++) {
int xR = wR + yR * ${t} - ${a};
if (xR < 0 || xR >= ${e.inHeight}) {
continue;
}
for (int yC = 0; yC < ${e.outWidth}; yC++) {
int xC = wC + yC * ${n} - ${r};
if (xC < 0 || xC >= ${e.inWidth}) {
continue;
}
float dyValue = getDy(b, yR, yC, d2);
float xValue = getX(b, xR, xC, d1);
dotProd += (xValue * dyValue);
}
}
}
setOutput(dotProd);
}
`}},DZ=class{constructor(e){this.variableNames=["dy","W"],this.outputShape=e.inShape;let t=e.filterHeight,n=e.filterWidth,a=e.strideHeight,r=e.strideWidth,s=t-1-e.padInfo.top,i=n-1-e.padInfo.left,o=e.outChannels/e.inChannels;this.userCode=`
const ivec2 pads = ivec2(${s}, ${i});
void main() {
ivec4 coords = getOutputCoords();
int batch = coords[0];
int d1 = coords[3];
ivec2 dyCorner = coords.yz - pads;
int dyRCorner = dyCorner.x;
int dyCCorner = dyCorner.y;
float dotProd = 0.0;
for (int wR = 0; wR < ${t}; wR++) {
float dyR = float(dyRCorner + wR) / ${a}.0;
if (dyR < 0.0 || dyR >= ${e.outHeight}.0 || fract(dyR) > 0.0) {
continue;
}
int idyR = int(dyR);
int wRPerm = ${t} - 1 - wR;
for (int wC = 0; wC < ${n}; wC++) {
float dyC = float(dyCCorner + wC) / ${r}.0;
if (dyC < 0.0 || dyC >= ${e.outWidth}.0 ||
fract(dyC) > 0.0) {
continue;
}
int idyC = int(dyC);
int wCPerm = ${n} - 1 - wC;
// TO DO: Vec4 over the channelMul
for (int dm = 0; dm < ${o}; dm++) {
int d2 = d1 * ${o} + dm;
float xValue = getDy(batch, idyR, idyC, d2);
float wValue = getW(wRPerm, wCPerm, d1, dm);
dotProd += xValue * wValue;
}
}
}
setOutput(dotProd);
}
`}};function RZ(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,dy:s}=t,{strides:i,dilations:o,pad:l,dimRoundingMode:u,filterShape:p}=a,d=C.computeConv2DInfo(r.shape,p,i,o,l,u,!0),c=new $Z(d);return n.runWebGLProgram(c,[r,s],"float32")}var MZ={kernelName:Am,backendName:"webgl",kernelFunc:RZ};function PZ(e){let{inputs:t,backend:n,attrs:a}=e,{dy:r,filter:s}=t,{strides:i,dilations:o,pad:l,dimRoundingMode:u,inputShape:p}=a,d=C.computeConv2DInfo(p,s.shape,i,o,l,u,!0),c=new DZ(d);return n.runWebGLProgram(c,[r,s],"float32")}var OZ={kernelName:$m,backendName:"webgl",kernelFunc:PZ},LZ=class{constructor(e){this.variableNames=["X"],this.outputShape=[e,e],this.userCode=`
void main() {
ivec2 coords = getOutputCoords();
float val = coords[0] == coords[1] ? getX(coords[0]) : 0.0;
setOutput(val);
}
`}};function zZ(e){let{inputs:t,backend:n}=e,{x:a}=t,r=[...a.shape,...a.shape],s=w.sizeFromShape(a.shape),i=me({inputs:{x:a},backend:n,attrs:{shape:[s]}}),o=new LZ(s),l=n.runWebGLProgram(o,[i],i.dtype),u=me({inputs:{x:l},backend:n,attrs:{shape:r}});return n.disposeIntermediateTensorInfo(i),n.disposeIntermediateTensorInfo(l),u}var BZ={kernelName:Dm,backendName:"webgl",kernelFunc:zZ},WZ=class{constructor(e){this.variableNames=["x","W"],this.outputShape=e.outShape;let{inHeight:t,inWidth:n,padInfo:a,strideHeight:r,strideWidth:s,filterHeight:i,filterWidth:o,dilationHeight:l,dilationWidth:u}=e,{top:p,left:d}=a;this.userCode=`
const ivec2 strides = ivec2(${r}, ${s});
const ivec2 pads = ivec2(${p}, ${d});
const float neg_infinity = -3.4e38;
void main() {
ivec4 coords = getOutputCoords();
int batch = coords.x;
int d1 = coords.w;
ivec2 outTopLeftCorner =
coords.yz * strides - pads;
int hBeg = outTopLeftCorner.x;
int wBeg = outTopLeftCorner.y;
float curVal = neg_infinity;
for (int h = 0; h < ${i}; h++) {
int hIn = hBeg + h * ${l};
if (hIn >= 0 && hIn < ${t}) {
for (int w = 0; w < ${o}; w++) {
int wIn = wBeg + w * ${u};
if (wIn >= 0 && wIn < ${n}) {
float xVal = getX(batch, hIn, wIn, d1);
float wVal = getW(h, w, d1);
float val = xVal + wVal;
if (val > curVal) {
curVal = val;
}
}
}
}
}
float result = curVal;
setOutput(result);
}
`}};function VZ(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,filter:s}=t,{strides:i,pad:o,dilations:l}=a,u=C.computeDilation2DInfo(r.shape,s.shape,i,o,"NHWC",l),p,d=new WZ(u);p=n.runWebGLProgram(d,[r,s],"float32");let c=me({inputs:{x:p},backend:n,attrs:{shape:u.outShape}});return n.disposeIntermediateTensorInfo(p),c}var UZ={kernelName:wc,backendName:"webgl",kernelFunc:VZ};function GZ(e){let{inputs:t,backend:n,attrs:a}=e,{equation:r}=a,s=t,{allDims:i,summedDims:o,idDims:l}=C.decodeEinsumEquation(r,s.length);C.checkEinsumDimSizes(i.length,l,s);let{path:u,steps:p}=C.getEinsumComputePath(o,l),d=p.length,c=null,h=i.length,m=[];for(let f=0;f<d;++f){for(let g of p[f]){let{permutationIndices:y,expandDims:b}=C.getEinsumPermutation(h,l[g]),x;C.isIdentityPermutation(y)?x=s[g]:(x=Wt({inputs:{x:s[g]},backend:n,attrs:{perm:y}}),m.push(x));let v=x.shape.slice();for(let k=0;k<b.length;++k)v.splice(b[k],0,1);w.arraysEqual(x.shape,v)||(x=me({inputs:{x},backend:n,attrs:{shape:v}}),m.push(x)),c===null?c=x:(c=V0({inputs:{a:x,b:c},backend:n}),m.push(c))}f<d-1&&(u[f]>=0&&(c=ug({inputs:{x:c},backend:n,attrs:{axis:u[f]-(i.length-h),keepDims:!1}}),m.push(c)),h--)}for(let f of m)f!==c&&n.disposeIntermediateTensorInfo(f);return c}var HZ={kernelName:Rm,backendName:"webgl",kernelFunc:GZ},jZ="return (x >= 0.0) ? x : (exp(x) - 1.0);",qZ=`
vec4 result;
result.r = (x.r >= 0.0) ? x.r : (exp(x.r) - 1.0);
result.g = (x.g >= 0.0) ? x.g : (exp(x.g) - 1.0);
result.b = (x.b >= 0.0) ? x.b : (exp(x.b) - 1.0);
result.a = (x.a >= 0.0) ? x.a : (exp(x.a) - 1.0);
return result;
`,KZ=Qe({opSnippet:jZ,packedOpSnippet:qZ}),XZ={kernelName:Ri,backendName:"webgl",kernelFunc:KZ},YZ="return (b >= 1.0) ? a : a * (b + 1.0);",QZ=`
vec4 bGTEZero = vec4(greaterThanEqual(b, vec4(0.)));
return (bGTEZero * a) + ((vec4(1.0) - bGTEZero) * (a * (b + vec4(1.0))));
`,JZ=e=>{let{inputs:t,backend:n}=e,{dy:a,y:r}=t,s=X().getBool("WEBGL_PACK_BINARY_OPERATIONS")?new dd(QZ,a.shape,r.shape):new Pl(YZ,a.shape,r.shape);return n.runWebGLProgram(s,[a,r],a.dtype)},ZZ={kernelName:Mm,backendName:"webgl",kernelFunc:JZ},eee=`
return vec4(equal(a, b));
`,tee="return float(a == b);",nee=cn({opSnippet:tee,packedOpSnippet:eee,dtype:"bool",cpuKernelImpl:BY}),aee={kernelName:eu,backendName:"webgl",kernelFunc:nee},ree=`
// Error function is calculated approximately with elementary function.
// See "Handbook of Mathematical Functions with Formulas,
// Graphs, and Mathematical Tables", Abramowitz and Stegun.
float p = ${C.ERF_P};
float a1 = ${C.ERF_A1};
float a2 = ${C.ERF_A2};
float a3 = ${C.ERF_A3};
float a4 = ${C.ERF_A4};
float a5 = ${C.ERF_A5};
float sign = sign(x);
x = abs(x);
float t = 1.0 / (1.0 + p * x);
return sign * (1.0 - (((((a5*t + a4)*t) + a3)*t + a2)*t + a1)*t*exp(-x*x));
`,see=Qe({opSnippet:ree}),iee={kernelName:Zl,backendName:"webgl",kernelFunc:see},oee=rp+`
return exp(x);
`,lee=`
vec4 result = exp(x);
bvec4 isNaN = isnan(x);
result.r = isNaN.r ? x.r : result.r;
result.g = isNaN.g ? x.g : result.g;
result.b = isNaN.b ? x.b : result.b;
result.a = isNaN.a ? x.a : result.a;
return result;
`,KC=Qe({opSnippet:oee,packedOpSnippet:lee,cpuKernelImpl:WY,dtype:"float32"}),uee={kernelName:Mi,backendName:"webgl",kernelFunc:KC};function Ix(e){let{inputs:t,attrs:n,backend:a}=e,{dim:r}=n,{input:s}=t,i=s.shape.length,o=s.shape.slice(),l=r;return r<0&&(w.assert(-(i+1)<=r,()=>`Axis must be in the interval [${-(i+1)}, ${i}]`),l=i+r+1),o.splice(l,0,1),me({inputs:{x:s},backend:a,attrs:{shape:o}})}var pee={kernelName:tu,backendName:"webgl",kernelFunc:Ix},lI="return exp(x) - 1.0;",cee=Qe({opSnippet:lI,packedOpSnippet:lI,cpuKernelImpl:VY}),dee={kernelName:nu,backendName:"webgl",kernelFunc:cee},uI=class{constructor(e,t,n){this.variableNames=["real","imag"];let a=t[1];this.outputShape=t;let r=n?`2.0 * ${Math.PI}`:`-2.0 * ${Math.PI}`,s=n?`${a}.0`:"1.0",i;if(e==="real")i="return real * expR - imag * expI;";else if(e==="imag")i="return real * expI + imag * expR;";else throw new Error(`FFT component must be either "real" or "imag", got ${e}.`);this.userCode=`
const float exponentMultiplier = ${r};
float unaryOpComplex(float real, float expR, float imag, float expI) {
${i}
}
float mulMatDFT(int batch, int index) {
float indexRatio = float(index) / float(${a});
float exponentMultiplierTimesIndexRatio =
exponentMultiplier * indexRatio;
float result = 0.0;
for (int i = 0; i < ${a}; i++) {
// x = (-2|2 * PI / N) * index * i;
float x = exponentMultiplierTimesIndexRatio * float(i);
float expR = cos(x);
float expI = sin(x);
float real = getReal(batch, i);
float imag = getImag(batch, i);
result +=
unaryOpComplex(real, expR, imag, expI) / ${s};
}
return result;
}
void main() {
ivec2 coords = getOutputCoords();
setOutput(mulMatDFT(coords[0], coords[1]));
}
`}};function XC(e,t,n){let a=n.texData.get(e.dataId),r=w.sizeFromShape(e.shape),s=e.shape[e.shape.length-1],i=r/s,o=me({inputs:{x:e},backend:n,attrs:{shape:[i,s]}}),l=o.shape,u=new uI("real",l,t),p=new uI("imag",l,t),d=[{dataId:a.complexTensorInfos.real.dataId,dtype:a.complexTensorInfos.real.dtype,shape:l},{dataId:a.complexTensorInfos.imag.dataId,dtype:a.complexTensorInfos.imag.dtype,shape:l}],c=n.runWebGLProgram(u,d,"float32"),h=n.runWebGLProgram(p,d,"float32"),m=Cs({inputs:{real:c,imag:h},backend:n});n.disposeIntermediateTensorInfo(c),n.disposeIntermediateTensorInfo(h);let f=me({inputs:{x:m},backend:n,attrs:{shape:e.shape}});return n.disposeIntermediateTensorInfo(o),n.disposeIntermediateTensorInfo(m),f}function hee(e){let{inputs:t,backend:n}=e,{input:a}=t;return XC(a,!1,n)}var mee={kernelName:Pm,backendName:"webgl",kernelFunc:hee},fee=class{constructor(e,t){this.outputShape=[],this.customUniforms=[{name:"value",type:"float"}],this.variableNames=["x"],this.outputShape=e,this.userCode=`
void main() {
// Input can be obtained from uniform value.
setOutput(value);
}
`}};function md(e){let{backend:t,attrs:n}=e,{shape:a,value:r}=n,{dtype:s}=n;if(s=s||w.inferDtype(r),s==="string"){let i=w.getArrayFromDType(s,w.sizeFromShape(a));return i.fill(r),t.makeTensorInfo(a,s,i)}else{let i=new fee(a,r),o=[[r]];return t.runWebGLProgram(i,[],s,o)}}var gee={kernelName:kc,backendName:"webgl",kernelFunc:md},yee=class{constructor(e){this.variableNames=["Image"],this.outputShape=[];let t=e[2];this.outputShape=e,this.userCode=`
void main() {
ivec4 coords = getOutputCoords();
int x = coords[2];
int coordX = ${t} - x - 1;
float outputValue;
if(coordX >= 0 && coordX < ${t}) {
outputValue = getImage(coords[0], coords[1], coordX, coords[3]);
} else {
outputValue = getImage(coords[0], coords[1], coords[2], coords[3]);
}
setOutput(outputValue);
}
`}},bee={kernelName:au,backendName:"webgl",kernelFunc:({inputs:e,backend:t})=>{let{image:n}=e,a=t,r=new yee(n.shape);return a.runWebGLProgram(r,[n],n.dtype)}},pI="return floor(x);",xee=Qe({opSnippet:pI,packedOpSnippet:pI,cpuKernelImpl:UY}),vee={kernelName:Pi,backendName:"webgl",kernelFunc:xee},wee=`
float s = sign(a) * sign(b);
int ia = round(a);
int ib = round(b);
if (ib != 0) {
// Windows (D3D) wants guaranteed non-zero int division at compile-time.
return float(idiv(ia, ib, s));
} else {
return NAN;
}
`,kee=`
ivec4 ia = round(a);
ivec4 ib = round(b);
bvec4 cond = notEqual(ib, ivec4(0));
ivec4 result = ivec4(0);
vec4 s = sign(a) * sign(b);
// Windows (D3D) wants guaranteed non-zero int division at compile-time.
if (cond[0]) {
result[0] = idiv(ia[0], ib[0], s[0]);
}
if (cond[1]) {
result[1] = idiv(ia[1], ib[1], s[1]);
}
if (cond[2]) {
result[2] = idiv(ia[2], ib[2], s[2]);
}
if (cond[3]) {
result[3] = idiv(ia[3], ib[3], s[3]);
}
return vec4(result);
`,Iee=cn({opSnippet:wee,packedOpSnippet:kee,dtype:"int32"}),See={kernelName:Oi,backendName:"webgl",kernelFunc:Iee},Nee=class{constructor(e){this.variableNames=["A"];let t=En(),[n,a]=e;this.outputShape=e,this.userCode=`
void main() {
ivec3 coords = getOutputCoords();
int texR = coords[0];
int texC = coords[1];
int depth = coords[2];
vec2 uv = (vec2(texC, texR) + halfCR) / vec2(${a}.0, ${n}.0);
vec4 values = ${t.texture2D}(A, uv);
float value;
if (depth == 0) {
value = values.r;
} else if (depth == 1) {
value = values.g;
} else if (depth == 2) {
value = values.b;
} else if (depth == 3) {
value = values.a;
}
setOutput(floor(value * 255.0 + 0.5));
}
`}},Tee=class{constructor(e){this.variableNames=["A"],this.packedInputs=!1,this.packedOutput=!0;let t=En(),[n,a]=e;this.outputShape=e,this.userCode=`
void main() {
ivec3 coords = getOutputCoords();
int texR = coords[0];
int texC = coords[1];
int depth = coords[2];
vec4 result = vec4(0.);
for(int row=0; row<=1; row++) {
for(int col=0; col<=1; col++) {
texC = coords[1] + row;
depth = coords[2] + col;
vec2 uv = (vec2(texC, texR) + halfCR) /
vec2(${a}.0, ${n}.0);
vec4 values = ${t.texture2D}(A, uv);
float value;
if (depth == 0) {
value = values.r;
} else if (depth == 1) {
value = values.g;
} else if (depth == 2) {
value = values.b;
} else if (depth == 3) {
value = values.a;
}
result[row * 2 + col] = floor(value * 255.0 + 0.5);
}
}
${t.output} = result;
}
`}},_ee={kernelName:jh,backendName:"webgl",kernelFunc:Cee},cl;function Cee(e){let{inputs:t,backend:n,attrs:a}=e,{pixels:r}=t,{numChannels:s}=a,i=typeof HTMLVideoElement!="undefined"&&r instanceof HTMLVideoElement,o=typeof HTMLImageElement!="undefined"&&r instanceof HTMLImageElement,[l,u]=i?[r.videoWidth,r.videoHeight]:[r.width,r.height],p=[u,l],d=[u,l,s];(o||i)&&(cl==null&&(cl=document.createElement("canvas").getContext("2d")),cl.canvas.width=l,cl.canvas.height=u,cl.drawImage(r,0,0,l,u),r=cl.canvas);let c=n.makeTensorInfo(p,"int32");n.texData.get(c.dataId).usage=da.PIXELS,n.gpgpu.uploadPixelDataToTexture(n.getTexture(c.dataId),r);let h=X().getBool("WEBGL_PACK")?new Tee(d):new Nee(d),m=n.runWebGLProgram(h,[c],"int32");return n.disposeData(c.dataId),m}function Eee(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,filter:s,bias:i,preluActivationWeights:o}=t,{strides:l,pad:u,dataFormat:p,dilations:d,dimRoundingMode:c,activation:h,leakyreluAlpha:m}=a,f=C.convertConv2DDataFormat(p),g=C.computeConv2DInfo(r.shape,s.shape,l,d,u,c,!1,f),y,b=[];if(g.filterHeight===1&&g.filterWidth===1&&g.dilationHeight===1&&g.dilationWidth===1&&g.strideHeight===1&&g.strideWidth===1&&(g.padInfo.type==="SAME"||g.padInfo.type==="VALID"))y=UC({x:r,filter:s,convInfo:g,backend:n,bias:i,activation:h,preluActivationWeights:o,leakyreluAlpha:m});else if(X().getBool("WEBGL_CONV_IM2COL")&&r.shape[0]===1)y=GC({x:r,filter:s,convInfo:g,backend:n,bias:i,activation:h,preluActivationWeights:o,leakyreluAlpha:m});else{let v=i!=null,k=o!=null,T=h==="leakyrelu",_=h?og(h,!1):null,E=new VC(g,v,_,k,T),A=[r,s],M=($,S)=>{if(S==="NCHW"&&$.shape.length===1&&$.shape[0]!==1){let P=me({inputs:{x:$},backend:n,attrs:{shape:[$.shape[0],1,1]}});return b.push(P),P}return $};if(v&&A.push(M(i,p)),k&&A.push(M(o,p)),T){let $=n.makeTensorInfo([],"float32",w.createScalarValue(m,"float32"));A.push($),b.push($)}y=n.runWebGLProgram(E,A,"float32")}let x=me({inputs:{x:y},backend:n,attrs:{shape:g.outShape}});return b.push(y),b.forEach(v=>n.disposeIntermediateTensorInfo(v)),x}var Fee={kernelName:ri,backendName:"webgl",kernelFunc:Eee};function Aee(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,filter:s,bias:i,preluActivationWeights:o}=t,{strides:l,pad:u,dilations:p,dimRoundingMode:d,activation:c,leakyreluAlpha:h}=a,m=[],f=p;f==null&&(f=[1,1]),w.assert(C.eitherStridesOrDilationsAreOne(l,f),()=>`Error in depthwiseConv2d: Either strides or dilations must be 1. Got strides ${l} and dilations '${f}'`);let g=C.computeConv2DInfo(r.shape,s.shape,l,f,u,d,!0),y=X().getBool("WEBGL_PACK_DEPTHWISECONV")&&g.strideWidth<=2&&g.outChannels/g.inChannels===1,b=c?og(c,y):null,x=[r,s],v=i!=null,k=o!=null,T=c==="leakyrelu";if(v&&x.push(i),k&&x.push(o),T){let M=n.makeTensorInfo([],"float32",w.createScalarValue(h,"float32"));x.push(M),m.push(M)}let _;y?_=new qC(g,v,b,k,T):_=new jC(g,v,b,k,T);let E=[[g.padInfo.top,g.padInfo.left],[g.strideHeight,g.strideWidth],[g.dilationHeight,g.dilationWidth],[g.inHeight,g.inWidth]],A=n.runWebGLProgram(_,x,"float32",E);return m.forEach(M=>n.disposeIntermediateTensorInfo(M)),A}var $ee={kernelName:si,backendName:"webgl",kernelFunc:Aee},Dee=class{constructor(e,t,n){this.sliceDim=e,this.strides=t,this.variableNames=["x","indices"],this.outputShape=n;let a=mt(t.length),r=mt(n.length),s=this.sliceDim>1?"strides[j]":"strides";this.userCode=`
${a} strides = ${a}(${this.strides});
void main() {
${r} coords = getOutputCoords();
int flattenIndex = 0;
for (int j = 0; j < ${this.sliceDim}; j++) {
int index = round(getIndices(coords[0], j));
flattenIndex += index * ${s};
}
setOutput(getX(flattenIndex, coords[1]));
}
`}};function Ree(e){let{inputs:t,backend:n}=e,{params:a,indices:r}=t,s=r.shape,i=s[s.length-1],o=w.sizeFromShape(a.shape),[l,u,p,d]=C.prepareAndValidate(a,r),c=me({inputs:{x:r},backend:n,attrs:{shape:[u,i]}}),h=me({inputs:{x:a},backend:n,attrs:{shape:[w.sizeFromShape(a.shape)/p,p]}});if(n.shouldExecuteOnCPU([a,r])||a.dtype==="string"){let y=n.readSync(r.dataId),b=n.bufferSync(a),x=GY(y,b,a.dtype,u,i,p,d,a.shape,o);return n.makeTensorInfo(l,a.dtype,x.values)}let m=new Dee(i,d,[u,p]),f=n.runWebGLProgram(m,[h,c],h.dtype),g=me({inputs:{x:f},backend:n,attrs:{shape:l}});return n.disposeIntermediateTensorInfo(c),n.disposeIntermediateTensorInfo(h),n.disposeIntermediateTensorInfo(f),g}var Mee={kernelName:su,backendName:"webgl",kernelFunc:Ree},Pee=class{constructor(e,t){this.variableNames=["A","indices"],this.outputShape=t,this.rank=t.length;let n=mt(this.rank),a=Oee(e,2);this.userCode=`
void main() {
${n} resRC = getOutputCoords();
int index = int(getIndices(resRC.x, resRC.z));
float inBounds = (index >= 0) && (index < ${e[2]}) ? 1.0 : 0.0;
setOutput(inBounds * getA(${a}));
}
`}};function Oee(e,t){let n=["resRC.x","resRC.y","resRC.z","resRC.w"],a=[];for(let r=0;r<e.length;r++)r===2?a.push("index"):a.push(`${n[r]}`);return a.join()}function YC(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,indices:s}=t,{axis:i,batchDims:o}=a,l=w.parseAxisParam(i,r.shape)[0];if(X().get("DEBUG")){let b=n.readSync(s.dataId),x=r.shape[l];for(let v=0;v<b.length;++v){let k=b[v];w.assert(k<=x-1&&k>=0,()=>`GatherV2: the index value ${k} is not in [0, ${x-1}]`)}}let u=C.segment_util.collectGatherOpShapeInfo(r,s,l,o),p=w.sizeFromShape(s.shape),d=[],c=me({inputs:{x:r},backend:n,attrs:{shape:[u.batchSize,u.outerSize,u.dimSize,u.sliceSize]}}),h=me({inputs:{x:s},backend:n,attrs:{shape:[u.batchSize,p/u.batchSize]}});d.push(c),d.push(h);let m=[u.batchSize,u.outerSize,p/u.batchSize,u.sliceSize];if(n.shouldExecuteOnCPU([r,s])||r.dtype==="string"){let b=n.bufferSync(h),x=n.bufferSync(c),v=HY(x,b,m);return d.forEach(k=>n.disposeIntermediateTensorInfo(k)),n.makeTensorInfo(u.outputShape,v.dtype,v.values)}let f=new Pee(c.shape,m),g=n.runWebGLProgram(f,[c,h],c.dtype);d.push(g);let y=me({inputs:{x:g},backend:n,attrs:{shape:u.outputShape}});return d.forEach(b=>n.disposeIntermediateTensorInfo(b)),y}var Lee={kernelName:ru,backendName:"webgl",kernelFunc:YC},zee="return float(a > b);",Bee=`
return vec4(greaterThan(a, b));
`,Wee=cn({opSnippet:zee,packedOpSnippet:Bee,cpuKernelImpl:jY,dtype:"bool"}),Vee={kernelName:iu,backendName:"webgl",kernelFunc:Wee},Uee="return float(a >= b);",Gee=`
return vec4(greaterThanEqual(a, b));
`,Hee=cn({opSnippet:Uee,packedOpSnippet:Gee,dtype:"bool",cpuKernelImpl:qY}),jee={kernelName:zi,backendName:"webgl",kernelFunc:Hee};function qee(e){let{inputs:t,backend:n}=e,{input:a}=t;return XC(a,!0,n)}var Kee={kernelName:Om,backendName:"webgl",kernelFunc:qee},Xee="return float(!isnan(x) && !isinf(x));",Yee=Qe({opSnippet:Xee,dtype:"bool"}),Qee={kernelName:ou,backendName:"webgl",kernelFunc:Yee},Jee="return float(isinf(x));",Zee=Qe({opSnippet:Jee,dtype:"bool"}),ete={kernelName:lu,backendName:"webgl",kernelFunc:Zee},tte="return float(isnan(x));",nte=Qe({opSnippet:tte,dtype:"bool"}),ate={kernelName:uu,backendName:"webgl",kernelFunc:nte},rte="return float(a < b);",ste=`
return vec4(lessThan(a, b));
`,ite=cn({opSnippet:rte,packedOpSnippet:ste,cpuKernelImpl:KY,dtype:"bool"}),ote={kernelName:pu,backendName:"webgl",kernelFunc:ite},lte="return float(a <= b);",ute=`
return vec4(lessThanEqual(a, b));
`,pte=cn({opSnippet:lte,packedOpSnippet:ute,cpuKernelImpl:XY,dtype:"bool"}),cte={kernelName:cu,backendName:"webgl",kernelFunc:pte};function dte(e){let{backend:t,attrs:n}=e,{start:a,stop:r,num:s}=n,i=YY(a,r,s);return t.makeTensorInfo([i.length],"float32",i)}var hte={kernelName:zm,backendName:"webgl",kernelFunc:dte},mte=rp+`
return x < 0.0 ? 0./0. : log(x);
`,fte=`
vec4 result = log(x);
bvec4 isNaN = isnan(x);
result.r = isNaN.r ? x.r : (x.r < 0.0 ? 0./0. : result.r);
result.g = isNaN.g ? x.g : (x.g < 0.0 ? 0./0. : result.g);
result.b = isNaN.b ? x.b : (x.b < 0.0 ? 0./0. : result.b);
result.a = isNaN.a ? x.a : (x.a < 0.0 ? 0./0. : result.a);
return result;
`,gte=Qe({opSnippet:mte,packedOpSnippet:fte,cpuKernelImpl:QY}),yte={kernelName:Vi,backendName:"webgl",kernelFunc:gte},bte=rp+`
return log(1.0 + x);
`,xte=Qe({opSnippet:bte}),vte={kernelName:du,backendName:"webgl",kernelFunc:xte},wte="return float(a >= 1.0 && b >= 1.0);",kte=`
return vec4(
vec4(greaterThanEqual(a, vec4(1.0))) *
vec4(greaterThanEqual(b, vec4(1.0))));
`,Ite=cn({opSnippet:wte,packedOpSnippet:kte,dtype:"bool"}),Ste={kernelName:hu,backendName:"webgl",kernelFunc:Ite},Nte="return float(!(x >= 1.0));",Tte=Qe({opSnippet:Nte}),_te={kernelName:Ic,backendName:"webgl",kernelFunc:Tte},Cte="return float(a >= 1.0 || b >= 1.0);",Ete=`
return min(
vec4(greaterThanEqual(a, vec4(1.0))) +
vec4(greaterThanEqual(b, vec4(1.0))),
vec4(1.0));
`,Fte=cn({opSnippet:Cte,packedOpSnippet:Ete,dtype:"bool"}),Ate={kernelName:Sc,backendName:"webgl",kernelFunc:Fte},$te=class{constructor(e,t,n,a,r){this.variableNames=["x"],this.outputShape=[];let s=t,i=e[3]-1;this.outputShape=e;let o,l=`float(${n}) + float(${a}) * sum`;r===.5?o=`inversesqrt(${l})`:r===1?o=`1.0/(${l})`:o=`exp(log(${l}) * float(-${r}));`,this.userCode=`
void main() {
ivec4 coords = getOutputCoords();
int b = coords[0];
int r = coords[1];
int c = coords[2];
int d = coords[3];
float x = getX(b, r, c, d);
float sum = 0.0;
for (int j = -${s}; j <= ${s}; j++) {
int idx = d + j;
if (idx >= 0 && idx <= ${i}) {
float z = getX(b, r, c, idx);
sum += z * z;
}
}
float val = x * ${o};
setOutput(val);
}
`}},Dte=class{constructor(e,t,n,a,r){this.variableNames=["x"],this.outputShape=[],this.packedInputs=!0,this.packedOutput=!0;let s=t,i=e[3]-1;this.outputShape=e;let o,l=`float(${n}) + float(${a}) * sum`;r===.5?o=`inversesqrt(${l})`:r===1?o=`1.0/(${l})`:o=`exp(log(${l}) * float(-${r}));`,this.userCode=`
void main() {
ivec4 coords = getOutputCoords();
int b = coords.x;
int r = coords.y;
int c = coords.z;
int d = coords.w;
bool hasNextCol = d < ${this.outputShape[3]};
bool hasNextRow = c < ${this.outputShape[2]};
vec4 sum = vec4(0.);
vec4 xFragAtOutputCoords = getX(b, r, c, d);
vec4 xAtOutputCoords = vec4(
getChannel(xFragAtOutputCoords, vec2(c, d)),
hasNextCol ?
getChannel(xFragAtOutputCoords, vec2(c, d + 1)) : 0.0,
hasNextRow ?
getChannel(xFragAtOutputCoords , vec2(c + 1, d)) : 0.0,
(hasNextRow && hasNextCol) ?
getChannel(xFragAtOutputCoords, vec2(c + 1, d + 1)) : 0.0
);
int firstChannel = d - ${s};
vec2 cache = vec2(0.);
if(firstChannel >= 0){
vec4 firstChannelFrag = getX(b, r, c, firstChannel);
cache.x = getChannel(firstChannelFrag, vec2(c, firstChannel));
if(hasNextRow){
cache.y = getChannel(firstChannelFrag, vec2(c + 1, firstChannel));
}
}
ivec2 depth = ivec2(d, d + 1);
for (int j = - ${s}; j <= ${s}; j++) {
ivec2 idx = depth + j;
bvec2 aboveLowerBound = greaterThanEqual(idx, ivec2(0));
bvec2 belowUpperBound = lessThanEqual(idx, ivec2(${i}));
bool depthInRange = aboveLowerBound.x && belowUpperBound.x;
bool depthPlusOneInRange = aboveLowerBound.y && belowUpperBound.y;
if(depthInRange || depthPlusOneInRange){
vec4 z = vec4(0.);
vec4 xFragAtCurrentDepth;
z.xz = cache.xy;
if(depthPlusOneInRange && hasNextCol){
xFragAtCurrentDepth = idx.y != d ?
getX(b, r, c, idx.y) : xFragAtOutputCoords;
z.y = getChannel(xFragAtCurrentDepth, vec2(c, idx.y));
if(hasNextRow){
z.w = getChannel(xFragAtCurrentDepth, vec2(c + 1, idx.y));
}
}
cache.xy = z.yw;
sum += z * z;
}
}
vec4 result = xAtOutputCoords * ${o};
setOutput(result);
}
`}},Rte=e=>{let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{depthRadius:s,bias:i,alpha:o,beta:l}=a,u=X().getBool("WEBGL_PACK_NORMALIZATION")?new Dte(r.shape,s,i,o,l):new $te(r.shape,s,i,o,l);return n.runWebGLProgram(u,[r],r.dtype)},Mte={kernelName:Nc,backendName:"webgl",kernelFunc:Rte},Pte=class{constructor(e,t,n,a,r){this.variableNames=["inputImage","outputImage","dy"],this.outputShape=[],this.outputShape=e,this.depth=e[3],this.depthRadius=t,this.bias=n,this.alpha=a,this.beta=r,this.userCode=`
void main() {
ivec4 coords = getOutputCoords();
int b = coords[0];
int r = coords[1];
int c = coords[2];
float result = 0.0;
for (int d = 0; d < ${this.depth}; ++d) {
int depthBegin = int(max(0.0, float(d - ${t})));
int depthEnd = int(min(float(${this.depth}),
float(d + ${t} + 1)));
const int MIN_DEPTH_BEGIN = 0;
const int MAX_DEPTH_END = ${this.depth};
float norm = 0.0;
for (int k = MIN_DEPTH_BEGIN; k < MAX_DEPTH_END; ++k) {
if (k < depthBegin){
continue;
}
else if (k >= depthBegin && k < depthEnd) {
norm += getInputImage(b, r, c, k) * getInputImage(b, r, c, k);
}
else {
break;
}
}
norm = float(${a}) * norm + float(${n});
for(int k = MIN_DEPTH_BEGIN; k < MAX_DEPTH_END; ++k){
if (k < depthBegin){
continue;
}
else if (k >= depthBegin && k < depthEnd){
float dyi = -2.0 * float(${a})
* float(${r})
* getInputImage(b ,r ,c, k) * getOutputImage(b, r, c, d)
/ norm;
if (k == d) {
dyi += pow(norm, -1.0 * ${r});
}
if (k == coords[3]) {
dyi *= getDy(b, r, c, d);
result += dyi;
}
}
else {
break;
}
}
}
setOutput(result);
}
`}},Ote=e=>{let{inputs:t,backend:n,attrs:a}=e,{x:r,y:s,dy:i}=t,{depthRadius:o,bias:l,alpha:u,beta:p}=a,d=new Pte(r.shape,o,l,u,p);return n.runWebGLProgram(d,[r,s,i],r.dtype)},Lte={kernelName:Bm,backendName:"webgl",kernelFunc:Ote};function zte(e,t,n,a){let r=w.sizeFromShape(t),s=w.sizeFromShape(e.shape)/r,i=me({inputs:{x:e},attrs:{shape:[s,r]},backend:a}),o=To(i,e.dtype,"max",a),l=me({inputs:{x:o},attrs:{shape:n},backend:a});return a.disposeIntermediateTensorInfo(i),a.disposeIntermediateTensorInfo(o),l}function QC(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{reductionIndices:s,keepDims:i}=a,o=r.shape.length,l=w.parseAxisParam(s,r.shape),u=l,p=C.getAxesPermutation(u,o),d=p!=null,c=n.shouldExecuteOnCPU([r]),h=r;if(d){if(c){let b=n.texData.get(h.dataId).values,x=new Array(o);for(let T=0;T<x.length;T++)x[T]=r.shape[p[T]];let v=W0(b,r.shape,r.dtype,p,x);h=n.makeTensorInfo(x,r.dtype);let k=n.texData.get(h.dataId);k.values=v}else h=lg(r,p,n);u=C.getInnerMostAxes(u.length,o)}C.assertAxesAreInnerMostDims("max",u,o);let[m,f]=C.computeOutAndReduceShapes(h.shape,u),g=m;i&&(g=C.expandShapeToKeepDim(m,l));let y;if(c){let b=n.texData.get(h.dataId).values,x=JY(b,w.sizeFromShape(f),g,r.dtype);y=n.makeTensorInfo(g,r.dtype);let v=n.texData.get(y.dataId);v.values=x}else y=zte(h,f,g,n);return d&&n.disposeIntermediateTensorInfo(h),y}var Bte={kernelName:Ui,backendName:"webgl",kernelFunc:QC},Wte=FC+`
return max(a, b);
`,Vte=`
vec4 result = vec4(max(a, b));
vec4 isNaN = min(vec4(isnan(a)) + vec4(isnan(b)), vec4(1.0));
`+ig+`
return result;
`,Ute=cn({opSnippet:Wte,packedOpSnippet:Vte,cpuKernelImpl:ZY}),Gte={kernelName:Gi,backendName:"webgl",kernelFunc:Ute};function Hte(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t;Zu(r,"maxPool");let{filterSize:s,strides:i,pad:o,dimRoundingMode:l}=a,u=1;w.assert(C.eitherStridesOrDilationsAreOne(i,u),()=>`Error in maxPool: Either strides or dilations must be 1. Got strides ${i} and dilations '${u}'`);let p=C.computePool2DInfo(r.shape,s,i,u,o,l);if(p.filterWidth===1&&p.filterHeight===1&&w.arraysEqual(p.inShape,p.outShape))return aa({inputs:{x:r},backend:n});let d=new hc(p,"max",!1);return n.runWebGLProgram(d,[r],r.dtype)}var jte={kernelName:Hi,backendName:"webgl",kernelFunc:Hte};function qte(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{filterSize:s,strides:i,pad:o,dataFormat:l,dimRoundingMode:u}=a,p=[1,1,1],d=C.computePool3DInfo(r.shape,s,i,p,o,u,l),c=new U0(d,"max",!1);return n.runWebGLProgram(c,[r],r.dtype)}var Kte={kernelName:Tc,backendName:"webgl",kernelFunc:qte},Xte=class{constructor(e){this.variableNames=["dy","maxPos"],this.outputShape=e.inShape;let t=e.strideHeight,n=e.strideWidth,a=e.dilationHeight,r=e.effectiveFilterHeight,s=e.effectiveFilterWidth,i=r-1-e.padInfo.top,o=s-1-e.padInfo.left,l=r*s-1;this.userCode=`
const ivec2 pads = ivec2(${i}, ${o});
void main() {
ivec4 coords = getOutputCoords();
int b = coords[0];
int d = coords[3];
ivec2 dyRCCorner = coords.yz - pads;
int dyRCorner = dyRCCorner.x;
int dyCCorner = dyRCCorner.y;
// Convolve dy(?, ?, d) with pos mask(:, :, d) to get dx(xR, xC, d).
// ? = to be determined. : = across all values in that axis.
float dotProd = 0.0;
for (int wR = 0; wR < ${r};
wR += ${a}) {
float dyR = float(dyRCorner + wR) / ${t}.0;
if (dyR < 0.0 || dyR >= ${e.outHeight}.0 || fract(dyR) > 0.0) {
continue;
}
int idyR = int(dyR);
for (int wC = 0; wC < ${s}; wC++) {
float dyC = float(dyCCorner + wC) / ${n}.0;
if (dyC < 0.0 || dyC >= ${e.outWidth}.0 ||
fract(dyC) > 0.0) {
continue;
}
int idyC = int(dyC);
float dyValue = getDy(b, idyR, idyC, d);
int maxPosValue = ${l} - int(getMaxPos(b, idyR, idyC, d));
// Get the current value, check it against the value from the
// position matrix.
int curPosValue = wR * ${s} + wC;
float mask = float(maxPosValue == curPosValue ? 1.0 : 0.0);
dotProd += dyValue * mask;
}
}
setOutput(dotProd);
}
`}},Yte=class{constructor(e){this.variableNames=["dy","maxPos"],this.outputShape=e.inShape;let t=e.strideDepth,n=e.strideHeight,a=e.strideWidth,r=e.dilationDepth,s=e.dilationHeight,i=e.dilationWidth,o=e.effectiveFilterDepth,l=e.effectiveFilterHeight,u=e.effectiveFilterWidth,p=o-1-e.padInfo.front,d=l-1-e.padInfo.top,c=u-1-e.padInfo.left,h=o*l*u-1;this.userCode=`
const ivec3 pads = ivec3(${p}, ${d}, ${c});
void main() {
ivec5 coords = getOutputCoords();
int batch = coords.x;
int ch = coords.u;
ivec3 dyCorner = ivec3(coords.y, coords.z, coords.w) - pads;
int dyDCorner = dyCorner.x;
int dyRCorner = dyCorner.y;
int dyCCorner = dyCorner.z;
// Convolve dy(?, ?, ?, ch) with pos mask(:, :, :, d) to get
// dx(xD, xR, xC, ch).
// ? = to be determined. : = across all values in that axis.
float dotProd = 0.0;
for (int wD = 0; wD < ${o};
wD += ${r}) {
float dyD = float(dyDCorner + wD) / ${t}.0;
if (dyD < 0.0 || dyD >= ${e.outDepth}.0 || fract(dyD) > 0.0) {
continue;
}
int idyD = int(dyD);
for (int wR = 0; wR < ${l};
wR += ${s}) {
float dyR = float(dyRCorner + wR) / ${n}.0;
if (dyR < 0.0 || dyR >= ${e.outHeight}.0 ||
fract(dyR) > 0.0) {
continue;
}
int idyR = int(dyR);
for (int wC = 0; wC < ${u};
wC += ${i}) {
float dyC = float(dyCCorner + wC) / ${a}.0;
if (dyC < 0.0 || dyC >= ${e.outWidth}.0 ||
fract(dyC) > 0.0) {
continue;
}
int idyC = int(dyC);
float dyValue = getDy(batch, idyD, idyR, idyC, ch);
int maxPosValue = ${h} -
int(getMaxPos(batch, idyD, idyR, idyC, ch));
// Get the current value, check it against the value from the
// position matrix.
int curPosValue =
wD * ${l} * ${u} +
wR * ${u} + wC;
float mask = float(maxPosValue == curPosValue ? 1.0 : 0.0);
dotProd += dyValue * mask;
}
}
}
setOutput(dotProd);
}
`}};function Qte(e){let{inputs:t,backend:n,attrs:a}=e,{dy:r,input:s}=t,i=s,{filterSize:o,strides:l,pad:u,dimRoundingMode:p}=a,d=[1,1,1],c=C.computePool3DInfo(i.shape,o,l,d,u,p),h=new U0(c,"max",!0),m=n.runWebGLProgram(h,[i],i.dtype),f=new Yte(c),g=n.runWebGLProgram(f,[r,m],i.dtype);return n.disposeIntermediateTensorInfo(m),g}var Jte={kernelName:Vm,backendName:"webgl",kernelFunc:Qte};function Zte(e){let{inputs:t,backend:n,attrs:a}=e,{dy:r,input:s,output:i}=t,o=s;Zu([s,i],"maxPoolGrad");let{filterSize:l,strides:u,pad:p,dimRoundingMode:d}=a,c=C.computePool2DInfo(o.shape,l,u,1,p,d),h=!0,m=new hc(c,"max",h),f=n.runWebGLProgram(m,[o],o.dtype),g=new Xte(c),y=n.runWebGLProgram(g,[r,f],o.dtype);return n.disposeIntermediateTensorInfo(f),y}var ene={kernelName:Wm,backendName:"webgl",kernelFunc:Zte};function tne(e,t,n,a){let r=new hc(n,"max",!1),s=a.runWebGLProgram(r,[e],"float32");r=new hc(n,"max",!0,!0,t);let i=a.runWebGLProgram(r,[e],"float32");return[s,i]}var nne={kernelName:Um,backendName:"webgl",kernelFunc:({inputs:e,attrs:t,backend:n})=>{let{x:a}=e,{filterSize:r,strides:s,pad:i,includeBatchInIndex:o}=t,l=n;w.assert(a.shape.length===4,()=>`Error in maxPool: input must be rank 4 but got rank ${a.shape.length}.`);let u=[1,1];w.assert(C.eitherStridesOrDilationsAreOne(s,u),()=>`Error in maxPool: Either strides or dilations must be 1. Got strides ${s} and dilations '${u}'`);let p=C.computePool2DInfo(a.shape,r,s,u,i),[d,c]=tne(a,o,p,l);return[d,c]}};function ane(e,t,n,a){let r=w.sizeFromShape(t),s=w.sizeFromShape(e.shape)/r,i=me({inputs:{x:e},attrs:{shape:[s,r]},backend:a}),o=To(i,"float32","mean",a),l=me({inputs:{x:o},attrs:{shape:n},backend:a});return a.disposeIntermediateTensorInfo(i),a.disposeIntermediateTensorInfo(o),l}var rne={kernelName:ji,backendName:"webgl",kernelFunc:({inputs:e,attrs:t,backend:n})=>{let{x:a}=e,{keepDims:r,axis:s}=t,i=n,o=a.shape.length,l=w.parseAxisParam(s,a.shape),u=l,p=C.getAxesPermutation(u,o),d=p!=null,c=i.shouldExecuteOnCPU([a]),h=[],m=a;if(d){if(c){let x=i.texData.get(m.dataId).values,v=new Array(o);for(let _=0;_<v.length;_++)v[_]=a.shape[p[_]];let k=W0(x,a.shape,a.dtype,p,v);m=i.makeTensorInfo(v,a.dtype);let T=i.texData.get(m.dataId);T.values=k}else m=lg(a,p,i);h.push(m),u=C.getInnerMostAxes(u.length,o)}C.assertAxesAreInnerMostDims("sum",u,o);let[f,g]=C.computeOutAndReduceShapes(m.shape,u),y=f;r&&(y=C.expandShapeToKeepDim(f,l));let b=ane(m,g,y,i);for(let x of h)i.disposeIntermediateTensorInfo(x);return b}};function sne(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{axis:s,keepDims:i}=a,o=r.shape.length,l=w.parseAxisParam(s,r.shape),u=l,p=C.getAxesPermutation(u,o),d=r;p!=null&&(d=Wt({inputs:{x:r},backend:n,attrs:{perm:p}}),u=C.getInnerMostAxes(u.length,r.shape.length)),C.assertAxesAreInnerMostDims("min",u,o);let[c,h]=C.computeOutAndReduceShapes(d.shape,u),m=w.sizeFromShape(h),f=me({inputs:{x:d},backend:n,attrs:{shape:[-1,m]}}),g=To(f,f.dtype,"min",n),y;if(i){let b=C.expandShapeToKeepDim(c,l);y=me({inputs:{x:g},backend:n,attrs:{shape:b}})}else y=me({inputs:{x:g},backend:n,attrs:{shape:c}});return n.disposeIntermediateTensorInfo(f),n.disposeIntermediateTensorInfo(g),p!=null&&n.disposeIntermediateTensorInfo(d),y}var ine={kernelName:qi,backendName:"webgl",kernelFunc:sne},one=FC+`
return min(a, b);
`,lne=`
vec4 result = vec4(min(a, b));
vec4 isNaN = min(vec4(isnan(a)) + vec4(isnan(b)), vec4(1.0));
`+ig+`
return result;
`,une=cn({opSnippet:one,packedOpSnippet:lne,cpuKernelImpl:e9}),pne={kernelName:Ki,backendName:"webgl",kernelFunc:une},cne=class{constructor(e,t,n){this.variableNames=["x"],this.outputShape=t.map((u,p)=>u[0]+e[p]+u[1]);let a=e.length,r=mt(a),s=t.map(u=>u[0]).join(","),i=t.map((u,p)=>u[0]+e[p]).join(","),o=["coords[0]","coords[1]","coords[2]","coords[3]"].slice(0,a),l=n==="reflect"?0:1;if(a===1){this.userCode=`
int start = ${s};
int end = ${i};
void main() {
int outC = getOutputCoords();
if (outC < start) {
outC = start * 2 - outC - ${l};
} else if(outC >= end) {
outC = (end - 1) * 2 - outC + ${l};
}
setOutput(getX(outC - start));
}
`;return}this.userCode=`
${r} start = ${r}(${s});
${r} end = ${r}(${i});
void main() {
${r} outC = getOutputCoords();
for (int i = 0; i < ${a}; i++) {
if (outC[i] < start[i]) {
outC[i] = start[i] * 2 - outC[i] - ${l};
} else if(outC[i] >= end[i]) {
outC[i] = (end[i] - 1) * 2 - outC[i] + ${l};
}
}
${r} coords = outC - start;
setOutput(getX(${o}));
}
`}},dne=class{constructor(e,t,n){this.variableNames=["x"],this.packedInputs=!0,this.packedOutput=!0,this.outputShape=t.map((h,m)=>h[0]+e[m]+h[1]);let a=e.length,r=mt(a),s=t.map(h=>h[0]).join(","),i=t.map((h,m)=>h[0]+e[m]).join(","),o=Nn("rc",a),l=Nn("source",a),u=`${o[a-1]} < ${this.outputShape[a-1]}`,p=a===1?"source":`vec2(${l.slice(-2).join()})`,d=n==="reflect"?0:1,c="";if(a===1){let h=`
${r} source = rc;
if (source < start) {
source = start * 2 - source - ${d};
} else if (source >= end) {
source = (end - 1) * 2 - source + ${d};
}
source -= start;
`;c=`
${r} rc = outputLoc;
${h}
result[0] = getChannel(getX(${l.join()}), ${p});
${o[a-1]} += 1;
if(${u}) {
${h}
result[1] = getChannel(getX(${l.join()}), ${p});
}
`}else{let h=`
${r} source = rc;
${r} lt = ${r}(lessThan(source, start));
${r} gte = ${r}(greaterThanEqual(source, end));
${r} orig = 1 - (lt + gte);
source = orig * source +
lt * (start * 2 - source - ${d}) +
gte * ((end - 1) * 2 - source + ${d});
source -= start;
`;c=`
${r} rc = outputLoc;
${h}
result[0] = getChannel(getX(${l.join()}), ${p});
${o[a-1]} += 1;
if(${u}) {
${h}
result[1] = getChannel(getX(${l.join()}), ${p});
}
rc = outputLoc;
${o[a-2]} += 1;
if(${o[a-2]} < ${this.outputShape[a-2]}) {
${h}
result[2] = getChannel(getX(${l.join()}), ${p});
${o[a-1]} += 1;
if(${u}) {
${h}
result[3] = getChannel(getX(${l.join()}), ${p});
}
}
`}this.userCode=`
const ${r} start = ${r}(${s});
const ${r} end = ${r}(${i});
void main() {
${r} outputLoc = getOutputCoords();
vec4 result = vec4(0.);
${c}
setOutput(result);
}
`}},hne=({inputs:e,backend:t,attrs:n})=>{let{x:a}=e,{paddings:r,mode:s}=n,i=X().getBool("WEBGL_PACK_ARRAY_OPERATIONS")?new dne(a.shape,r,s):new cne(a.shape,r,s);return t.runWebGLProgram(i,[a],a.dtype)},mne={kernelName:Xi,backendName:"webgl",kernelFunc:hne},fne=`if (b == 0.0) return NAN;
return mod(a, b);`,gne=`
vec4 result = mod(a, b);
vec4 isNaN = vec4(equal(b, vec4(0.0)));
`+ig+`
return result;
`,yne=cn({opSnippet:fne,packedOpSnippet:gne}),bne={kernelName:mu,backendName:"webgl",kernelFunc:yne},xne=class{constructor(e,t,n){this.variableNames=["probs"],this.customUniforms=[{name:"seed",type:"float"}],this.outputShape=[e,n],this.userCode=`
void main() {
ivec2 coords = getOutputCoords();
int batch = coords[0];
float r = random(seed);
float cdf = 0.0;
for (int i = 0; i < ${t-1}; i++) {
cdf += getProbs(batch, i);
if (r < cdf) {
setOutput(float(i));
return;
}
}
// If no other event happened, last event happened.
setOutput(float(${t-1}));
}
`}},vne=`
if (a == b) {
return 1.0;
};
return a / b;`,wne=`
// vec4 one = vec4(equal(a, b));
// return one + (vec4(1.0) - one) * a / b;
vec4 result = a / b;
if(a.x == b.x) {
result.x = 1.;
}
if(a.y == b.y) {
result.y = 1.;
}
if(a.z == b.z) {
result.z = 1.;
}
if(a.w == b.w) {
result.w = 1.;
}
return result;
`,JC=cn({opSnippet:vne,packedOpSnippet:wne,checkOutOfBounds:!0}),kne={kernelName:Di,backendName:"webgl",kernelFunc:JC},cI="return a - b;",ZC=cn({opSnippet:cI,packedOpSnippet:cI,supportsComplex:!0,cpuKernelImpl:y9}),Ine={kernelName:fo,backendName:"webgl",kernelFunc:ZC};function eE(e){let{inputs:t,backend:n,attrs:a}=e,{logits:r}=t,{dim:s}=a,i=w.parseAxisParam([s],r.shape),o=QC({inputs:{x:r},backend:n,attrs:{reductionIndices:i,keepDims:!1}}),l=C.expandShapeToKeepDim(o.shape,i),u=me({inputs:{x:o},backend:n,attrs:{shape:l}}),p=ZC({inputs:{a:r,b:u},backend:n}),d=KC({inputs:{x:p},backend:n}),c=ug({inputs:{x:d},backend:n,attrs:{axis:i,keepDims:!1}}),h=me({inputs:{x:c},backend:n,attrs:{shape:l}}),m=JC({inputs:{a:d,b:h},backend:n});return n.disposeIntermediateTensorInfo(o),n.disposeIntermediateTensorInfo(u),n.disposeIntermediateTensorInfo(p),n.disposeIntermediateTensorInfo(d),n.disposeIntermediateTensorInfo(c),n.disposeIntermediateTensorInfo(h),m}var Sne={kernelName:ho,backendName:"webgl",kernelFunc:eE};function Nne(e){let{inputs:t,backend:n,attrs:a}=e,{logits:r}=t,{numSamples:s,seed:i,normalized:o}=a,l=o?r:eE({inputs:{logits:r},backend:n,attrs:{dim:r.shape.length-1}}),u=l.shape[0],p=l.shape[1],d=new xne(u,p,s),c=[[i]],h=n.runWebGLProgram(d,[l],"int32",c);return o||n.disposeIntermediateTensorInfo(l),h}var Tne={kernelName:Gm,backendName:"webgl",kernelFunc:Nne},_ne=Aa+`
return -x;
`,Cne=`
vec4 result = -x;
bvec4 isNaN = isnan(x);
result.r = isNaN.r ? x.r : result.r;
result.g = isNaN.g ? x.g : result.g;
result.b = isNaN.b ? x.b : result.b;
result.a = isNaN.a ? x.a : result.a;
return result;
`;function Ene(e){let{inputs:t,backend:n}=e,{x:a}=t;if(n.shouldExecuteOnCPU([a])){let s=n.texData.get(a.dataId),[i,o]=n9(s.values,a.shape,a.dtype);return n.makeTensorInfo(o,a.dtype,i)}let r;return X().getBool("WEBGL_PACK_UNARY_OPERATIONS")?r=new Js(a.shape,Cne):r=new _r(a.shape,_ne),n.runWebGLProgram(r,[a],a.dtype)}var Fne={kernelName:fu,backendName:"webgl",kernelFunc:Ene},Ane=gr.nonMaxSuppressionV3Impl;function $ne(e){C.warn("tf.nonMaxSuppression() in webgl locks the UI thread. Call tf.nonMaxSuppressionAsync() instead");let{inputs:t,backend:n,attrs:a}=e,{boxes:r,scores:s}=t,{maxOutputSize:i,iouThreshold:o,scoreThreshold:l}=a,u=n.readSync(r.dataId),p=n.readSync(s.dataId),{selectedIndices:d}=Ane(u,p,i,o,l);return n.makeTensorInfo([d.length],"int32",new Int32Array(d))}var Dne={kernelName:yu,backendName:"webgl",kernelFunc:$ne},Rne=gr.nonMaxSuppressionV4Impl;function Mne(e){C.warn("tf.nonMaxSuppression() in webgl locks the UI thread. Call tf.nonMaxSuppressionAsync() instead");let{inputs:t,backend:n,attrs:a}=e,{boxes:r,scores:s}=t,{maxOutputSize:i,iouThreshold:o,scoreThreshold:l,padToMaxOutputSize:u}=a,p=n.readSync(r.dataId),d=n.readSync(s.dataId),{selectedIndices:c,validOutputs:h}=Rne(p,d,i,o,l,u);return[n.makeTensorInfo([c.length],"int32",new Int32Array(c)),n.makeTensorInfo([],"int32",new Int32Array([h]))]}var Pne={kernelName:bu,backendName:"webgl",kernelFunc:Mne},One=gr.nonMaxSuppressionV5Impl;function Lne(e){C.warn("tf.nonMaxSuppression() in webgl locks the UI thread. Call tf.nonMaxSuppressionAsync() instead");let{inputs:t,backend:n,attrs:a}=e,{boxes:r,scores:s}=t,{maxOutputSize:i,iouThreshold:o,scoreThreshold:l,softNmsSigma:u}=a,p=n.readSync(r.dataId),d=n.readSync(s.dataId),c=i,h=o,m=l,f=u,{selectedIndices:g,selectedScores:y}=One(p,d,c,h,m,f);return[n.makeTensorInfo([g.length],"int32",new Int32Array(g)),n.makeTensorInfo([y.length],"float32",new Float32Array(y))]}var zne={kernelName:xu,backendName:"webgl",kernelFunc:Lne},Bne=class{constructor(e,t,n,a){this.variableNames=["indices"],this.outputShape=[e,t],this.userCode=`
void main() {
ivec2 coords = getOutputCoords();
int index = round(getIndices(coords.x));
setOutput(mix(float(${a}), float(${n}),
float(index == coords.y)));
}
`}},Wne=e=>{let{inputs:t,backend:n,attrs:a}=e,{indices:r}=t,{depth:s,onValue:i,offValue:o}=a,l=w.sizeFromShape(r.shape),u=new Bne(l,s,i,o),p=me({inputs:{x:r},backend:n,attrs:{shape:[l]}}),d=n.runWebGLProgram(u,[p],r.dtype);n.disposeIntermediateTensorInfo(p);let c=[...r.shape,s],h=me({inputs:{x:d},backend:n,attrs:{shape:c}});return n.disposeIntermediateTensorInfo(d),h},Vne={kernelName:Qi,backendName:"webgl",kernelFunc:Wne};function ym(e){let{inputs:t,backend:n}=e,{x:a}=t;if(a.dtype==="complex64"){let r=hd({inputs:{input:a},backend:n}),s=ym({inputs:{x:r},backend:n}),i=pg({inputs:{input:a},backend:n}),o=ym({inputs:{x:i},backend:n}),l=Cs({inputs:{real:s,imag:o},backend:n});return n.disposeIntermediateTensorInfo(r),n.disposeIntermediateTensorInfo(s),n.disposeIntermediateTensorInfo(i),n.disposeIntermediateTensorInfo(o),l}else return md({attrs:{shape:a.shape,dtype:a.dtype,value:a.dtype==="string"?"":0},backend:n})}var Une={kernelName:Lu,backendName:"webgl",kernelFunc:ym};function tE(e){let{inputs:t,backend:n}=e,{x:a}=t;if(a.dtype==="string")throw new Error("onesLike is not supported under string dtype");if(a.dtype==="complex64"){let r=hd({inputs:{input:a},backend:n}),s=tE({inputs:{x:r},backend:n}),i=pg({inputs:{input:a},backend:n}),o=ym({inputs:{x:i},backend:n}),l=Cs({inputs:{real:s,imag:o},backend:n});return n.disposeIntermediateTensorInfo(r),n.disposeIntermediateTensorInfo(s),n.disposeIntermediateTensorInfo(i),n.disposeIntermediateTensorInfo(o),l}else return md({attrs:{shape:a.shape,dtype:a.dtype,value:1},backend:n})}var Gne={kernelName:vu,backendName:"webgl",kernelFunc:tE};function Hne(e){let{inputs:t,backend:n,attrs:a}=e,{axis:r}=a;if(t.length===1)return Ix({inputs:{input:t[0]},backend:n,attrs:{dim:r}});let s=t[0].shape,i=t[0].dtype;t.forEach(p=>{w.assertShapesMatch(s,p.shape,"All tensors passed to stack must have matching shapes"),w.assert(i===p.dtype,()=>"All tensors passed to stack must have matching dtypes")});let o=[],l=t.map(p=>{let d=Ix({inputs:{input:p},backend:n,attrs:{dim:r}});return o.push(d),d}),u=WC({inputs:l,backend:n,attrs:{axis:r}});return o.forEach(p=>n.disposeIntermediateTensorInfo(p)),u}var jne={kernelName:wu,backendName:"webgl",kernelFunc:Hne},qne=class{constructor(e,t,n){this.variableNames=["x"],this.customUniforms=[{name:"value",type:"float"}],this.outputShape=t.map((l,u)=>l[0]+e[u]+l[1]);let a=e.length,r=mt(a),s=t.map(l=>l[0]).join(","),i=t.map((l,u)=>l[0]+e[u]).join(","),o=["coords[0]","coords[1]","coords[2]","coords[3]"].slice(0,a);if(a===1){this.userCode=`
int start = ${s};
int end = ${i};
void main() {
int outC = getOutputCoords();
if (outC < start || outC >= end) {
setOutput(value);
} else {
setOutput(getX(outC - start));
}
}
`;return}this.userCode=`
${r} start = ${r}(${s});
${r} end = ${r}(${i});
void main() {
${r} outC = getOutputCoords();
if (any(lessThan(outC, start)) || any(greaterThanEqual(outC, end))) {
setOutput(value);
} else {
${r} coords = outC - start;
setOutput(getX(${o}));
}
}
`}},Kne=class{constructor(e,t,n){this.variableNames=["x"],this.packedInputs=!0,this.packedOutput=!0,this.customUniforms=[{name:"value",type:"float"}],this.outputShape=t.map((m,f)=>m[0]+e[f]+m[1]);let a=e.length,r=mt(a),s=t.map(m=>m[0]).join(","),i=t.map((m,f)=>m[0]+e[f]).join(","),o=Nn("rc",a),l=Nn("source",a),u=`${o[a-1]} < ${this.outputShape[a-1]}`,p=a===1?"source":`vec2(${l.slice(-2).join()})`,d=[`${r} rc = outputLoc;`,`${o[a-1]} += 1;
if(${u}) {
`,a===1?"":`}
rc = outputLoc;
${o[a-2]} += 1;
if(${o[a-2]} < ${this.outputShape[a-2]}) {`,a===1?"":` ${o[a-1]} += 1;
if(${u}) {`],c=a===1?"rc < start || rc >= end":"any(lessThan(rc, start)) || any(greaterThanEqual(rc, end))",h="";for(let m=0,f=a===1?2:4;m<f;m++)h+=`
${d[m]}
if (${c}) {
result[${m}] = float(value);
} else {
${r} source = rc - start;
result[${m}] = getChannel(getX(${l.join()}), ${p});
}
`;h+=a===1?"} ":"}}",this.userCode=`
const ${r} start = ${r}(${s});
const ${r} end = ${r}(${i});
void main() {
${r} outputLoc = getOutputCoords();
vec4 result = vec4(0.);
${h}
setOutput(result);
}
`}},nE=e=>{let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{paddings:s,constantValue:i}=a;if(w.sizeFromShape(r.shape)===0){let u=s.map((p,d)=>p[0]+r.shape[d]+p[1]);return md({backend:n,attrs:{shape:u,value:i,dtype:r.dtype}})}let o=X().getBool("WEBGL_PACK_ARRAY_OPERATIONS")?new Kne(r.shape,s,i):new qne(r.shape,s,i),l=[[i]];return n.runWebGLProgram(o,[r],r.dtype,l)},Xne={kernelName:Ji,backendName:"webgl",kernelFunc:nE},Yne=`
if(a < 0.0 && floor(b) < b){
return NAN;
}
if (b == 0.0) {
return 1.0;
}
return (round(mod(b, 2.0)) != 1) ?
pow(abs(a), b) : sign(a) * pow(abs(a), b);
`,Qne=`
// isModRound1 has 1 for components with round(mod(b, 2.0)) == 1, 0 otherwise.
vec4 isModRound1 = vec4(equal(round(mod(b, 2.0)), ivec4(1)));
vec4 multiplier = sign(a) * isModRound1 + (vec4(1.0) - isModRound1);
vec4 result = multiplier * pow(abs(a), b);
// Ensure that a^0 = 1, including 0^0 = 1 as this correspond to TF and JS
bvec4 isExpZero = equal(b, vec4(0.0));
result.r = isExpZero.r ? 1.0 : result.r;
result.g = isExpZero.g ? 1.0 : result.g;
result.b = isExpZero.b ? 1.0 : result.b;
result.a = isExpZero.a ? 1.0 : result.a;
vec4 isNaN = vec4(lessThan(a, vec4(0.0))) * vec4(lessThan(floor(b), b));
`+ig+`
return result;
`,Jne=cn({opSnippet:Yne,packedOpSnippet:Qne}),Zne={kernelName:Zi,backendName:"webgl",kernelFunc:Jne};function eae(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{axis:s,keepDims:i}=a,o=r.shape.length,l=[],u=w.parseAxisParam(s,r.shape),p=u,d=C.getAxesPermutation(p,o),c=r;d!=null&&(c=Wt({inputs:{x:r},backend:n,attrs:{perm:d}}),p=C.getInnerMostAxes(p.length,o),l.push(c)),C.assertAxesAreInnerMostDims("prod",p,o);let h;if(n.shouldExecuteOnCPU([c])){let m=n.texData.get(c.dataId).values,{outVals:f,outShape:g,outDtype:y}=r9(c.shape,c.dtype,m,p);h=n.makeTensorInfo(g,y,f)}else{let[m,f]=C.computeOutAndReduceShapes(c.shape,p),g=w.sizeFromShape(f),y=me({inputs:{x:c},backend:n,attrs:{shape:[-1,g]}}),b=tf(r.dtype),x=To(y,b,"prod",n);h=me({inputs:{x},backend:n,attrs:{shape:m}}),l.push(y),l.push(x)}if(i){l.push(h);let m=C.expandShapeToKeepDim(h.shape,u);h=me({inputs:{x:h},backend:n,attrs:{shape:m}})}return l.forEach(m=>n.disposeIntermediateTensorInfo(m)),h}var tae={kernelName:to,backendName:"webgl",kernelFunc:eae},aE=e=>{let{backend:t,attrs:n}=e,{start:a,stop:r,step:s,dtype:i}=n,o=s9(a,r,s,i);return t.makeTensorInfo([o.length],i,o)},nae={kernelName:_c,backendName:"webgl",kernelFunc:aE},aae="return 1.0 / x;",rae=Qe({opSnippet:aae}),sae={kernelName:ku,backendName:"webgl",kernelFunc:rae},iae=Aa+`
return (x < 0.0) ? 0.0 : x;
`,oae=`
vec4 result = x * vec4(greaterThanEqual(x, vec4(0.0)));
bvec4 isNaN = isnan(x);
result.r = isNaN.r ? x.r : result.r;
result.g = isNaN.g ? x.g : result.g;
result.b = isNaN.b ? x.b : result.b;
result.a = isNaN.a ? x.a : result.a;
return result;
`,lae=Qe({opSnippet:iae,packedOpSnippet:oae}),uae={kernelName:no,backendName:"webgl",kernelFunc:lae},pae=Aa+`
return (x < 0.0) ? 0.0 : min(6.0, x);
`,cae=`
vec4 result = min(x, vec4(6.)) * vec4(greaterThanEqual(x, vec4(0.0)));
bvec4 isNaN = isnan(x);
result.r = isNaN.r ? x.r : result.r;
result.g = isNaN.g ? x.g : result.g;
result.b = isNaN.b ? x.b : result.b;
result.a = isNaN.a ? x.a : result.a;
return result;
`,dae=Qe({opSnippet:pae,packedOpSnippet:cae}),hae={kernelName:ro,backendName:"webgl",kernelFunc:dae},mae=class{constructor(e,t,n,a,r){this.variableNames=["A"],this.outputShape=[];let[s,i,o,l]=e;this.outputShape=[s,t,n,l];let u=[a&&t>1?i-1:i,a&&n>1?o-1:o],p=[a&&t>1?t-1:t,a&&n>1?n-1:n],d;r?d="(vec2(yRC) + vec2(0.5)) * effectiveInputOverOutputRatioRC - vec2(0.5)":d="vec2(yRC) * effectiveInputOverOutputRatioRC",this.userCode=`
const vec2 effectiveInputOverOutputRatioRC = vec2(
${u[0]/p[0]},
${u[1]/p[1]});
const vec2 inputShapeRC = vec2(${i}.0, ${o}.0);
void main() {
ivec4 coords = getOutputCoords();
int b = coords[0];
int d = coords[3];
ivec2 yRC = coords.yz;
// Fractional source index.
vec2 sourceFracIndexRC = ${d};
// Compute the four integer indices.
ivec2 sourceFloorRC = ivec2(max(sourceFracIndexRC, vec2(0.0)));
ivec2 sourceCeilRC = ivec2(
min(inputShapeRC - 1.0, ceil(sourceFracIndexRC)));
float topLeft = getA(b, sourceFloorRC.x, sourceFloorRC.y, d);
float bottomLeft = getA(b, sourceCeilRC.x, sourceFloorRC.y, d);
float topRight = getA(b, sourceFloorRC.x, sourceCeilRC.y, d);
float bottomRight = getA(b, sourceCeilRC.x, sourceCeilRC.y, d);
vec2 fracRC = sourceFracIndexRC - vec2(sourceFloorRC);
float top = topLeft + (topRight - topLeft) * fracRC.y;
float bottom = bottomLeft + (bottomRight - bottomLeft) * fracRC.y;
float newValue = top + (bottom - top) * fracRC.x;
setOutput(newValue);
}
`}},fae=class{constructor(e,t,n,a,r){this.variableNames=["A"],this.packedInputs=!0,this.packedOutput=!0,this.outputShape=[];let[s,i,o,l]=e;this.outputShape=[s,t,n,l];let u=[a&&t>1?i-1:i,a&&n>1?o-1:o],p=[a&&t>1?t-1:t,a&&n>1?n-1:n],d;r?d="(vec3(yRC) + vec3(0.5)) * effectiveInputOverOutputRatioRC - vec3(0.5)":d="vec3(yRC) * effectiveInputOverOutputRatioRC",this.userCode=`
const vec3 effectiveInputOverOutputRatioRC = vec3(
${u[0]/p[0]},
${u[1]/p[1]},
${u[1]/p[1]});
const vec3 inputShapeRC = vec3(${i}.0, ${o}.0,
${o}.0);
float getAValue(int b, int r, int c, int d) {
return getChannel(getA(b, r, c, d), vec2(c, d));
}
void main() {
ivec4 coords = getOutputCoords();
int b = coords[0];
int d = coords[3];
// Calculate values for next column in yRC.z.
ivec3 yRC = coords.yzz + ivec3(0, 0, 1);
// Fractional source index.
vec3 sourceFracIndexRC = ${d};
// Compute the four integer indices.
ivec3 sourceFloorRC = ivec3(max(sourceFracIndexRC, vec3(0.0)));
ivec3 sourceCeilRC = ivec3(
min(inputShapeRC - 1.0, ceil(sourceFracIndexRC)));
// Should we calculate next column and row elements in 2x2 packed cell.
bool hasNextCol = d < ${l-1};
bool hasNextRow = coords.z < ${n-1};
// In parallel, construct four corners for all four components in
// packed 2x2 cell.
vec4 topLeft = vec4(
getAValue(b, sourceFloorRC.x, sourceFloorRC.y, d),
hasNextCol ? getAValue(b, sourceFloorRC.x, sourceFloorRC.y, d + 1)
: 0.0,
hasNextRow ? getAValue(b, sourceFloorRC.x, sourceFloorRC.z, d)
: 0.0,
(hasNextRow && hasNextCol) ?
getAValue(b, sourceFloorRC.x, sourceFloorRC.z, d + 1) : 0.0);
vec4 bottomLeft = vec4(
getAValue(b, sourceCeilRC.x, sourceFloorRC.y, d),
hasNextCol ? getAValue(b, sourceCeilRC.x, sourceFloorRC.y, d + 1)
: 0.0,
hasNextRow ? getAValue(b, sourceCeilRC.x, sourceFloorRC.z, d)
: 0.0,
(hasNextRow && hasNextCol) ?
getAValue(b, sourceCeilRC.x, sourceFloorRC.z, d + 1) : 0.0);
vec4 topRight = vec4(
getAValue(b, sourceFloorRC.x, sourceCeilRC.y, d),
hasNextCol ? getAValue(b, sourceFloorRC.x, sourceCeilRC.y, d + 1)
: 0.0,
hasNextRow ? getAValue(b, sourceFloorRC.x, sourceCeilRC.z, d)
: 0.0,
(hasNextRow && hasNextCol) ?
getAValue(b, sourceFloorRC.x, sourceCeilRC.z, d + 1) : 0.0);
vec4 bottomRight = vec4(
getAValue(b, sourceCeilRC.x, sourceCeilRC.y, d),
hasNextCol ? getAValue(b, sourceCeilRC.x, sourceCeilRC.y, d + 1)
: 0.0,
hasNextRow ? getAValue(b, sourceCeilRC.x, sourceCeilRC.z, d)
: 0.0,
(hasNextRow && hasNextCol) ?
getAValue(b, sourceCeilRC.x, sourceCeilRC.z, d + 1) : 0.0);
vec3 fracRC = sourceFracIndexRC - vec3(sourceFloorRC);
vec4 top = mix(topLeft, topRight, fracRC.yyzz);
vec4 bottom = mix(bottomLeft, bottomRight, fracRC.yyzz);
vec4 newValue = mix(top, bottom, fracRC.x);
setOutput(newValue);
}
`}};function gae(e){let{inputs:t,backend:n,attrs:a}=e,{images:r}=t,{alignCorners:s,halfPixelCenters:i,size:o}=a,[l,u]=o,p=X().getBool("WEBGL_PACK_IMAGE_OPERATIONS")?new fae(r.shape,l,u,s,i):new mae(r.shape,l,u,s,i);return n.runWebGLProgram(p,[r],"float32")}var yae={kernelName:ao,backendName:"webgl",kernelFunc:gae},bae=class{constructor(e,t,n){this.variableNames=["dy"],this.outputShape=[],this.outputShape=t;let[,a,r]=t,[,s,i]=e,o=[n&&s>1?a-1:a,n&&i>1?r-1:r],l=[n&&s>1?s-1:s,n&&i>1?i-1:i],u=o[0]/l[0],p=o[1]/l[1],d=1/u,c=1/p,h=Math.ceil(d)*2+2,m=Math.ceil(c)*2+2;this.userCode=`
void main() {
ivec4 coords = getOutputCoords();
int b = coords[0];
int d = coords[3];
int r = coords[1];
int c = coords[2];
float accumulator = 0.0;
const float heightScale = float(${u});
const float widthScale = float(${p});
const float invHeightScale = float(${d});
const float invWidthScale = float(${c});
const int winHeight = int(${h});
const int winWidth = int(${m});
// Compute bounds for where in dy we will look
float startRLerp = floor(float(r) * invHeightScale);
int startDyR = int(startRLerp - float(winHeight / 2));
float startCLerp = floor(float(c) * invWidthScale);
int startDyC = int(startCLerp - float(winWidth / 2));
// Loop over dy
for (int dyROffset = 0; dyROffset < winHeight; dyROffset++) {
int dyR = dyROffset + startDyR;
// Guard against the window exceeding the bounds of dy
if (dyR < 0 || dyR >= ${s}) {
continue;
}
for (int dyCOffset = 0; dyCOffset < winWidth; dyCOffset++) {
int dyC = dyCOffset + startDyC;
// Guard against the window exceeding the bounds of dy
if (dyC < 0 || dyC >= ${i}) {
continue;
}
float dxR = float(dyR) * heightScale;
int topDxRIndex = int(floor(dxR));
int bottomDxRIndex = int(min(ceil(dxR), ${a-1}.0));
float dxRLerp = dxR - float(topDxRIndex);
float inverseDxRLerp = 1.0 - dxRLerp;
float dxC = float(dyC) * widthScale;
int leftDxCIndex = int(floor(dxC));
int rightDxCIndex = int(min(ceil(dxC), ${r-1}.0));
float dxCLerp = dxC - float(leftDxCIndex);
float inverseDxCLerp = 1.0 - dxCLerp;
if (r == topDxRIndex && c == leftDxCIndex) {
// topLeft
accumulator +=
getDy(b, dyR, dyC, d) * inverseDxRLerp * inverseDxCLerp;
}
if (r == topDxRIndex && c == rightDxCIndex) {
// topRight
accumulator += getDy(b, dyR, dyC, d) * inverseDxRLerp * dxCLerp;
}
if (r == bottomDxRIndex && c == leftDxCIndex) {
// bottomLeft
accumulator += getDy(b, dyR, dyC, d) * dxRLerp * inverseDxCLerp;
}
if (r == bottomDxRIndex && c == rightDxCIndex) {
// bottomRight
accumulator += getDy(b, dyR, dyC, d) * dxRLerp * dxCLerp;
}
}
}
// End loop over dy
setOutput(accumulator);
}
`}};function xae(e){let{inputs:t,backend:n,attrs:a}=e,{images:r,dy:s}=t,{alignCorners:i}=a,o=new bae(s.shape,r.shape,i);return n.runWebGLProgram(o,[s],s.dtype)}var vae={kernelName:qm,backendName:"webgl",kernelFunc:xae},wae=class{constructor(e,t,n,a,r){this.variableNames=["A"],this.outputShape=[];let[s,i,o,l]=e;this.outputShape=[s,t,n,l];let u=[a&&t>1?i-1:i,a&&n>1?o-1:o],p=[a&&t>1?t-1:t,a&&n>1?n-1:n],d=a?"0.5":"0.0",c;r?c="max((vec2(yRC) + vec2(0.5)) * effectiveInputOverOutputRatioRC, vec2(0.0))":c="vec2(yRC) * effectiveInputOverOutputRatioRC",this.userCode=`
const vec2 effectiveInputOverOutputRatioRC = vec2(
${u[0]/p[0]},
${u[1]/p[1]});
const vec2 inputShapeRC = vec2(${i}.0, ${o}.0);
void main() {
ivec4 coords = getOutputCoords();
int b = coords[0];
int d = coords[3];
ivec2 yRC = coords.yz;
// Fractional source index.
vec2 sourceFracIndexRC = ${c};
// Compute the coordinators of nearest neighbor point.
ivec2 sourceNearestRC = ivec2(
min(inputShapeRC - 1.0, floor(sourceFracIndexRC + ${d})));
float newValue = getA(b, sourceNearestRC.x, sourceNearestRC.y, d);
setOutput(newValue);
}
`}},kae=class{constructor(e,t,n,a,r){this.variableNames=["A"],this.packedInputs=!0,this.packedOutput=!0,this.outputShape=[];let[s,i,o,l]=e;this.outputShape=[s,t,n,l];let u=[a&&t>1?i-1:i,a&&n>1?o-1:o],p=[a&&t>1?t-1:t,a&&n>1?n-1:n],d=a?"0.5":"0.0",c;r?c="max((vec3(yRC) + vec3(0.5)) * effectiveInputOverOutputRatioRC, vec3(0.0))":c="vec3(yRC) * effectiveInputOverOutputRatioRC",this.userCode=`
const vec3 effectiveInputOverOutputRatioRC = vec3(
${u[0]/p[0]},
${u[1]/p[1]},
${u[1]/p[1]});
const vec3 inputShapeRC = vec3(${i}.0, ${o}.0,
${o}.0);
float getAValue(int b, int r, int c, int d) {
return getChannel(getA(b, r, c, d), vec2(c, d));
}
void main() {
ivec4 coords = getOutputCoords();
int b = coords[0];
int d = coords[3];
// Calculate values for next column in yRC.z.
ivec3 yRC = coords.yzz + ivec3(0, 0, 1);
// Fractional source index.
vec3 sourceFracIndexRC = ${c};
// Compute the coordinators of nearest neighbor point.
ivec3 sourceNearestRC = ivec3(
min(inputShapeRC - 1.0, floor(sourceFracIndexRC + ${d})));
// Should we calculate next column and row elements in 2x2 packed cell.
bool hasNextCol = d < ${l-1};
bool hasNextRow = coords.z < ${n-1};
vec4 newValue = vec4(
getAValue(b, sourceNearestRC.x, sourceNearestRC.y, d),
hasNextCol ? getAValue(b, sourceNearestRC.x, sourceNearestRC.y, d + 1)
: 0.0,
hasNextRow ? getAValue(b, sourceNearestRC.x, sourceNearestRC.z, d)
: 0.0,
(hasNextRow && hasNextCol) ?
getAValue(b, sourceNearestRC.x, sourceNearestRC.z, d + 1) : 0.0);
setOutput(newValue);
}
`}};function Iae(e){let{inputs:t,backend:n,attrs:a}=e,{images:r}=t,{alignCorners:s,halfPixelCenters:i,size:o}=a,[l,u]=o,p=X().getBool("WEBGL_PACK_IMAGE_OPERATIONS")?new kae(r.shape,l,u,s,i):new wae(r.shape,l,u,s,i);return n.runWebGLProgram(p,[r],r.dtype)}var Sae={kernelName:Cc,backendName:"webgl",kernelFunc:Iae},Nae=class{constructor(e,t,n){this.variableNames=["dy"],this.outputShape=[],this.outputShape=t;let[,a,r]=t,[,s,i]=e,o=[n&&s>1?a-1:a,n&&i>1?r-1:r],l=[n&&s>1?s-1:s,n&&i>1?i-1:i],u=o[0]/l[0],p=o[1]/l[1],d=1/u,c=1/p,h=Math.ceil(d)*2+2,m=Math.ceil(c)*2+2;this.userCode=`
void main() {
ivec4 coords = getOutputCoords();
int b = coords[0];
int d = coords[3];
int r = coords[1];
int c = coords[2];
float accumulator = 0.0;
const float heightScale = float(${u});
const float widthScale = float(${p});
const float invHeightScale = float(${d});
const float invWidthScale = float(${c});
const int winHeight = int(${h});
const int winWidth = int(${m});
// Compute bounds for where in dy we will look
float startRLerp = floor(float(r) * invHeightScale);
int startDyR = int(floor(startRLerp - float(winHeight / 2)));
float startCLerp = floor(float(c) * invWidthScale);
int startDyC = int(floor(startCLerp - float(winWidth / 2)));
// Loop over dy
for (int dyROffset = 0; dyROffset < winHeight; dyROffset++) {
int dyR = dyROffset + startDyR;
// Guard against the window exceeding the bounds of dy
if (dyR < 0 || dyR >= ${s}) {
continue;
}
for (int dyCOffset = 0; dyCOffset < winWidth; dyCOffset++) {
int dyC = dyCOffset + startDyC;
// Guard against the window exceeding the bounds of dy
if (dyC < 0 || dyC >= ${i}) {
continue;
}
float sourceFracRow =
float(${o[0]}) *
(float(dyR) / float(${l[0]}));
float sourceFracCol =
float(${o[1]}) *
(float(dyC) / float(${l[1]}));
int sourceNearestRow = int(min(
float(int(${a}) - 1),
${n} ? float(round(sourceFracRow)) :
float(floor(sourceFracRow))));
int sourceNearestCol = int(min(
float(int(${r}) - 1),
${n} ? float(round(sourceFracCol)) :
float(floor(sourceFracCol))));
if (r == sourceNearestRow && c == sourceNearestCol) {
accumulator += getDy(b, dyR, dyC, d);
}
}
}
// End loop over dy
setOutput(accumulator);
}
`}};function Tae(e){let{inputs:t,backend:n,attrs:a}=e,{images:r,dy:s}=t,{alignCorners:i}=a,o=new Nae(s.shape,r.shape,i);return n.runWebGLProgram(o,[s],s.dtype)}var _ae={kernelName:jm,backendName:"webgl",kernelFunc:Tae},Cae=class{constructor(e,t){this.variableNames=["x"];let n=e.length;if(n>4)throw new Error(`WebGL backend: Reverse of rank-${n} tensor is not yet supported`);if(this.outputShape=e,n===1){this.userCode=`
void main() {
int coord = getOutputCoords();
setOutput(getX(${e[0]} - coord - 1));
}
`;return}let a=i=>t.indexOf(i)!==-1&&e[i]!==1?`${e[i]} - coords[${i}] - 1`:`coords[${i}]`,r=e.map((i,o)=>a(o)).join(","),s=mt(n);this.userCode=`
void main() {
${s} coords = getOutputCoords();
setOutput(getX(${r}));
}
`}},Eae=class{constructor(e,t){this.variableNames=["x"],this.packedInputs=!0,this.packedOutput=!0;let n=e.length;if(n>4)throw new Error(`WebGL backend: Reverse of rank-${n} tensor is not yet supported`);this.outputShape=e;let a=Nn("rc",n),r=`${a[n-1]} + 1 < ${this.outputShape[n-1]}`,s=`${a[n-2]} + 1 < ${this.outputShape[n-2]}`,i=mt(n);n===1?this.userCode=`
void main(){
int rc = getOutputCoords();
vec4 result = vec4(0.);
result.r = getChannel(getX(${e[0]} - rc - 1),
${e[0]} - rc - 1);
if(${r}){
result.g = getChannel(getX(${e[0]} - (rc + 1) - 1),
${e[0]} - (rc + 1) - 1);
}
setOutput(result);
}
`:this.userCode=`
void main() {
${i} rc = getOutputCoords();
vec4 result = vec4(0.);
result.r = ${o(a.slice())};
if(${r}){
result.g = ${l(a.slice())};
}
if(${s}) {
result.b = ${u(a.slice())};
if(${r}) {
result.a = ${p(a.slice())};
}
}
setOutput(result);
}
`;function o(h){return d(h)}function l(h){return h[n-1]="("+h[n-1]+" + 1)",d(h)}function u(h){return h[n-2]="("+h[n-2]+" + 1)",d(h)}function p(h){return h[n-1]="("+h[n-1]+" + 1)",h[n-2]="("+h[n-2]+" + 1)",d(h)}function d(h){let m=e.map((y,b)=>c(b,h)),f=m.join(","),g=m.slice(-2).join(",");return`getChannel(getX(${f}), vec2(${g}))`}function c(h,m){return t.indexOf(h)!==-1&&e[h]!==1?`${e[h]} - ${m[h]} - 1`:`${m[h]}`}}};function Fae(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{dims:s}=a,i=r.shape.length,o=w.parseAxisParam(s,r.shape);if(i===0)return aa({inputs:{x:r},backend:n});let l=X().getBool("WEBGL_PACK_ARRAY_OPERATIONS")?new Eae(r.shape,o):new Cae(r.shape,o);return n.runWebGLProgram(l,[r],r.dtype)}var Aae={kernelName:so,backendName:"webgl",kernelFunc:Fae},$ae=class{constructor(e,t){this.variableNames=["Image"],this.outputShape=[],this.customUniforms=[{name:"params",type:"vec4"}];let n=e[1],a=e[2];this.outputShape=e;let r="";typeof t=="number"?r=`float outputValue = ${t.toFixed(2)};`:r=`
vec3 fill = vec3(${t.join(",")});
float outputValue = fill[coords[3]];`,this.userCode=`
void main() {
ivec4 coords = getOutputCoords();
int x = coords[2];
int y = coords[1];
float coordXFloat = (float(x) - params[0]) * params[3] -
(float(y) - params[1]) * params[2];
float coordYFloat = (float(x) - params[0]) * params[2] +
(float(y) - params[1]) * params[3];
int coordX = int(round(coordXFloat + params[0]));
int coordY = int(round(coordYFloat + params[1]));
${r}
if(coordX >= 0 && coordX < ${a} && coordY >= 0 && coordY < ${n}) {
outputValue = getImage(coords[0], coordY, coordX, coords[3]);
}
setOutput(outputValue);
}
`}},Dae={kernelName:zu,backendName:"webgl",kernelFunc:({inputs:e,attrs:t,backend:n})=>{let{image:a}=e,{radians:r,fillValue:s,center:i}=t,o=n,l=new $ae(a.shape,s),[u,p]=C.getImageCenter(i,a.shape[1],a.shape[2]),d=[[u,p,Math.sin(r),Math.cos(r)]];return o.runWebGLProgram(l,[a],a.dtype,d)}},Rae=`
// OpenGL ES does not support round function.
// The algorithm is based on banker's rounding.
float base = floor(x);
if ((x - base) < 0.5) {
return floor(x);
} else if ((x - base) > 0.5) {
return ceil(x);
} else {
if (mod(base, 2.0) == 0.0) {
return base;
} else {
return base + 1.0;
}
}
`,Mae=Qe({opSnippet:Rae}),Pae={kernelName:io,backendName:"webgl",kernelFunc:Mae},Oae="return inversesqrt(x);",Lae=Qe({opSnippet:Oae,cpuKernelImpl:i9}),zae={kernelName:oo,backendName:"webgl",kernelFunc:Lae},rE=class{constructor(e,t,n,a,r,s,i=!0){this.variableNames=["updates","indices","defaultValue"],this.outputShape=s;let o=mt(r.length),l=mt(s.length),u="";n===1?u="i":n===2&&(u="i, j");let p=`getIndices(${u})`,d="";a===1?d="i":a===2&&(d="i, coords[1]");let c=`getUpdates(${d})`,h=t>1?"strides[j]":"strides";this.userCode=`
${o} strides = ${o}(${r});
void main() {
${l} coords = getOutputCoords();
float sum = 0.0;
bool found = false;
for (int i = 0; i < ${e}; i++) {
int flattenedIndex = 0;
for (int j = 0; j < ${t}; j++) {
int index = round(${p});
flattenedIndex += index * ${h};
}
if (flattenedIndex == coords[0]) {
sum += ${c};
found = true;
}
}
setOutput(mix(getDefaultValue(), sum, float(found)));
}
`}};function Bae(e){let{inputs:t,backend:n,attrs:a}=e,{indices:r,updates:s}=t,{shape:i}=a,{sliceRank:o,numUpdates:l,sliceSize:u,strides:p,outputSize:d}=C.calculateShapes(s,r,i),c=[d/u,u];if(d===0)return n.makeTensorInfo(i,r.dtype);let h=me({inputs:{x:r},backend:n,attrs:{shape:[l,o]}}),m=me({inputs:{x:s},backend:n,attrs:{shape:[l,u]}}),f=n.makeTensorInfo([],"float32",new Float32Array([0])),g=new rE(l,o,h.shape.length,m.shape.length,p,c),y=n.runWebGLProgram(g,[m,h,f],m.dtype),b=me({inputs:{x:y},backend:n,attrs:{shape:i}});return n.disposeIntermediateTensorInfo(h),n.disposeIntermediateTensorInfo(m),n.disposeIntermediateTensorInfo(y),n.disposeIntermediateTensorInfo(f),b}var Wae={kernelName:Su,backendName:"webgl",kernelFunc:Bae},Vae=class{constructor(e,t,n,a){this.variableNames=["sortedSequence","values"],this.customUniforms=[{name:"numInputs",type:"int"}],this.outputShape=[e,n];let r="while (left < right) {",s=`for (int i = 0; i < ${Math.ceil(Math.log2(t+1))}; ++i) { if (left >= right) break;`,i=X().getNumber("WEBGL_VERSION")===2?r:s,o=a==="left"?"<":"<=";this.userCode=`
int findBound(int batch, float value) {
int left = 0;
int right = numInputs;
int mid;
${i}
mid = (left + right) / 2;
if (getSortedSequence(batch, mid) ${o} value) {
left = mid + 1;
} else {
right = mid;
}
}
return right;
}
void main() {
ivec2 coords = getOutputCoords();
int batch = coords[0];
int valueIndex = coords[1];
float value = getValues(batch, valueIndex);
setOutput(float(findBound(batch, value)));
}
`}};function Uae(e){let{inputs:t,backend:n,attrs:a}=e,{sortedSequence:r,values:s}=t,{side:i}=a,o=new Vae(r.shape[0],r.shape[1],s.shape[1],i),l=[[r.shape[1]]];return n.runWebGLProgram(o,[r,s],"int32",l)}var Gae={kernelName:Km,backendName:"webgl",kernelFunc:Uae},Hae=class{constructor(e,t,n){this.variableNames=["c","a","b"],this.outputShape=t;let a,r;if(n>4)throw Error(`Where for rank ${n} is not yet supported`);if(n===1)r="resRC",a="resRC";else{let i=["resRC.x","resRC.y","resRC.z","resRC.w"],o=[],l=[];for(let u=0;u<t.length;u++)l.push(`${i[u]}`),u<e&&o.push(`${i[u]}`);a=o.join(),r=l.join()}let s=mt(n);this.userCode=`
void main() {
${s} resRC = getOutputCoords();
float cVal = getC(${a});
if (cVal >= 1.0) {
setOutput(getA(${r}));
} else {
setOutput(getB(${r}));
}
}
`}};function jae(e){let{inputs:t,backend:n}=e,{condition:a,t:r,e:s}=t,i=new Hae(a.shape.length,r.shape,r.shape.length);return n.runWebGLProgram(i,[a,r,s],fa(r.dtype,s.dtype))}var qae={kernelName:Nu,backendName:"webgl",kernelFunc:jae},Kae=`
// Stable and Attracting Fixed Point (0, 1) for Normalized Weights.
// see: https://arxiv.org/abs/1706.02515
float scaleAlpha = ${C.SELU_SCALEALPHA};
float scale = ${C.SELU_SCALE};
return (x >= 0.0) ? scale * x : scaleAlpha * (exp(x) - 1.0);
`,Xae=Qe({opSnippet:Kae}),Yae={kernelName:Tu,backendName:"webgl",kernelFunc:Xae},Qae=rp+`
return 1.0 / (1.0 + exp(-1.0 * x));
`,Jae=`
vec4 result = 1.0 / (1.0 + exp(-1.0 * x));
bvec4 isNaN = isnan(x);
result.r = isNaN.r ? x.r : result.r;
result.g = isNaN.g ? x.g : result.g;
result.b = isNaN.b ? x.b : result.b;
result.a = isNaN.a ? x.a : result.a;
return result;
`,Zae=Qe({opSnippet:Qae,packedOpSnippet:Jae,cpuKernelImpl:l9}),ere={kernelName:uo,backendName:"webgl",kernelFunc:Zae},tre=`
if (isnan(x)) { return 0.0; }
return sign(x);
`,nre=Qe({opSnippet:tre}),are={kernelName:Eu,backendName:"webgl",kernelFunc:nre},rre=rp+`
return sin(x);
`,sre=Qe({opSnippet:rre}),ire={kernelName:lo,backendName:"webgl",kernelFunc:sre},ore=`
float e2x = exp(x);
return (e2x - 1.0 / e2x) / 2.0;
`,lre=Qe({opSnippet:ore}),ure={kernelName:Cu,backendName:"webgl",kernelFunc:lre},pre=`
float epsilon = 1.1920928955078125e-7;
float threshold = log(epsilon) + 2.0;
bool too_large = x > -threshold;
bool too_small = x < threshold;
float result;
float exp_x = exp(x);
if (too_large){
result = x;
}
else if (too_small){
result = exp_x;
}
else{
result = log(exp_x + 1.0);
}
return result;
`,cre=Qe({opSnippet:pre}),dre={kernelName:Fu,backendName:"webgl",kernelFunc:cre},hre=e=>{let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{blockShape:s,paddings:i}=a;w.assert(r.shape.length<=4,()=>"spaceToBatchND for rank > 4 with a WebGL backend not implemented yet");let o=s.reduce((y,b)=>y*b),l=[[0,0]];l.push(...i);for(let y=1+s.length;y<r.shape.length;++y)l.push([0,0]);let u=[],p=nE({inputs:{x:r},backend:n,attrs:{paddings:l,constantValue:0}}),d=C.getReshaped(p.shape,s,o,!1),c=C.getPermuted(d.length,s.length,!1),h=C.getReshapedPermuted(p.shape,s,o,!1),m=me({inputs:{x:p},backend:n,attrs:{shape:d}}),f=Wt({inputs:{x:m},backend:n,attrs:{perm:c}}),g=me({inputs:{x:f},backend:n,attrs:{shape:h}});return u.push(p),u.push(m),u.push(f),u.forEach(y=>n.disposeIntermediateTensorInfo(y)),g},mre={kernelName:Au,backendName:"webgl",kernelFunc:hre};function fre(e){let{inputs:t,backend:n}=e,{indices:a,values:r,denseShape:s,defaultValue:i}=t;if(s.shape.length!==1)throw new Error(`Dense shape must be a vector, saw:
${s.shape}`);if(a.shape.length!==2)throw new Error(`Indices must be a matrix, saw:
${a.shape}`);if(r.shape.length!==1)throw new Error(`Values must be a vector, saw:
${r.shape}`);if(i.shape.length!==0)throw new Error(`Default value must be a scalar, saw:
${i.shape}`);let o=n.readSync(a.dataId),l=n.readSync(r.dataId),u=n.readSync(s.dataId),p=n.readSync(i.dataId)[0],[d,c,h,m,f]=p9(o,a.shape,a.dtype,l,r.dtype,u,p);return[n.makeTensorInfo(c,a.dtype,d),n.makeTensorInfo([c[0]],r.dtype,h),n.makeTensorInfo([m.length],"bool",new Uint8Array(m.map(g=>Number(g)))),n.makeTensorInfo([f.length],a.dtype,new Int32Array(f))]}var gre={kernelName:Ec,backendName:"webgl",kernelFunc:fre};function yre(e){let{inputs:t,backend:n}=e,{inputIndices:a,inputShape:r,newShape:s}=t;if(a.shape.length!==2)throw new Error(`Input indices should be a matrix but received shape ${a.shape}`);if(r.shape.length!==1)throw new Error(`Input shape should be a vector but received shape ${r.shape}`);if(s.shape.length!==1)throw new Error(`Target shape should be a vector but received shape ${s.shape}`);let i=Array.from(n.readSync(r.dataId)),o=n.readSync(a.dataId),l=Array.from(n.readSync(s.dataId)),[u,p,d]=c9(o,a.shape,a.dtype,i,l);return[n.makeTensorInfo(p,a.dtype,u),n.makeTensorInfo([d.length],s.dtype,new Int32Array(d))]}var bre={kernelName:Du,backendName:"webgl",kernelFunc:yre};function xre(e){let{inputs:t,backend:n}=e,{data:a,indices:r,segmentIds:s}=t;if(a.shape.length<1)throw new Error("Data should be at least 1 dimensional but received scalar");if(r.shape.length!==1)throw new Error(`Indices should be a vector but received shape
${r.shape}`);if(s.shape.length!==1)throw new Error(`Segment ids should be a vector but received shape
${s.shape}`);let i=n.readSync(a.dataId),o=n.readSync(r.dataId),l=n.readSync(s.dataId),[u,p]=TC(i,a.shape,a.dtype,o,l,!0);return n.makeTensorInfo(p,a.dtype,u)}var vre={kernelName:Fc,backendName:"webgl",kernelFunc:xre};function wre(e){let{inputs:t,backend:n}=e,{data:a,indices:r,segmentIds:s}=t;if(a.shape.length<1)throw new Error("Data should be at least 1 dimensional but received scalar");if(r.shape.length!==1)throw new Error(`Indices should be a vector but received shape
${r.shape}`);if(s.shape.length!==1)throw new Error(`Segment ids should be a vector but received shape
${s.shape}`);let i=n.readSync(a.dataId),o=n.readSync(r.dataId),l=n.readSync(s.dataId),[u,p]=TC(i,a.shape,a.dtype,o,l);return n.makeTensorInfo(p,a.dtype,u)}var kre={kernelName:Ac,backendName:"webgl",kernelFunc:wre};function Ire(e){let{inputs:t,backend:n,attrs:a}=e,{sparseIndices:r,sparseValues:s,defaultValue:i}=t,{outputShape:o}=a,{sliceRank:l,numUpdates:u,sliceSize:p,strides:d,outputSize:c}=C.calculateShapes(s,r,o),h=!1;if(s.dtype==="string"){let y=n.bufferSync(r),b=n.bufferSync(s),x=w.decodeString(n.readSync(i.dataId)[0]),v=o9(y,b,o,c,p,u,l,d,x,h);return n.makeTensorInfo(o,v.dtype,v.values)}let m=new rE(u,l,r.shape.length,s.shape.length,d,[c,1],h),f=n.runWebGLProgram(m,[s,r,i],s.dtype),g=me({inputs:{x:f},backend:n,attrs:{shape:o}});return n.disposeIntermediateTensorInfo(f),g}var Sre={kernelName:Xm,backendName:"webgl",kernelFunc:Ire};function Nre(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{numOrSizeSplits:s,axis:i}=a,o=w.parseAxisParam(i,r.shape)[0],l=C.prepareSplitSize(r,s,o),u=r.shape.length,p=new Array(u).fill(0),d=r.shape.slice();return l.map(c=>{let h=[...d];h[o]=c;let m=sp({inputs:{x:r},backend:n,attrs:{begin:p,size:h}});return p[o]+=c,m})}var Tre={kernelName:$u,backendName:"webgl",kernelFunc:Nre},dI="return sqrt(x);",_re=Qe({opSnippet:dI,packedOpSnippet:dI,cpuKernelImpl:d9}),Cre={kernelName:po,backendName:"webgl",kernelFunc:_re},Ere="return x * x;",Fre=Qe({opSnippet:Ere}),Are={kernelName:$c,backendName:"webgl",kernelFunc:Fre},hI="return (a - b) * (a - b);",$re=cn({opSnippet:hI,packedOpSnippet:hI}),Dre={kernelName:mo,backendName:"webgl",kernelFunc:$re};function Rre({inputs:e,attrs:t,backend:n}){let{x:a}=e,r=Aa+`
return x > 0.0 ? 1.0 : float(${t.alpha});
`,s=new _r(a.shape,r);return n.runWebGLProgram(s,[a],a.dtype)}var Mre={kernelName:ws,backendName:"webgl",kernelFunc:Rre},Pre=class{constructor(e,t,n){this.variableNames=["x"],this.outputShape=n;let a=n.length,r=mt(n.length),s=mt(n.length),i="";if(a===1)i="coords * strides + begin";else{let o=0;i=n.map((l,u)=>(o++,n.length===1?`coords * strides[${u}] + begin[${u}]`:`coords[${o-1}] * strides[${u}] + begin[${u}]`)).join(",")}this.userCode=`
${r} begin = ${r}(${e});
${r} strides = ${r}(${t});
void main() {
${s} coords = getOutputCoords();
setOutput(getX(${i}));
}
`}};function Ore(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{begin:s,end:i,strides:o,beginMask:l,endMask:u,ellipsisMask:p,newAxisMask:d,shrinkAxisMask:c}=a,{finalShapeSparse:h,finalShape:m,isIdentity:f,sliceDim0:g,isSimpleSlice:y,begin:b,end:x,strides:v}=Kt.sliceInfo(r.shape,s,i,o,l,u,p,d,c),k;if(f)k=me({inputs:{x:r},backend:n,attrs:{shape:m}});else if(g||y){w.assert(r.shape.length>=1,()=>`Input must have rank at least 1, got: ${r.shape.length}`);let _=Kt.computeOutShape(b,x,v),E=sp({inputs:{x:r},backend:n,attrs:{begin:b,size:_}});k=me({inputs:{x:E},backend:n,attrs:{shape:m}}),n.disposeIntermediateTensorInfo(E)}else if(n.shouldExecuteOnCPU([r])){let _=n.readSync(r.dataId),E=Ve(r.shape,r.dtype,_),A=h9(h,E,v,b);k=n.makeTensorInfo(m,r.dtype,A.values)}else{let _=new Pre(b,v,h);k=n.runWebGLProgram(_,[r],r.dtype)}let T=me({inputs:{x:k},backend:n,attrs:{shape:m}});return n.disposeIntermediateTensorInfo(k),T}var Lre={kernelName:Ru,backendName:"webgl",kernelFunc:Ore};function zre(e){let{inputs:t,backend:n,attrs:a}=e,{separator:r,nGramWidths:s,leftPad:i,rightPad:o,padWidth:l,preserveShortSequences:u}=a,{data:p,dataSplits:d}=t,c=n.readSync(p.dataId),h=n.readSync(d.dataId),[m,f]=m9(c,h,r,s,i,o,l,u);return[n.makeTensorInfo([m.length],"string",m),n.makeTensorInfo(d.shape,"int32",f)]}var Bre={kernelName:Ym,backendName:"webgl",kernelFunc:zre};function Wre(e){let{inputs:t,backend:n,attrs:a}=e,{skipEmpty:r}=a,{input:s,delimiter:i}=t;if(s.dtype!=="string")throw new Error("Input must be of datatype string");if(s.shape.length!==1)throw new Error(`Input must be a vector, got shape: ${s.shape}`);if(i.shape.length!==0)throw new Error(`Delimiter must be a scalar, got shape: ${i.shape}`);let o=n.readSync(s.dataId),l=n.readSync(i.dataId)[0],[u,p,d]=f9(o,l,r),c=p.length;return[n.makeTensorInfo([c,2],"int32",u),n.makeTensorInfo([c],"string",p),n.makeTensorInfo([2],"int32",new Int32Array(d))]}var Vre={kernelName:Qm,backendName:"webgl",kernelFunc:Wre};function Ure(e){let{inputs:t,backend:n,attrs:a}=e,{numBuckets:r}=a,{input:s}=t;if(s.dtype!=="string")throw new Error("Input must be of datatype string");if(r<=0)throw new Error("Number of buckets must be at least 1");let i=n.readSync(s.dataId),o=g9(i,r);return n.makeTensorInfo(s.shape,"int32",o)}var Gre={kernelName:Jm,backendName:"webgl",kernelFunc:Ure},Hre="return tan(x);",jre=Qe({opSnippet:Hre}),qre={kernelName:go,backendName:"webgl",kernelFunc:jre},Kre=`
float e2x = exp(-2.0 * abs(x));
return sign(x) * (1.0 - e2x) / (1.0 + e2x);
`,Xre=Qe({opSnippet:Kre}),Yre={kernelName:yo,backendName:"webgl",kernelFunc:Xre},Qre=class{constructor(e,t){this.variableNames=["A"];let n=new Array(e.length);for(let s=0;s<n.length;s++)n[s]=e[s]*t[s];this.outputShape=n,this.rank=n.length;let a=mt(this.rank),r=Jre(e);this.userCode=`
void main() {
${a} resRC = getOutputCoords();
setOutput(getA(${r}));
}
`}};function Jre(e){let t=e.length;if(t>5)throw Error(`Tile for rank ${t} is not yet supported`);if(t===1)return`imod(resRC, ${e[0]})`;let n=["resRC.x","resRC.y","resRC.z","resRC.w","resRC.u"],a=[];for(let r=0;r<e.length;r++)a.push(`imod(${n[r]}, ${e[r]})`);return a.join()}function sE(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{reps:s}=a;if(r.dtype==="string"||r.shape.length>5){let o=n.readSync(r.dataId),l=r.dtype==="string"?o.map(d=>w.decodeString(d)):o,u=Ve(r.shape,r.dtype,l),p=b9(u,s);return n.makeTensorInfo(p.shape,p.dtype,p.values)}let i=new Qre(r.shape,s);return n.runWebGLProgram(i,[r],r.dtype)}var Zre={kernelName:vs,backendName:"webgl",kernelFunc:sE},ese=class{constructor(e){this.variableNames=["x","indices"],this.customUniforms=[{name:"n",type:"int"},{name:"firstPass",type:"int"},{name:"negativeInf",type:"float"},{name:"dir",type:"int"},{name:"inc",type:"int"}],this.outputShape=e,this.userCode=`
void main() {
ivec2 coords = getOutputCoords();
int batch = coords[0];
int elemIdx = coords[1];
// We compare elements pair-wise within a group of size 2 * inc.
// The comparing rule for each group alternates between ascending
// and descending. Within each group, we compare each pair at
// positions i and i+inc. To decide whether an element at position i
// is x0 or x1, we mod it by 2 * inc, if the result is smaller than
// inc, it is in the first half of the group, we denote it as x0,
// otherwise we denote it as x1.
// For example, as shown in the Bitonic top K paper referenced above,
// Figure5(a) shows that element[1] is in the
// second half of the group when group size is 2, but it is in the
// first half of the group when group size is 4.
bool isFirstInPair = imod(elemIdx, 2 * inc) < inc;
int i = isFirstInPair ? elemIdx : elemIdx - inc;
int i0 = firstPass == 1 ? i : int(getIndices(batch, i));
int i1 = firstPass == 1 ? i + inc : int(getIndices(batch, i + inc));
float x0 = i0 < n ? getX(batch, i0) : negativeInf;
float x1 = i1 < n ? getX(batch, i1) : negativeInf;
// Denotes which direction indices are in (ascending or descending).
bool reverse = imod(elemIdx, 2 * dir) >= dir;
bool isGreater = x0 > x1 || (x0 == x1 && i1 > i0);
if (reverse == isGreater) { // Elements in opposite order of direction
int iTemp = i0;
i0 = i1;
i1 = iTemp;
}
if (isFirstInPair) {
setOutput(float(i0));
} else {
setOutput(float(i1));
}
}
`}},tse=class{constructor(e){this.variableNames=["x","indices"],this.customUniforms=[{name:"n",type:"int"},{name:"firstPass",type:"int"},{name:"k",type:"int"}],this.outputShape=e,this.userCode=`
void main() {
// Takes max of indices (0, k), (1, k + 1), (2, k + 2) ...
ivec2 coords = getOutputCoords();
int batch = coords[0];
int elemIdx = coords[1];
// The output size is half of the previous size.
// If the previous sequence is | | | | _ _ _ _ | | | | _ _ _ _ (k=4),
// we only need to output the indices at positions |, the indices at
// positions _ can be thrown away, see Figure5(b) After Phase 2
// (Merge phase) in the Bitonic Top K paper referenced above.
// For example, the paper shows we only need to output the orange bars.
// The output sequence should look like this | | | | | | | |.
// Because the sequence is halved, to map the output index back
// to the previous sequence to find the corresponding value,
// we need to double the index. When we double the index,
// we basically interpolate a position, so 2i looks like
// | _ | _ | _ | _ | _ | _ | _. We move the | to the first k position
// of each 2k positions by - elemIdx % k. E.g. for output at
// index 4,5,6,7, we want to get the corresponding element at
// original index 8,9,10,11, for output at index 8,9,10,11,
// we want to get the corresponding element at original index
// 16,17,18,19, so on and so forth.
int i = elemIdx < k ? elemIdx : (elemIdx * 2 - imod(elemIdx, k));
int i0 = firstPass == 1 ? i : int(getIndices(batch, i));
int i1 = firstPass == 1 ? i + k : int(getIndices(batch, i + k));
float x0 = getX(batch, i0);
float x1 = i1 < n ? getX(batch, i1) : x0;
setOutput(x0 >= x1 ? float(i0) : float(i1));
}
`}};function Us(e,t){t!==null&&e.disposeIntermediateTensorInfo(t)}function mI(e){let t=1;for(;t<e;)t*=2;return t}function nse(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{k:s,sorted:i}=a,o=X().getNumber("TOPK_LAST_DIM_CPU_HANDOFF_SIZE_THRESHOLD"),l=X().getNumber("TOPK_K_CPU_HANDOFF_THRESHOLD"),u=r.shape,p=u[u.length-1];if(n.shouldExecuteOnCPU([r])||p<o||s>l){let A=n.readSync(r.dataId),[M,$]=x9(A,u,r.dtype,s,i);return[n.makeTensorInfo(M.shape,M.dtype,M.values),n.makeTensorInfo($.shape,$.dtype,$.values)]}if(s===0)return u[u.length-1]=0,[n.makeTensorInfo(u,r.dtype,[]),n.makeTensorInfo(u,"int32",[])];if(p===1)return[r,md({attrs:{shape:u,dtype:"int32",value:0},backend:n})];let d=n.texData.get(r.dataId),c=d!==null&&d.isPacked,h=c?n.unpackTensor(r):r,m=w.sizeFromShape(u)/p,f=me({inputs:{x:h},attrs:{shape:[m,p]},backend:n});c&&Us(n,h);let g=mI(s),y=mI(p),b=null,x=()=>b===null?[f,f]:[f,b],v=(A,M,$)=>{let S=x(),P=new ese($),V=[[p],[b===null?1:0],[Number.NEGATIVE_INFINITY],[A],[M]],j=b;b=n.runWebGLProgram(P,S,"int32",V),Us(n,j)};for(let A=1;A<g;A*=2){let M=A*2;for(let $=A;$>=1;$/=2)v(M,$,[m,y])}for(let A=y;A>g;A/=2){let M=x(),$=new tse([m,A/2]),S=[[p],[b===null?1:0],[g]],P=b;b=n.runWebGLProgram($,M,"int32",S),Us(n,P);let V=g/2,j=V*2;for(let q=V;q>=1;q/=2)v(j,q,b.shape)}let k=b;b=sp({inputs:{x:b},backend:n,attrs:{begin:0,size:[m,s]}}),Us(n,k);let T=YC({inputs:{x:f,indices:b},backend:n,attrs:{axis:1,batchDims:1}});Us(n,f);let _=u.slice(0,-1);_.push(s),k=b,b=me({inputs:{x:b},attrs:{shape:_},backend:n}),Us(n,k);let E=T;return T=me({inputs:{x:T},attrs:{shape:_},backend:n}),Us(n,E),[T,b]}var ase={kernelName:Mu,backendName:"webgl",kernelFunc:nse},rse=class{constructor(e,t,n,a,r,s){this.variableNames=["Image","Transforms"],this.outputShape=s;let i=n==="nearest"?1:2,o;switch(a){case"constant":o=1;break;case"reflect":o=2;break;case"wrap":o=3;break;case"nearest":o=4;break;default:o=1;break}this.userCode=`
float mapCoord(float outCoord, float len) {
float inCoord = outCoord;
if(${o} == 2) {
if (inCoord < 0.0) {
if (len <= 1.0) {
inCoord = 0.0;
} else {
float sz2 = 2.0 * len;
if (inCoord < sz2) {
inCoord = sz2 * float(int(float(-inCoord / sz2))) +
inCoord;
}
inCoord = inCoord < -len ? inCoord + sz2 : -inCoord - 1.0;
}
} else if (inCoord > len - 1.0) {
if (len <= 1.0) {
inCoord = 0.0;
} else {
float sz2 = 2.0 * len;
inCoord -= sz2 * float(int(float(inCoord / sz2)));
if (inCoord >= len) {
inCoord = sz2 - inCoord - 1.0;
}
}
}
return clamp(inCoord, 0.0, len - 1.0);
} else if (${o} == 3) {
if (inCoord < 0.0) {
if (len <= 1.0) {
inCoord = 0.0;
} else {
float sz = len - 1.0;
inCoord += len * (float(int(float(-inCoord / sz))) + 1.0);
}
} else if (inCoord > len - 1.0) {
if (len <= 1.0) {
inCoord = 0.0;
} else {
float sz = len - 1.0;
inCoord -= len * float(int(float(inCoord / sz)));
}
}
return clamp(inCoord, 0.0, len - 1.0);
} else if (${o} == 4) {
return clamp(outCoord, 0.0, len - 1.0);
} else {
return outCoord;
}
}
float readWithFillValue(int batch, int coordY, int coordX,
int channel) {
float outputValue;
if (0 <= coordY && coordY < ${e} && 0 <= coordX && coordX < ${t}) {
outputValue = getImage(batch, coordY, coordX, channel);
} else {
outputValue = float(${r});
}
return outputValue;
}
void main() {
ivec4 coords = getOutputCoords();
float outputValue;
int batch = coords[0];
int x = coords[2];
int y = coords[1];
int channel = coords[3];
float xf = float(x);
float yf = float(y);
float a1 = getTransforms(batch, 0);
float a2 = getTransforms(batch, 1);
float a3 = getTransforms(batch, 2);
float b1 = getTransforms(batch, 3);
float b2 = getTransforms(batch, 4);
float b3 = getTransforms(batch, 5);
float c1 = getTransforms(batch, 6);
float c2 = getTransforms(batch, 7);
float projection = c1 * xf + c2 * yf + 1.0;
if (projection == 0.0) {
outputValue = float(${r});
} else {
float inX = (a1 * xf + a2 * yf + a3) / projection;
float inY = (b1 * xf + b2 * yf + b3) / projection;
float mapX = mapCoord(inX, float(${t}));
float mapY = mapCoord(inY, float(${e}));
if (${i} == 1) {
int coordY = int(round(mapY));
int coordX = int(round(mapX));
outputValue = readWithFillValue(batch, coordY, coordX,
channel);
} else {
float yFloor = floor(mapY);
float xFloor = floor(mapX);
float yCeil = yFloor + 1.0;
float xCeil = xFloor + 1.0;
float valueYFloor = (xCeil - mapX) *
readWithFillValue(batch, int(yFloor), int(xFloor), channel) +
(mapX - xFloor) *
readWithFillValue(batch, int(yFloor), int(xCeil), channel);
float valueYCeil = (xCeil - mapX) *
readWithFillValue(batch, int(yCeil), int(xFloor), channel) +
(mapX - xFloor) *
readWithFillValue(batch, int(yCeil), int(xCeil), channel);
outputValue = (yCeil - mapY) * valueYFloor +
(mapY - yFloor) * valueYCeil;
}
}
setOutput(outputValue);
}
`}};function sse(e){let{inputs:t,backend:n,attrs:a}=e,{image:r,transforms:s}=t,{interpolation:i,fillMode:o,fillValue:l,outputShape:u}=a,[p,d,c,h]=r.shape,[m,f]=u!=null?u:[d,c],g=[p,m,f,h],y=new rse(d,c,i,o,l,g);return n.runWebGLProgram(y,[r,s],"float32")}var ise={kernelName:Pu,backendName:"webgl",kernelFunc:sse};function ose(e){let{inputs:t,attrs:n,backend:a}=e,{axis:r}=n,{x:s}=t;Zu(s,"unique"),console.warn("WARNING: ","UI might be locked temporarily as data is being downloaded");let i=a.readSync(s.dataId),{outputValues:o,outputShape:l,indices:u}=v9(i,r,s.shape,s.dtype);return[a.makeTensorInfo(l,s.dtype,o),a.makeTensorInfo([u.length],"int32",u)]}var lse={kernelName:Zm,backendName:"webgl",kernelFunc:ose};function use(e){let{inputs:t,backend:n,attrs:a}=e,{value:r}=t,{axis:s}=a;s<0&&(s+=r.shape.length);let i=r,o=i.shape.length,l=r.shape[s],u=new Array(o-1),p=0;for(let f=0;f<o;f++)f!==s&&(u[p++]=i.shape[f]);let d=[],c=new Array(o).fill(0),h=i.shape.slice();h[s]=1;let m=new Array(l);for(let f=0;f<m.length;f++){c[s]=f;let g=sp({inputs:{x:i},backend:n,attrs:{begin:c,size:h}}),y=me({inputs:{x:g},backend:n,attrs:{shape:u}});m[f]=y,d.push(g)}return d.forEach(f=>n.disposeIntermediateTensorInfo(f)),m}var pse={kernelName:Ou,backendName:"webgl",kernelFunc:use},cse=class{constructor(e,t){this.variableNames=["x","segmentIds"];let n=e.windowSize,a=e.batchSize,r=e.inSize,s=e.numSegments,i=s*Math.ceil(r/n);this.outputShape=[a,i];let o="0.0",l="sumValue",u=Math.floor(n/4)*4,p=n%4,d=`
sumValue += dot(values, segFilter);
`,c="";r%n>0&&(c=`
if (inIdx < 0 || inIdx >= ${r}) {
return initializationValue;
}
`);let h="";r%n>0&&(h=`
if (inIdx < 0 || inIdx >= ${r}) {
return -1.0;
}
`),this.userCode=`
const float initializationValue = ${o};
float getValue(int batch, int inIdx) {
${c}
return getX(batch, inIdx);
}
float getSegmentIdAtIndex(int inIdx) {
${h}
return getSegmentIds(inIdx);
}
void main() {
ivec2 coords = getOutputCoords();
int batch = coords[0];
int outIdx = coords[1];
int inOffset = int(floor(float(outIdx) / float(
${s})) * float(${n}));
int currentSeg = int(mod(float(outIdx), float(${s})));
float sumValue = 0.0;
for (int i = 0; i < ${u}; i += 4) {
int inIdx = inOffset + i;
vec4 values = vec4(
getValue(batch, inIdx),
getValue(batch, inIdx + 1),
getValue(batch, inIdx + 2),
getValue(batch, inIdx + 3)
);
vec4 segFilter = vec4(
int(getSegmentIdAtIndex(inIdx)) == currentSeg ? 1 : 0,
int(getSegmentIdAtIndex(inIdx + 1)) == currentSeg ? 1 : 0,
int(getSegmentIdAtIndex(inIdx + 2)) == currentSeg ? 1 : 0,
int(getSegmentIdAtIndex(inIdx + 3)) == currentSeg ? 1 : 0
);
${d}
}
int inIdx = inOffset + ${u};
if (${p===1}) {
vec4 values = vec4(
getValue(batch, inIdx),
initializationValue,
initializationValue,
initializationValue
);
int inIdxSeg = int(getSegmentIdAtIndex(inIdx));
vec4 segFilter = vec4(
int(getSegmentIdAtIndex(inIdx)) == currentSeg ? 1 : 0,
0,
0,
0
);
${d}
} else if (${p===2}) {
vec4 values = vec4(
getValue(batch, inIdx),
getValue(batch, inIdx + 1),
initializationValue,
initializationValue
);
vec4 segFilter = vec4(
int(getSegmentIdAtIndex(inIdx)) == currentSeg ? 1 : 0,
int(getSegmentIdAtIndex(inIdx + 1)) == currentSeg ? 1 : 0,
0,
0
);
${d}
} else if (${p===3}) {
vec4 values = vec4(
getValue(batch, inIdx),
getValue(batch, inIdx + 1),
getValue(batch, inIdx + 2),
initializationValue
);
vec4 segFilter = vec4(
int(getSegmentIdAtIndex(inIdx)) == currentSeg ? 1 : 0,
int(getSegmentIdAtIndex(inIdx + 1)) == currentSeg ? 1 : 0,
int(getSegmentIdAtIndex(inIdx + 2)) == currentSeg ? 1 : 0,
0
);
${d}
}
setOutput(${l});
}
`}};function dse(e){let{inputs:t,backend:n,attrs:a}=e,{x:r,segmentIds:s}=t,{numSegments:i}=a,o=r.shape.length,l=[],u=0,p=C.getAxesPermutation([u],o),d=r;p!=null&&(d=Wt({inputs:{x:r},backend:n,attrs:{perm:p}}),l.push(d),u=C.getInnerMostAxes(1,o)[0]);let c=C.segment_util.computeOutShape(d.shape,u,i),h=w.sizeFromShape([d.shape[u]]),m=me({inputs:{x:d},backend:n,attrs:{shape:[-1,h]}});l.push(m);let f=tf(r.dtype),g=(v,k,T,_,E)=>{let A=v.shape[0],M=v.shape[1],$=C.segment_util.segOpComputeOptimalWindowSize(M,E),S={windowSize:$,inSize:M,batchSize:A,numSegments:E},P=new cse(S,k),V=n.compileAndRun(P,[v,T],_);if(l.push(V),V.shape[1]===E)return V;let j=aE({backend:n,attrs:{start:0,stop:E,step:1,dtype:"float32"}}),q=sE({inputs:{x:j},backend:n,attrs:{reps:[M/$]}});return l.push(j),l.push(q),g(V,k,q,_,E)},y=g(m,"unsortedSegmentSum",s,f,i),b=me({inputs:{x:y},backend:n,attrs:{shape:c}}),x=b;if(p!=null){l.push(b);let v=C.getUndoAxesPermutation(p);x=Wt({inputs:{x},backend:n,attrs:{perm:v}})}return l.forEach(v=>n.disposeIntermediateTensorInfo(v)),x}var hse={kernelName:Dc,backendName:"webgl",kernelFunc:dse},mse=[fQ,yQ,vQ,IQ,NQ,CQ,FQ,$Q,PQ,LQ,WQ,GQ,qQ,QQ,eJ,nJ,rJ,lJ,pJ,dJ,gJ,IJ,NJ,_J,DJ,MJ,zJ,Y9,VJ,qJ,QJ,aZ,sZ,oZ,uZ,cZ,mZ,yZ,vZ,kZ,SZ,TZ,EZ,AZ,MZ,OZ,BZ,UZ,HZ,XZ,ZZ,aee,iee,uee,pee,dee,mee,gee,bee,vee,See,_ee,Fee,$ee,Mee,Lee,Vee,jee,X9,Kee,HJ,Qee,ete,ate,J9,ote,cte,hte,yte,vte,Ste,_te,Ate,Mte,Lte,Bte,Gte,jte,Kte,Jte,ene,nne,rne,ine,pne,mne,bne,Tne,aQ,Fne,Dne,Pne,zne,EJ,Vne,Gne,jne,Xne,Zne,eQ,tae,nae,FJ,kne,sae,uae,hae,sQ,yae,vae,Sae,_ae,Aae,Dae,Pae,zae,Wae,Gae,qae,Yae,ere,are,ire,ure,wJ,Sne,dre,mre,gre,bre,vre,kre,Sre,Tre,Cre,Are,Dre,Mre,Lre,Bre,Vre,Gre,Ine,dQ,qre,Yre,Zre,ase,ise,hQ,lse,pse,hse,Une];for(let e of mse)Rc(e);var $t;(function(e){e[e.float32=0]="float32",e[e.int32=1]="int32",e[e.bool=2]="bool",e[e.string=3]="string",e[e.complex64=4]="complex64"})($t||($t={}));var fc;(function(e){e[e.linear=0]="linear",e[e.relu=1]="relu",e[e.relu6=2]="relu6",e[e.prelu=3]="prelu",e[e.leakyrelu=4]="leakyrelu",e[e.sigmoid=5]="sigmoid",e[e.elu=6]="elu"})(fc||(fc={}));var iE;function fse(e){iE=e.wasm.cwrap(ai,null,["number","array","number","number","array","number","number","number","number","number","number","number","number"])}function gse(e){let{inputs:t,backend:n,attrs:a}=e,{a:r,b:s,bias:i,preluActivationWeights:o}=t;if(r.dtype!=="float32"||s.dtype!=="float32")throw new Error("_FusedMatMul for non non-float32 tensors not yet supported.");let{transposeA:l,transposeB:u,activation:p,leakyreluAlpha:d}=a,c=n.dataIdMap.get(r.dataId).id,h=n.dataIdMap.get(s.dataId).id,m=0;if(i!=null){let E=n.dataIdMap.get(i.dataId);if(E.shape.length!==1)throw new Error(`_FusedMatMul only supports rank-1 bias but got rank ${E.shape.length}.`);m=E.id}let f=o==null?0:n.dataIdMap.get(o.dataId).id,g=fc[p];if(g==null)throw new Error(`${p} activation not yet supported for FusedConv2D in the wasm backend.`);let y=l?r.shape[2]:r.shape[1],b=u?s.shape[1]:s.shape[2],x=Bu.assertAndGetBroadcastShape(r.shape.slice(0,-2),s.shape.slice(0,-2)),v=n.makeOutput([...x,y,b],r.dtype),k=n.dataIdMap.get(v.dataId).id,T=new Uint8Array(new Int32Array(r.shape).buffer),_=new Uint8Array(new Int32Array(s.shape).buffer);return iE(c,T,r.shape.length,h,_,s.shape.length,l,u,g,m,f,d||0,k),v}var yse={kernelName:ai,backendName:"wasm",setupFunc:fse,kernelFunc:gse};function dn(e,t){let n;function a(s){n=s.wasm.cwrap(e,null,["number","number","number"])}function r(s){let{backend:i,inputs:{x:o}}=s,l=i.dataIdMap.get(o.dataId).id,u=i.makeOutput(o.shape,t||o.dtype),p=i.dataIdMap.get(u.dataId).id;return w.sizeFromShape(u.shape)===0||n(l,$t[o.dtype],p),u}return{kernelName:e,backendName:"wasm",setupFunc:a,kernelFunc:r}}var bse=dn(Ll);function Fn(e,t,n){let a;function r(i){a=i.wasm.cwrap(e,null,["number","array","number","number","array","number","number","number"])}function s(i){let{backend:o,inputs:l}=i,{a:u,b:p}=l,d=o.dataIdMap.get(u.dataId).id,c=o.dataIdMap.get(p.dataId).id,h=n!=null?n:u.dtype,m=C.assertAndGetBroadcastShape(u.shape,p.shape),f=o.makeOutput(m,h);if(w.sizeFromShape(m)===0)return f;let g=new Uint8Array(new Int32Array(u.shape).buffer),y=new Uint8Array(new Int32Array(p.shape).buffer),b=o.dataIdMap.get(f.dataId).id;return a(d,g,u.shape.length,c,y,p.shape.length,$t[u.dtype],b),f}return{kernelName:e,backendName:"wasm",setupFunc:r,kernelFunc:s}}var xse=!0,vse=Fn(bs,xse),oE;function wse(e){oE=e.wasm.cwrap(wi,null,["array","number","number","number"])}function kse(e){let{inputs:t,backend:n}=e,a=n.makeOutput(t[0].shape,t[0].dtype);if(w.sizeFromShape(a.shape)===0)return a;let r=t.map(o=>n.dataIdMap.get(o.dataId).id),s=new Uint8Array(new Int32Array(r).buffer),i=n.dataIdMap.get(a.dataId).id;return oE(s,r.length,$t[a.dtype],i),a}var Ise={kernelName:wi,backendName:"wasm",setupFunc:wse,kernelFunc:kse};function cg(e){let{inputs:{x:t},backend:n}=e,a=n.makeOutput(t.shape,t.dtype),r=n.typedArrayFromHeap(t);return n.typedArrayFromHeap(a).set(r),a}var Sse={kernelName:Bi,backendName:"wasm",kernelFunc:cg},lE;function Nse(e){lE=e.wasm.cwrap(Cr,null,["number","array","number","number","number","array","number"])}function gs(e){let{inputs:t,backend:n,attrs:a}=e,[r,s]=_se(t.x.shape,a.perm),i=!0;for(let m=0;m<s.length;m++)s[m]!==m&&(i=!1);let o=Tse(t.x.shape,a.perm),l={dataId:t.x.dataId,shape:r,dtype:t.x.dtype};if(i){let m=cg({inputs:t,backend:n});return m.shape=o,m}let u=n.makeOutput(o,l.dtype),p=n.dataIdMap.get(l.dataId).id,d=n.dataIdMap.get(u.dataId).id,c=new Uint8Array(new Int32Array(s).buffer),h=new Uint8Array(new Int32Array(l.shape).buffer);return lE(p,h,l.shape.length,$t[l.dtype],d,c,s.length),u}function Tse(e,t){let n=new Array(e.length);for(let a=0;a<n.length;a++)n[a]=e[t[a]];return n}function _se(e,t){let n=[],a=[];for(let r=0;r<e.length;++r)e[r]!==1&&n.push(e[r]),e[t[r]]!==1&&a.push(t[r]);for(let r=0;r<a.length;++r){let s=-1;for(let i=0;i<a.length;++i)a[i]>=r&&(s===-1||a[s]>a[i])&&(s=i);a[s]=r}return[n,a]}var Cse={kernelName:Cr,backendName:"wasm",kernelFunc:gs,setupFunc:Nse};function Es(e,t,n){let a=e.shape,r=e.shape.length,s=w.parseAxisParam(t,a),i=s,o=C.getAxesPermutation(i,r),l=null,u=!1;if(o!=null){let p=new Array(r);for(let c=0;c<p.length;c++)p[c]=a[o[c]];i=C.getInnerMostAxes(i.length,r),l=gs({inputs:{x:e},attrs:{perm:o},backend:n});let d=n.dataIdMap.get(e.dataId).id;n.dataIdMap.get(l.dataId).id!==d&&(u=!0)}return{transposed:l,originalAxes:s,axes:i,inputWasTransposed:u}}var uE;function Ese(e){uE=e.wasm.cwrap(Wl,null,["number, number, number"])}function Fse(e){let{backend:t,inputs:n,attrs:a}=e,{axis:r,keepDims:s}=a,{x:i}=n,o=t.dataIdMap.get(i.dataId).id,l=i,{transposed:u,axes:p,originalAxes:d,inputWasTransposed:c}=Es(i,r,t);if(c){let b=t.dataIdMap.get(u.dataId).id;l=u,o=b}let h=l.shape.length;C.assertAxesAreInnerMostDims("all",p,h);let[m,f]=C.computeOutAndReduceShapes(l.shape,p),g=w.sizeFromShape(f),y=t.makeOutput(m,i.dtype);if(w.sizeFromShape(l.shape)!==0){let b=t.dataIdMap.get(y.dataId).id;uE(o,g,b)}if(c&&t.disposeData(u.dataId),s){let b=C.expandShapeToKeepDim(y.shape,d);y.shape=b}return y}var Ase={kernelName:Wl,backendName:"wasm",setupFunc:Ese,kernelFunc:Fse},pE;function $se(e){pE=e.wasm.cwrap(Vl,null,["number, number, number"])}function Dse(e){let{backend:t,inputs:n,attrs:a}=e,{axis:r,keepDims:s}=a,{x:i}=n,o=t.dataIdMap.get(i.dataId).id,l=i,{transposed:u,axes:p,originalAxes:d,inputWasTransposed:c}=Es(i,r,t);if(c){let b=t.dataIdMap.get(u.dataId).id;l=u,o=b}let h=l.shape.length;C.assertAxesAreInnerMostDims("any",p,h);let[m,f]=C.computeOutAndReduceShapes(l.shape,p),g=w.sizeFromShape(f),y=t.makeOutput(m,i.dtype);if(w.sizeFromShape(l.shape)!==0){let b=t.dataIdMap.get(y.dataId).id;pE(o,g,b)}if(c&&t.disposeData(u.dataId),s){let b=C.expandShapeToKeepDim(y.shape,d);y.shape=b}return y}var Rse={kernelName:Vl,backendName:"wasm",setupFunc:$se,kernelFunc:Dse},cE;function Mse(e){cE=e.wasm.cwrap(ki,null,["number","number","number","number","number"])}function Pse(e){let{backend:t,inputs:n,attrs:a}=e,{axis:r}=a,{x:s}=n,i=t.dataIdMap.get(s.dataId).id,o=i,l=s,{transposed:u,axes:p,inputWasTransposed:d}=Es(s,r,t);if(d){let y=t.dataIdMap.get(u.dataId).id;y!==i&&(l=u,o=y)}let c=l.shape.slice(0,-1),h=t.makeOutput(c,"int32"),m=t.dataIdMap.get(h.dataId).id,f=w.sizeFromShape(h.shape),g=l.shape[p[0]];return cE(o,$t[l.dtype],f,g,m),d&&t.disposeData(u.dataId),h}var Ose={kernelName:ki,backendName:"wasm",kernelFunc:Pse,setupFunc:Mse},dE;function Lse(e){dE=e.wasm.cwrap(Ii,null,["number","number","number","number","number","number","number","number","number","number","number","number","number","number"])}function zse(e){let{inputs:t,attrs:n,backend:a}=e,r=t.x,s=a.dataIdMap.get(r.dataId).id,{filterSize:i,strides:o,pad:l,dimRoundingMode:u}=n,p=C.computePool2DInfo(r.shape,i,o,1,l,u),d=p.filterHeight,c=p.filterWidth,h=p.padInfo.top,m=p.padInfo.right,f=p.padInfo.bottom,g=p.padInfo.left,y=p.strideHeight,b=p.strideWidth,x=p.inChannels;if(p.dataFormat!=="channelsLast")throw new Error(`wasm backend does not support dataFormat:'${p.dataFormat}'. Please use 'channelsLast'.`);if(p.dilationWidth!==1||p.dilationHeight!==1)throw new Error(`was backend only supports average pooling with dilation = [1, 1], got [${p.dilationHeight}, ${p.dilationWidth}].`);let v=a.makeOutput(p.outShape,"float32"),k=a.dataIdMap.get(v.dataId).id;return dE(s,r.shape[0],r.shape[1],r.shape[2],d,c,h,m,f,g,y,b,x,k),v}var Bse={kernelName:Ii,backendName:"wasm",setupFunc:Lse,kernelFunc:zse};function Bn(e){let{inputs:t,attrs:n}=e,{x:a}=t,{shape:r}=n,s=w.sizeFromShape(a.shape),i=w.inferFromImplicitShape(r,s);return w.assert(s===w.sizeFromShape(i),()=>`new shape: ${i}, old shape: ${a.shape}. New shape and old shape must have the same number of elements.`),e.backend.incRef(a.dataId),{dataId:a.dataId,shape:i,dtype:a.dtype}}var Wse={kernelName:Iu,backendName:"wasm",kernelFunc:Bn},hE;function Vse(e){hE=e.wasm.cwrap(Si,null,["number","array","number","number","array","number","number","number","number"])}function Use(e){let{inputs:t,backend:n,attrs:a}=e,{a:r,b:s}=t,{transposeA:i,transposeB:o}=a;if(r.dtype!=="float32"||s.dtype!=="float32")throw new Error("BatchMatMul for non non-float32 tensors not yet supported.");let l=r.shape.length,u=s.shape.length,p=i?r.shape[l-2]:r.shape[l-1],d=o?s.shape[u-1]:s.shape[u-2],c=i?r.shape[l-1]:r.shape[l-2],h=o?s.shape[u-2]:s.shape[u-1],m=r.shape.slice(0,-2),f=s.shape.slice(0,-2),g=w.sizeFromShape(m),y=w.sizeFromShape(f),b=Bu.assertAndGetBroadcastShape(r.shape.slice(0,-2),s.shape.slice(0,-2)).concat([c,h]);w.assert(p===d,()=>`Error in matMul: inner shapes (${p}) and (${d}) of Tensors with shapes ${r.shape} and ${s.shape} and transposeA=${i} and transposeB=${o} must match.`);let x=i?[g,p,c]:[g,c,p],v=o?[y,h,d]:[y,d,h],k=Bn({inputs:{x:r},backend:n,attrs:{shape:x}}),T=Bn({inputs:{x:s},backend:n,attrs:{shape:v}}),_=n.dataIdMap.get(k.dataId).id,E=n.dataIdMap.get(T.dataId).id,A=i?k.shape[2]:k.shape[1],M=o?T.shape[1]:T.shape[2],$=Math.max(g,y),S=n.makeOutput([$,A,M],k.dtype),P=n.dataIdMap.get(S.dataId).id,V=new Uint8Array(new Int32Array(k.shape).buffer),j=new Uint8Array(new Int32Array(T.shape).buffer);return hE(_,V,k.shape.length,E,j,T.shape.length,i,o,P),n.disposeData(k.dataId),n.disposeData(T.dataId),S.shape=b,S}var Gse={kernelName:Si,backendName:"wasm",setupFunc:Vse,kernelFunc:Use};function bi(e){let{inputs:{x:t},attrs:{begin:n,size:a},backend:r}=e,[s,i]=Kt.parseSliceParams(t,n,a),o=Kt.isSliceContinous(t.shape,s,i),l=r.readSync(t.dataId),u=r.makeOutput(i,t.dtype),p=w.computeStrides(t.shape),d=r.dataIdMap.get(u.dataId);if(o){let m=Kt.computeFlatOffset(s,p);return t.dtype==="string"?d.stringBytes=l.slice(m,m+w.sizeFromShape(i)):r.typedArrayFromHeap(u).set(l.subarray(m,m+w.sizeFromShape(i))),u}if(t.dtype==="string"){let m=dm(l,s,i,t.shape,t.dtype);return d.stringBytes=m,u}let c=r.typedArrayFromHeap(u),h=t.shape.length;if(h===2)Hse(l,p[0],c,s,i);else if(h===3)jse(l,p[0],p[1],c,s,i);else if(h===4)qse(l,p[0],p[1],p[2],c,s,i);else{let m=dm(l,s,i,t.shape,t.dtype);c.set(m)}return u}function Hse(e,t,n,a,r){let s=0,i=a[0],o=a[1],l=i+r[0];for(let u=i;u<l;u++){let p=u*t+o;n.set(e.subarray(p,p+r[1]),s),s+=r[1]}}function jse(e,t,n,a,r,s){let i=0,o=r[0],l=r[1],u=r[2],p=o+s[0],d=l+s[1];for(let c=o;c<p;c++)for(let h=l;h<d;h++){let m=c*t+h*n+u;a.set(e.subarray(m,m+s[2]),i),i+=s[2]}}function qse(e,t,n,a,r,s,i){let o=0,l=s[0],u=s[1],p=s[2],d=l+i[0],c=u+i[1],h=p+i[2],m=s[3];for(let f=l;f<d;f++)for(let g=u;g<c;g++)for(let y=p;y<h;y++){let b=f*t+g*n+y*a+m;r.set(e.subarray(b,b+i[3]),o),o+=i[3]}}var Kse={kernelName:_u,backendName:"wasm",kernelFunc:bi};function Xse(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{blockShape:s,crops:i}=a,o=s.reduce((y,b)=>y*b),l=C.getReshaped(r.shape,s,o),u=C.getPermuted(l.length,s.length),p=C.getReshapedPermuted(r.shape,s,o),d=C.getSliceBeginCoords(i,s.length),c=C.getSliceSize(p,i,s.length),h=Bn({inputs:{x:r},backend:n,attrs:{shape:l}}),m=gs({inputs:{x:h},backend:n,attrs:{perm:u}}),f=Bn({inputs:{x:m},backend:n,attrs:{shape:p}}),g=bi({inputs:{x:f},backend:n,attrs:{begin:d,size:c}});return n.disposeData(h.dataId),n.disposeData(m.dataId),n.disposeData(h.dataId),g}var Yse={kernelName:Kl,backendName:"wasm",kernelFunc:Xse};function fd(e){let{inputs:{x:t},attrs:{dtype:n},backend:a}=e,r=a.makeOutput(t.shape,n),s=a.typedArrayFromHeap(t);return a.typedArrayFromHeap(r).set(s),r}var Qse={kernelName:Ni,backendName:"wasm",kernelFunc:fd},Jse=dn(Ti),mE;function Zse(e){mE=e.wasm.cwrap(xs,null,["number","number","number","number"])}function eie(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{clipValueMin:s,clipValueMax:i}=a,o=n.dataIdMap.get(r.dataId).id,l=n.makeOutput(r.shape,r.dtype),u=n.dataIdMap.get(l.dataId).id;return mE(o,s,i,u),l}var tie={kernelName:xs,backendName:"wasm",setupFunc:Zse,kernelFunc:eie};function fE(e){let{inputs:t,backend:n}=e,a=w.parseAxisParam(e.attrs.axis,t[0].shape)[0],r=C.computeOutShape(t.map(h=>h.shape),a),s=t.filter(h=>w.sizeFromShape(h.shape)>0);if(s.length===1)return cg({inputs:{x:s[0]},backend:n});let i=n.makeOutput(r,t[0].dtype);if(w.sizeFromShape(r)===0)return i;let o=s.map(h=>h.shape);if(C.assertParamsConsistent(o,a),s[0].dtype==="string"){let h=s.map(x=>{let v=w.sizeFromShape(x.shape.slice(a));return Bn({inputs:{x},backend:n,attrs:{shape:[-1,v]}})}),m=h.map(x=>({vals:n.readSync(x.dataId),shape:x.shape}));r=C.computeOutShape(h.map(x=>x.shape),1);let f=h[0].shape[0]===1,g=k0(m,r,t[0].dtype,f),y=C.computeOutShape(s.map(x=>x.shape),a);i.shape=y;let b=n.dataIdMap.get(i.dataId);return b.stringBytes=C.fromStringArrayToUint8(g),h.forEach(x=>n.disposeData(x.dataId)),i}let l=w.sizeFromShape(s[0].shape.slice(0,a)),u=0,p=s.map(h=>{let m=w.sizeFromShape(h.shape.slice(a));return u+=m,m}),d=s.map(h=>n.typedArrayFromHeap(h)),c=n.typedArrayFromHeap(i);for(let h=0;h<l;h++){let m=h*u;for(let f=0;f<d.length;f++){let g=p[f],y=h*g,b=d[f].subarray(y,y+g);c.set(b,m),m+=g}}return i}var nie={kernelName:Xl,backendName:"wasm",kernelFunc:fE},gE;function aie(e){gE=e.wasm.cwrap(_i,null,["number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number"])}function rie(e){let{inputs:t,attrs:n,backend:a}=e,{x:r,filter:s}=t,i=a.dataIdMap.get(r.dataId).id,o=a.dataIdMap.get(s.dataId).id,{strides:l,dilations:u,pad:p,dimRoundingMode:d,dataFormat:c}=n,h=C.convertConv2DDataFormat(c),m=C.computeConv2DInfo(r.shape,s.shape,l,u,p,d,!1,h),f=m.filterHeight,g=m.filterWidth,y=m.padInfo.top,b=m.padInfo.right,x=m.padInfo.bottom,v=m.padInfo.left,k=m.dilationHeight,T=m.dilationWidth,_=m.strideHeight,E=m.strideWidth,A=m.inChannels,M=m.outChannels,$=m.padInfo.type==="SAME"?1:0;if(m.dataFormat!=="channelsLast")throw new Error(`wasm backend Conv2D does not support dataFormat:'${m.dataFormat}'. Please use 'channelsLast'.`);let S=a.makeOutput(m.outShape,"float32"),P=a.dataIdMap.get(S.dataId).id;return gE(i,r.shape[0],r.shape[1],r.shape[2],o,f,g,y,b,x,v,$,k,T,_,E,A,M,P),S}var sie={kernelName:_i,backendName:"wasm",setupFunc:aie,kernelFunc:rie},yE;function iie(e){yE=e.wasm.cwrap(Ci,null,["number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number"])}function oie(e){let{backend:t,inputs:n,attrs:a}=e,{dy:r,filter:s}=n,{strides:i,pad:o,dataFormat:l,dimRoundingMode:u,inputShape:p}=a,d=1,c=C.convertConv2DDataFormat(l),h=C.computeConv2DInfo(p,s.shape,i,d,o,u,!1,c),{batchSize:m,filterHeight:f,filterWidth:g,inChannels:y,inHeight:b,inWidth:x,outChannels:v,outHeight:k,outWidth:T,strideHeight:_,strideWidth:E}=h,A=f-1-h.padInfo.top,M=g-1-h.padInfo.left,$=h.dataFormat==="channelsLast",S=w.computeStrides(h.inShape),P=w.computeStrides(r.shape),[V,j,q]=w.computeStrides(s.shape),K=S[0],Z=$?S[1]:S[2],ee=$?S[2]:1,re=$?1:S[1],Y=P[0],ie=$?P[1]:P[2],ae=$?P[2]:1,le=$?1:P[1],ue=t.makeOutput(h.inShape,"float32"),ke=t.dataIdMap.get(ue.dataId).id,ye=t.dataIdMap.get(r.dataId).id,Ie=t.dataIdMap.get(s.dataId).id;return yE(ye,Ie,m,f,g,b,x,y,k,T,v,_,E,A,M,V,j,q,K,Z,ee,re,Y,ie,ae,le,ke),ue}var lie={kernelName:Ci,backendName:"wasm",setupFunc:iie,kernelFunc:oie},uie=dn(Ei),pie=dn(Fi),Sx;(function(e){e[e.bilinear=0]="bilinear",e[e.nearest=1]="nearest"})(Sx||(Sx={}));var bE;function cie(e){bE=e.wasm.cwrap(Ql,null,["number","number","number","number","array","number","number","number","number","number"])}function die(e){let{backend:t,inputs:n,attrs:a}=e,{method:r,extrapolationValue:s,cropSize:i}=a,{image:o,boxes:l,boxInd:u}=n,p=l.shape[0],[d,c]=i,h=[p,d,c,o.shape[3]],m=t.dataIdMap.get(o.dataId),f;o.dtype!=="float32"&&(f=fd({backend:t,inputs:{x:o},attrs:{dtype:"float32"}}),m=t.dataIdMap.get(f.dataId));let g=m.id,y=t.dataIdMap.get(l.dataId).id,b=t.dataIdMap.get(u.dataId).id,x=t.makeOutput(h,"float32"),v=t.dataIdMap.get(x.dataId).id,k=new Uint8Array(new Int32Array(o.shape).buffer);return bE(g,y,b,p,k,d,c,Sx[r],s,v),f!=null&&t.disposeData(f.dataId),x}var hie={kernelName:Ql,backendName:"wasm",setupFunc:cie,kernelFunc:die},xE;function mie(e){xE=e.wasm.cwrap(Yl,null,["number","number","number","number","number","number"])}function fie(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{axis:s,exclusive:i,reverse:o}=a,l=r.shape.length;w.assert(r.dtype==="float32"||r.dtype==="int32",()=>`cumprod does not support ${r.dtype} tensors in the WASM backend`);let u=C.getAxesPermutation([s],l),p=r;u!==null&&(p=gs({inputs:{x:r},attrs:{perm:u},backend:n}));let d=C.getInnerMostAxes(1,l)[0];C.assertAxesAreInnerMostDims("cumprod",[d],l);let c=n.makeOutput(p.shape,p.dtype),h=p.shape[d],m=n.dataIdMap.get(p.dataId).id,f=n.dataIdMap.get(c.dataId).id;xE(m,i?1:0,o?1:0,h,f,$t[r.dtype]);let g=c;if(u!==null){let y=C.getUndoAxesPermutation(u);g=gs({inputs:{x:c},attrs:{perm:y},backend:n}),n.disposeData(p.dataId),n.disposeData(c.dataId)}return g}var gie={kernelName:Yl,backendName:"wasm",setupFunc:mie,kernelFunc:fie},vE;function yie(e){vE=e.wasm.cwrap(Ai,null,["number","number","number","number","number","number"])}function bie(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{axis:s,exclusive:i,reverse:o}=a,l=r.shape.length;w.assert(r.dtype==="float32"||r.dtype==="int32",()=>`cumsum does not support ${r.dtype} tensors in the WASM backend`);let u=C.getAxesPermutation([s],l),p=r;u!==null&&(p=gs({inputs:{x:r},attrs:{perm:u},backend:n}));let d=C.getInnerMostAxes(1,l)[0];C.assertAxesAreInnerMostDims("cumsum",[d],l);let c=n.makeOutput(p.shape,p.dtype),h=p.shape[d],m=n.dataIdMap.get(p.dataId).id,f=n.dataIdMap.get(c.dataId).id;vE(m,i?1:0,o?1:0,h,f,$t[r.dtype]);let g=c;if(u!==null){let y=C.getUndoAxesPermutation(u);g=gs({inputs:{x:c},attrs:{perm:y},backend:n}),n.disposeData(p.dataId),n.disposeData(c.dataId)}return g}var xie={kernelName:Ai,backendName:"wasm",setupFunc:yie,kernelFunc:bie},wE;function vie(e){wE=e.wasm.cwrap(Jl,null,["number","number","number","array","number","array","array","number","number"])}function wie(e){let{backend:t,inputs:n,attrs:a}=e,{x:r}=n,{blockSize:s,dataFormat:i}=a,o=r.shape[0],l=i==="NHWC"?r.shape[1]:r.shape[2],u=i==="NHWC"?r.shape[2]:r.shape[3],p=i==="NHWC"?r.shape[3]:r.shape[1],d=l*s,c=u*s,h=p/(s*s),m=i==="NHWC"?[o,d,c,h]:[o,h,d,c],f=t.makeOutput(m,"float32"),g=t.dataIdMap.get(r.dataId).id,y=new Uint8Array(new Int32Array(w.computeStrides(r.shape)).buffer),b=new Uint8Array(new Int32Array(m).buffer),x=new Uint8Array(new Int32Array(w.computeStrides(m)).buffer),v=t.dataIdMap.get(f.dataId).id;return wE(g,s,i==="NHWC"?1:0,y,r.shape.length-1,b,x,m.length,v),f}var kie={kernelName:Jl,backendName:"wasm",setupFunc:vie,kernelFunc:wie},kE;function Iie(e){kE=e.wasm.cwrap($i,null,["number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number"])}function Sie(e){let{inputs:t,attrs:n,backend:a}=e,{x:r,filter:s}=t,i=a.dataIdMap.get(r.dataId).id,o=a.dataIdMap.get(s.dataId).id,{strides:l,dilations:u,pad:p,dimRoundingMode:d}=n,c=u==null?[1,1]:u,h=C.computeConv2DInfo(r.shape,s.shape,l,c,p,d,!0),m=h.filterHeight,f=h.filterWidth,g=h.padInfo.top,y=h.padInfo.right,b=h.padInfo.bottom,x=h.padInfo.left,v=h.dilationHeight,k=h.dilationWidth,T=h.strideHeight,_=h.strideWidth,E=h.inChannels,A=h.outChannels,M=h.padInfo.type==="SAME"?1:0;if(h.dataFormat!=="channelsLast")throw new Error(`wasm backend DepthwiseConv2dNative does not support dataFormat:'${h.dataFormat}'. Please use 'channelsLast'.`);let $=a.makeOutput(h.outShape,"float32"),S=a.dataIdMap.get($.dataId).id;return kE(i,r.shape[0],r.shape[1],r.shape[2],o,m,f,g,y,b,x,M,v,k,T,_,E,A,S),$}var Nie={kernelName:$i,backendName:"wasm",setupFunc:Iie,kernelFunc:Sie},Tie=dn(Ri),_ie=!1,Cie=Fn(eu,_ie,"bool"),Eie=dn(Mi,"float32");function Nx(e){let{inputs:t,attrs:n,backend:a}=e,{input:r}=t,{dim:s}=n,i=r.shape.length,o=r.shape.slice(),l=s;return s<0&&(w.assert(-(i+1)<=s,()=>`Axis must be in the interval [${-(i+1)}, ${i}]`),l=i+s+1),o.splice(l,0,1),Bn({inputs:{x:r},backend:a,attrs:{shape:o}})}var Fie={kernelName:tu,backendName:"wasm",kernelFunc:Nx};function IE(e){let{attrs:{shape:t,value:n,dtype:a},backend:r}=e,s=r.makeOutput(t,a);return r.typedArrayFromHeap(s).fill(n),s}var Aie={kernelName:kc,backendName:"wasm",kernelFunc:IE},SE;function $ie(e){SE=e.wasm.cwrap(au,null,["number","number","number","number","number","number"])}function Die(e){let{inputs:t,backend:n}=e,{image:a}=t,r=n.makeOutput(a.shape,a.dtype),s=n.dataIdMap.get(a.dataId).id,i=n.dataIdMap.get(r.dataId).id,[o,l,u,p]=a.shape;return SE(s,o,l,u,p,i),r}var Rie={kernelName:au,backendName:"wasm",kernelFunc:Die,setupFunc:$ie},Mie=dn(Pi),Pie=!1,Oie=Fn(Oi,Pie),NE;function Lie(e){NE=e.wasm.cwrap(Li,null,["number","number","number","number","number","number","number"])}function zie(e){let{backend:t,inputs:n,attrs:a}=e,{varianceEpsilon:r}=a,{x:s,mean:i,variance:o,offset:l,scale:u}=n,p=t.dataIdMap.get(s.dataId).id,d=t.dataIdMap.get(i.dataId).id,c=t.dataIdMap.get(o.dataId).id,h=l!=null?t.dataIdMap.get(l.dataId).id:0,m=u!=null?t.dataIdMap.get(u.dataId).id:0,f=t.makeOutput(s.shape,s.dtype);if(w.sizeFromShape(s.shape)===0)return f;let g=t.dataIdMap.get(f.dataId).id;return NE(p,d,c,h,m,r,g),f}var Bie={kernelName:Li,backendName:"wasm",setupFunc:Lie,kernelFunc:zie},TE;function Wie(e){TE=e.wasm.cwrap(ri,null,["number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number"])}function Vie(e){let{inputs:t,attrs:n,backend:a}=e,{x:r,filter:s,bias:i,preluActivationWeights:o}=t,{strides:l,pad:u,dilations:p,dataFormat:d,dimRoundingMode:c,activation:h,leakyreluAlpha:m}=n,f=C.computeConv2DInfo(r.shape,s.shape,l,p,u,c),g=fc[h];if(g==null)throw new Error(`${h} activation not yet supported for FusedConv2D in the wasm backend.`);let y=a.dataIdMap.get(r.dataId).id,b=a.dataIdMap.get(s.dataId).id,x=f.outChannels,v=0;if(i!=null){let ae=a.dataIdMap.get(i.dataId);if(ae.shape.length!==1)throw new Error(`FusedConv2D only supports rank-1 bias but got rank ${ae.shape.length}.`);if(ae.shape[0]!==x)throw new Error(`FusedConv2D bias shape (${ae.shape}) does not match the number of output channels (${x})`);v=ae.id}let k=f.filterHeight,T=f.filterWidth,_=f.padInfo.top,E=f.padInfo.right,A=f.padInfo.bottom,M=f.padInfo.left,$=f.dilationHeight,S=f.dilationWidth,P=f.strideHeight,V=f.strideWidth,j=f.inChannels,q=f.padInfo.type==="SAME"?1:0,K=f.batchSize,Z=f.inHeight,ee=f.inWidth;if(d!=="NHWC")throw new Error(`wasm backend FusedConv2D does not support dataFormat:'${d}'. Please use 'NHWC'.`);let re=a.makeOutput(f.outShape,"float32"),Y=a.dataIdMap.get(re.dataId).id,ie=o==null?0:a.dataIdMap.get(o.dataId).id;return TE(y,K,Z,ee,b,k,T,v,_,E,A,M,q,$,S,P,V,j,x,g,ie,m||0,Y),re}var Uie={kernelName:ri,backendName:"wasm",setupFunc:Wie,kernelFunc:Vie},_E;function Gie(e){_E=e.wasm.cwrap(si,null,["number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number"])}function Hie(e){let{inputs:t,attrs:n,backend:a}=e,{x:r,filter:s,bias:i,preluActivationWeights:o}=t,{strides:l,pad:u,dilations:p,dataFormat:d,dimRoundingMode:c,activation:h,leakyreluAlpha:m}=n,f=C.computeConv2DInfo(r.shape,s.shape,l,p,u,c,!0),g=fc[h];if(g==null)throw new Error(`${h} activation not yet supported for FusedDepthwiseConv2D in the wasm backend.`);let y=a.dataIdMap.get(r.dataId).id,b=a.dataIdMap.get(s.dataId).id,x=f.outChannels,v=0;if(i!=null){let ae=a.dataIdMap.get(i.dataId);if(ae.shape.length!==1)throw new Error(`FusedDepthwiseConv2D only supports rank-1 bias but got rank ${ae.shape.length}.`);if(ae.shape[0]!==x)throw new Error(`FusedDepthwiseConv2D bias shape (${ae.shape}) does not match the number of output channels (${x})`);v=ae.id}let k=f.filterHeight,T=f.filterWidth,_=f.padInfo.top,E=f.padInfo.right,A=f.padInfo.bottom,M=f.padInfo.left,$=f.dilationHeight,S=f.dilationWidth,P=f.strideHeight,V=f.strideWidth,j=f.inChannels,q=f.padInfo.type==="SAME"?1:0,K=f.batchSize,Z=f.inHeight,ee=f.inWidth;if(d!=="NHWC")throw new Error(`wasm backend FusedDepthwiseConv2D does not support dataFormat:'${d}'. Please use 'NHWC'.`);let re=a.makeOutput(f.outShape,"float32"),Y=a.dataIdMap.get(re.dataId).id,ie=o==null?0:a.dataIdMap.get(o.dataId).id;return _E(y,K,Z,ee,b,k,T,v,_,E,A,M,q,$,S,P,V,j,x,g,ie,m||0,Y),re}var jie={kernelName:si,backendName:"wasm",setupFunc:Gie,kernelFunc:Hie},CE;function qie(e){CE=e.wasm.cwrap(su,null,["number","number","number","number","number","number","array","number"])}function Kie(e){let{backend:t,inputs:n}=e,{params:a,indices:r}=n,[s,i,o,l]=Gx.prepareAndValidate(a,r),u=t.makeOutput(s,a.dtype);if(i===0)return u;let p=r.shape,d=p[p.length-1],c=t.dataIdMap.get(a.dataId).id,h=t.dataIdMap.get(r.dataId).id,m=new Uint8Array(new Int32Array(l).buffer),f=t.dataIdMap.get(u.dataId).id;return CE(c,$t[a.dtype],h,i,d,o,m,f),u}var Xie={kernelName:su,backendName:"wasm",setupFunc:qie,kernelFunc:Kie},EE;function Yie(e){EE=e.wasm.cwrap("Gather",null,["number","number","array","number","number","number","array","number"])}function Qie(e){let{backend:t,inputs:n,attrs:a}=e,{x:r,indices:s}=n,{axis:i,batchDims:o}=a,l=w.parseAxisParam(i,r.shape)[0],u=t.readSync(s.dataId),p=r.shape[l];for(let _=0;_<u.length;++_){let E=u[_];w.assert(E<=p-1&&E>=0,()=>`GatherV2: the index value ${E} is not in [0, ${p-1}]`)}let d=C.segment_util.collectGatherOpShapeInfo(r,s,l,o),c=Bn({inputs:{x:r},attrs:{shape:[d.batchSize,d.outerSize,d.dimSize,d.sliceSize]},backend:t}),h=w.sizeFromShape(s.shape),m=Bn({inputs:{x:s},attrs:{shape:[d.batchSize,h/d.batchSize]},backend:t}),f=[d.batchSize,d.outerSize,h/d.batchSize,d.sliceSize],g=t.makeOutput(f,r.dtype);if(w.sizeFromShape(r.shape)===0)return g;let y=c.shape.length-1,b=t.dataIdMap.get(c.dataId).id,x=t.dataIdMap.get(m.dataId).id,v=t.dataIdMap.get(g.dataId).id,k=new Uint8Array(new Int32Array(w.computeStrides(c.shape)).buffer),T=new Uint8Array(new Int32Array(w.computeStrides(f)).buffer);return EE(b,$t[r.dtype],k,y,x,d.batchSize,T,v),t.disposeData(c.dataId),t.disposeData(m.dataId),g.shape=d.outputShape,g}var Jie={kernelName:ru,backendName:"wasm",setupFunc:Yie,kernelFunc:Qie},Zie=!1,eoe=Fn(iu,Zie,"bool"),toe=!1,noe=Fn(zi,toe,"bool"),FE;function aoe(e){FE=e.wasm.cwrap(Wi,null,["number","number","number","number"])}function roe(e){let{inputs:{x:t},attrs:{alpha:n},backend:a}=e,r=a.dataIdMap.get(t.dataId).id,s=a.makeOutput(t.shape,"float32");if(w.sizeFromShape(t.shape)!==0){let i=a.dataIdMap.get(s.dataId).id;FE(r,$t[t.dtype],n,i)}return s}var soe={kernelName:Wi,backendName:"wasm",setupFunc:aoe,kernelFunc:roe},ioe=!1,ooe=Fn(pu,ioe,"bool"),loe=!1,uoe=Fn(cu,loe,"bool"),poe=dn(Vi),coe=!1,doe=Fn(hu,coe,"bool"),AE;function hoe(e){AE=e.wasm.cwrap(Ui,null,["number","number","number","number"])}function moe(e){let{backend:t,inputs:n,attrs:a}=e,{reductionIndices:r,keepDims:s}=a,{x:i}=n,o=t.dataIdMap.get(i.dataId).id,l=i,{transposed:u,axes:p,originalAxes:d,inputWasTransposed:c}=Es(i,r,t);if(c){let b=t.dataIdMap.get(u.dataId).id;l=u,o=b}let h=l.shape.length;C.assertAxesAreInnerMostDims("max",p,h);let[m,f]=C.computeOutAndReduceShapes(l.shape,p),g=w.sizeFromShape(f),y=t.makeOutput(m,i.dtype);if(w.sizeFromShape(l.shape)!==0){let b=t.dataIdMap.get(y.dataId).id;AE(o,$t[i.dtype],g,b)}if(c&&t.disposeData(u.dataId),s){let b=C.expandShapeToKeepDim(y.shape,d);y.shape=b}return y}var foe={kernelName:Ui,backendName:"wasm",setupFunc:hoe,kernelFunc:moe},goe=!1,yoe=Fn(Gi,goe),$E;function boe(e){$E=e.wasm.cwrap(Hi,null,["number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number","number"])}function xoe(e){let{inputs:t,attrs:n,backend:a}=e,r=t.x,s=a.dataIdMap.get(r.dataId).id;w.assert(r.dtype==="float32",()=>`Error in MaxPool: only float32 input is supported. Got ${r.dtype}.`);let{filterSize:i,strides:o,pad:l,dimRoundingMode:u}=n,p=C.computePool2DInfo(r.shape,i,o,1,l,u),d=p.filterHeight,c=p.filterWidth,h=p.padInfo.top,m=p.padInfo.right,f=p.padInfo.bottom,g=p.padInfo.left,y=p.dilationHeight,b=p.dilationWidth,x=p.strideHeight,v=p.strideWidth,k=p.inChannels,T=p.outChannels;if(p.dataFormat!=="channelsLast")throw new Error(`wasm backend does not support dataFormat:'${p.dataFormat}'. Please use 'channelsLast'.`);let _=a.makeOutput(p.outShape,"float32"),E=a.dataIdMap.get(_.dataId).id;return $E(s,r.shape[0],r.shape[1],r.shape[2],d,c,h,m,f,g,y,b,x,v,k,T,E),_}var voe={kernelName:Hi,backendName:"wasm",setupFunc:boe,kernelFunc:xoe},DE;function woe(e){DE=e.wasm.cwrap(ji,null,["number, number, number"])}function koe(e){let{backend:t,inputs:n,attrs:a}=e,{axis:r,keepDims:s}=a,{x:i}=n,o=t.dataIdMap.get(i.dataId).id,l=o,u=i,{transposed:p,axes:d,originalAxes:c,inputWasTransposed:h}=Es(i,r,t),m=d;if(h){let v=t.dataIdMap.get(p.dataId).id;v!==o&&(u=p,l=v,m=C.getInnerMostAxes(m.length,u.shape.length))}C.assertAxesAreInnerMostDims("mean",m,u.shape.length);let[f,g]=C.computeOutAndReduceShapes(u.shape,m),y=w.sizeFromShape(g),b=u;u.dtype!=="float32"&&(b=fd({backend:t,inputs:{x:u},attrs:{dtype:"float32"}}),l=t.dataIdMap.get(b.dataId).id);let x=t.makeOutput(f,"float32");if(w.sizeFromShape(u.shape)!==0){let v=t.dataIdMap.get(x.dataId).id;DE(l,y,v)}if(h&&t.disposeData(p.dataId),s){let v=C.expandShapeToKeepDim(x.shape,c);x.shape=v}return u.dtype!=="float32"&&t.disposeData(b.dataId),x}var Ioe={kernelName:ji,backendName:"wasm",setupFunc:woe,kernelFunc:koe},RE;function Soe(e){RE=e.wasm.cwrap(qi,null,["number","number","number","number"])}function Noe(e){let{backend:t,inputs:n,attrs:a}=e,{axis:r,keepDims:s}=a,{x:i}=n,o=t.dataIdMap.get(i.dataId).id,l=o,u=i,{transposed:p,axes:d,originalAxes:c,inputWasTransposed:h}=Es(i,r,t);if(h){let x=t.dataIdMap.get(p.dataId).id;x!==o&&(u=p,l=x)}let m=u.shape.length;C.assertAxesAreInnerMostDims("min",d,m);let[f,g]=C.computeOutAndReduceShapes(u.shape,d),y=w.sizeFromShape(g),b=t.makeOutput(f,u.dtype);if(w.sizeFromShape(u.shape)!==0){let x=t.dataIdMap.get(b.dataId).id;RE(l,$t[i.dtype],y,x)}if(h&&t.disposeData(p.dataId),s){let x=C.expandShapeToKeepDim(b.shape,c);b.shape=x}return b}var Toe={kernelName:qi,backendName:"wasm",setupFunc:Soe,kernelFunc:Noe},_oe=!1,Coe=Fn(Ki,_oe),Tx;(function(e){e[e.reflect=0]="reflect",e[e.symmetric=1]="symmetric"})(Tx||(Tx={}));var ME;function Eoe(e){ME=e.wasm.cwrap(Xi,null,["number","array","number","number","array","array","number","number"])}function Foe(e){let{inputs:{x:t},backend:n,attrs:{paddings:a,mode:r}}=e,s=a.map((m,f)=>m[0]+t.shape[f]+m[1]),i=n.dataIdMap.get(t.dataId).id,o=n.makeOutput(s,t.dtype),l=n.dataIdMap.get(o.dataId).id,u=new Uint8Array(new Int32Array(t.shape).buffer),p=a.map(m=>m[0]),d=a.map(m=>m[1]),c=new Uint8Array(new Int32Array(p).buffer),h=new Uint8Array(new Int32Array(d).buffer);return ME(i,u,t.shape.length,$t[t.dtype],c,h,Tx[r],l),o}var Aoe={kernelName:Xi,backendName:"wasm",kernelFunc:Foe,setupFunc:Eoe},$oe=!0,Doe=Fn(Yi,$oe),Roe=dn(fu);function G0(e,t){let n=new Int32Array(e.wasm.HEAPU8.buffer,t,4),a=n[0],r=n[1],s=n[2],i=n[3];return e.wasm._free(t),{pSelectedIndices:a,selectedSize:r,pSelectedScores:s,pValidOutputs:i}}var PE;function Moe(e){PE=e.wasm.cwrap(yu,"number",["number","number","number","number","number"])}function Poe(e){let{backend:t,inputs:n,attrs:a}=e,{iouThreshold:r,maxOutputSize:s,scoreThreshold:i}=a,{boxes:o,scores:l}=n,u=t.dataIdMap.get(o.dataId).id,p=t.dataIdMap.get(l.dataId).id,d=PE(u,p,s,r,i),{pSelectedIndices:c,selectedSize:h,pSelectedScores:m,pValidOutputs:f}=G0(t,d);return t.wasm._free(m),t.wasm._free(f),t.makeOutput([h],"int32",c)}var Ooe={kernelName:yu,backendName:"wasm",setupFunc:Moe,kernelFunc:Poe},OE;function Loe(e){OE=e.wasm.cwrap(bu,"number",["number","number","number","number","number","bool"])}function zoe(e){let{backend:t,inputs:n,attrs:a}=e,{iouThreshold:r,maxOutputSize:s,scoreThreshold:i,padToMaxOutputSize:o}=a,{boxes:l,scores:u}=n,p=t.dataIdMap.get(l.dataId).id,d=t.dataIdMap.get(u.dataId).id,c=OE(p,d,s,r,i,o),{pSelectedIndices:h,selectedSize:m,pSelectedScores:f,pValidOutputs:g}=G0(t,c);t.wasm._free(f);let y=t.makeOutput([m],"int32",h),b=t.makeOutput([],"int32",g);return[y,b]}var Boe={kernelName:bu,backendName:"wasm",setupFunc:Loe,kernelFunc:zoe},LE;function Woe(e){LE=e.wasm.cwrap(xu,"number",["number","number","number","number","number","number"])}function Voe(e){let{backend:t,inputs:n,attrs:a}=e,{iouThreshold:r,maxOutputSize:s,scoreThreshold:i,softNmsSigma:o}=a,{boxes:l,scores:u}=n,p=t.dataIdMap.get(l.dataId).id,d=t.dataIdMap.get(u.dataId).id,c=LE(p,d,s,r,i,o),{pSelectedIndices:h,selectedSize:m,pSelectedScores:f,pValidOutputs:g}=G0(t,c);t.wasm._free(g);let y=t.makeOutput([m],"int32",h),b=t.makeOutput([m],"float32",f);return[y,b]}var Uoe={kernelName:xu,backendName:"wasm",setupFunc:Woe,kernelFunc:Voe},Goe=!1,Hoe=Fn(gu,Goe,"bool"),zE;function joe(e){zE=e.wasm.cwrap(Qi,null,["number","number","number","number","number"])}function qoe(e){let{inputs:t,backend:n,attrs:a}=e,{indices:r}=t,{depth:s,onValue:i,offValue:o}=a,l=n.makeOutput([...r.shape,s],"int32"),u=n.dataIdMap.get(l.dataId).id,p=n.dataIdMap.get(r.dataId).id;return zE(p,s,i,o,u),l}var Koe={kernelName:Qi,backendName:"wasm",setupFunc:joe,kernelFunc:qoe};function Xoe(e){let{inputs:{x:t},backend:n}=e,a=n.makeOutput(t.shape,t.dtype);return n.typedArrayFromHeap(a).fill(1),a}var Yoe={kernelName:vu,backendName:"wasm",kernelFunc:Xoe};function Qoe(e){let{inputs:t,backend:n,attrs:a}=e,{axis:r}=a;if(t.length===1)return Nx({inputs:{input:t[0]},backend:n,attrs:{dim:r}});let s=t[0].shape,i=t[0].dtype;t.forEach(p=>{w.assertShapesMatch(s,p.shape,"All tensors passed to stack must have matching shapes"),w.assert(i===p.dtype,()=>"All tensors passed to stack must have matching dtypes")});let o=[],l=t.map(p=>{let d=Nx({inputs:{input:p},backend:n,attrs:{dim:r}});return o.push(d),d}),u=fE({inputs:l,backend:n,attrs:{axis:r}});return o.forEach(p=>n.disposeData(p.dataId)),u}var Joe={kernelName:wu,backendName:"wasm",kernelFunc:Qoe},BE;function Zoe(e){BE=e.wasm.cwrap(Ji,null,["number","array","number","number","array","array","number","number"])}function ele(e){let{inputs:{x:t},backend:n,attrs:{paddings:a,constantValue:r}}=e,s=a.map((m,f)=>m[0]+t.shape[f]+m[1]);if(w.sizeFromShape(t.shape)===0)return IE({backend:n,attrs:{shape:s,value:r,dtype:t.dtype}});let i=n.dataIdMap.get(t.dataId).id,o=n.makeOutput(s,t.dtype),l=n.dataIdMap.get(o.dataId).id,u=new Uint8Array(new Int32Array(t.shape).buffer),p=a.map(m=>m[0]),d=a.map(m=>m[1]),c=new Uint8Array(new Int32Array(p).buffer),h=new Uint8Array(new Int32Array(d).buffer);return BE(i,u,t.shape.length,$t[t.dtype],c,h,r,l),o}var WE={kernelName:Ji,backendName:"wasm",kernelFunc:ele,setupFunc:Zoe},tle=!1,nle=Fn(Zi,tle),VE;function ale(e){VE=e.wasm.cwrap(eo,null,["number","number","number"])}function rle(e){let{inputs:t,backend:n}=e,{x:a,alpha:r}=t,s=n.dataIdMap.get(a.dataId).id,i=n.dataIdMap.get(r.dataId).id,o=s,l=a,u=l;l.dtype!=="float32"&&(u=fd({backend:n,inputs:{x:a},attrs:{dtype:"float32"}}),o=n.dataIdMap.get(u.dataId).id);let p=n.makeOutput(a.shape,"float32"),d=n.dataIdMap.get(p.dataId).id;return VE(o,i,d),l.dtype!=="float32"&&n.disposeData(u.dataId),p}var sle={kernelName:eo,backendName:"wasm",setupFunc:ale,kernelFunc:rle},UE;function ile(e){UE=e.wasm.cwrap(to,null,["number","number","number","number"])}function ole(e){let{backend:t,inputs:n,attrs:a}=e,{axis:r,keepDims:s}=a,{x:i}=n,o=t.dataIdMap.get(i.dataId).id,l=o,u=i,{transposed:p,axes:d,originalAxes:c,inputWasTransposed:h}=Es(i,r,t),m=d;if(h){let x=t.dataIdMap.get(p.dataId).id;x!==o&&(u=p,l=x,m=C.getInnerMostAxes(m.length,u.shape.length))}C.assertAxesAreInnerMostDims("prod",m,u.shape.length);let[f,g]=C.computeOutAndReduceShapes(u.shape,m),y=w.sizeFromShape(g),b=t.makeOutput(f,u.dtype);if(w.sizeFromShape(u.shape)!==0){let x=t.dataIdMap.get(b.dataId).id;UE(l,y,$t[b.dtype],x)}if(h&&t.disposeData(p.dataId),s){let x=C.expandShapeToKeepDim(b.shape,c);b.shape=x}return b}var lle={kernelName:to,backendName:"wasm",setupFunc:ile,kernelFunc:ole},ule=e=>{let{backend:t,attrs:n}=e,{start:a,stop:r,step:s,dtype:i}=n,o=N0(a,r,s,i),l=t.makeOutput([o.length],i);return t.typedArrayFromHeap(l).set(o),l},ple={kernelName:_c,backendName:"wasm",kernelFunc:ule},cle=!0,dle=Fn(Di,cle),hle=dn(no),mle=dn(ro),GE;function fle(e){GE=e.wasm.cwrap(ao,null,["number","number","number","number","number","number","number","number","number","number"])}function gle(e){let{backend:t,inputs:n,attrs:a}=e,{images:r}=n,{alignCorners:s,halfPixelCenters:i,size:o}=a,[l,u]=o,[p,d,c,h]=r.shape,m=[p,l,u,h],f=t.dataIdMap.get(r.dataId),g;f.dtype!=="float32"&&(g=fd({backend:t,inputs:{x:r},attrs:{dtype:"float32"}}),f=t.dataIdMap.get(g.dataId));let y=f.id,b=t.makeOutput(m,"float32");if(w.sizeFromShape(r.shape)===0)return b;let x=t.dataIdMap.get(b.dataId).id;return GE(y,p,d,c,h,l,u,s?1:0,i?1:0,x),g!=null&&t.disposeData(g.dataId),b}var yle={kernelName:ao,backendName:"wasm",setupFunc:fle,kernelFunc:gle},HE;function ble(e){HE=e.wasm.cwrap(so,null,["number","array","number","array","number","number"])}function xle(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{dims:s}=a,i=w.parseAxisParam(s,r.shape);if(r.shape.length===0)return cg({inputs:{x:r},backend:n});let o=n.makeOutput(r.shape,r.dtype),l=n.dataIdMap.get(r.dataId).id,u=n.dataIdMap.get(o.dataId).id,p=new Uint8Array(new Int32Array(i).buffer),d=new Uint8Array(new Int32Array(r.shape).buffer);HE(l,p,i.length,d,r.shape.length,u);let c=Bn({inputs:{x:o},attrs:{shape:r.shape},backend:n});return n.disposeData(o.dataId),c}var vle={kernelName:so,backendName:"wasm",kernelFunc:xle,setupFunc:ble},jE;function wle(e){jE=e.wasm.cwrap(zu,null,["number","number","number","number","number","number","number","number","array","number","number"])}function kle(e){let{inputs:t,backend:n,attrs:a}=e,{image:r}=t,{radians:s,fillValue:i,center:o}=a,l=n.makeOutput(r.shape,r.dtype),u=n.dataIdMap.get(r.dataId).id,p=n.dataIdMap.get(l.dataId).id,[d,c,h,m]=r.shape,[f,g]=C.getImageCenter(o,c,h),y=i===0,b=255,x=typeof i=="number"?[i,i,i,y?0:b]:[...i,b],v=new Uint8Array(new Int32Array(x).buffer);return jE(u,d,c,h,m,s,f,g,v,x.length,p),l}var Ile={kernelName:zu,backendName:"wasm",kernelFunc:kle,setupFunc:wle},Sle=dn(io),Nle=dn(oo),qE;function Tle(e){qE=e.wasm.cwrap(Su,null,["number","number","number","number","number","number","array","number","number"])}function _le(e){let{backend:t,inputs:n,attrs:a}=e,{indices:r,updates:s}=n,{shape:i}=a,o=t.makeOutput(i,s.dtype);if(w.sizeFromShape(i)===0)return o;let{sliceRank:l,numUpdates:u,sliceSize:p,strides:d,outputSize:c}=Hx.calculateShapes(s,r,i),h=t.dataIdMap.get(r.dataId).id,m=t.dataIdMap.get(s.dataId).id,f=new Uint8Array(new Int32Array(d).buffer),g=t.dataIdMap.get(o.dataId).id;return qE(h,m,$t[s.dtype],l,u,p,f,c,g),o}var Cle={kernelName:Su,backendName:"wasm",setupFunc:Tle,kernelFunc:_le},KE;function Ele(e){KE=e.wasm.cwrap("SelectV2",null,["number","number","number","number","number"])}function Fle(e){let{inputs:t,backend:n}=e,{condition:a,t:r,e:s}=t,i=n.dataIdMap.get(a.dataId).id,o=n.dataIdMap.get(r.dataId).id,l=n.dataIdMap.get(s.dataId).id,u=n.makeOutput(r.shape,r.dtype),p=n.dataIdMap.get(u.dataId).id,d=a.shape.length,c=r.shape.length,h=d===0||d>1||c===1?1:w.sizeFromShape(r.shape.slice(1));return KE(i,o,l,h,p),u}var Ale={kernelName:Nu,backendName:"wasm",kernelFunc:Fle,setupFunc:Ele},XE;function $le(e){XE=e.wasm.cwrap(uo,null,["number","number"])}function Dle(e){let{backend:t,inputs:{x:n}}=e,a=t.dataIdMap.get(n.dataId).id,r=t.makeOutput(n.shape,n.dtype),s=t.dataIdMap.get(r.dataId).id;return w.sizeFromShape(r.shape)===0||XE(a,s),r}var Rle={kernelName:"Sigmoid",backendName:"wasm",setupFunc:$le,kernelFunc:Dle},Mle=dn(lo),YE;function Ple(e){YE=e.wasm.cwrap(ho,null,["number","number","number","number"])}function Ole(e){let{backend:t,inputs:{logits:n},attrs:{dim:a}}=e,r=t.dataIdMap.get(n.dataId).id,s=t.makeOutput(n.shape,n.dtype),i=t.dataIdMap.get(s.dataId).id,o=n.shape[a],l=w.sizeFromShape(n.shape)/o;return w.sizeFromShape(s.shape)===0||YE(r,i,o,l),s}var Lle={kernelName:ho,backendName:"wasm",setupFunc:Ple,kernelFunc:Ole};function zle(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,{blockShape:s,paddings:i}=a,o=w.sizeFromShape(s),l=[[0,0]];l.push(...i);for(let g=1+s.length;g<r.shape.length;++g)l.push([0,0]);let u=WE.kernelFunc({inputs:{x:r},backend:n,attrs:{paddings:l,constantValue:0}}),p=C.getReshaped(u.shape,s,o,!1),d=C.getPermuted(p.length,s.length,!1),c=C.getReshapedPermuted(u.shape,s,o,!1),h=Bn({inputs:{x:u},backend:n,attrs:{shape:p}}),m=gs({inputs:{x:h},backend:n,attrs:{perm:d}}),f=Bn({inputs:{x:m},backend:n,attrs:{shape:c}});return n.disposeData(u.dataId),n.disposeData(h.dataId),n.disposeData(m.dataId),f}var Ble={kernelName:Au,backendName:"wasm",kernelFunc:zle},QE;function Wle(e){QE=e.wasm.cwrap("SparseFillEmptyRows","number",["number","number","number","number","number","number","number","number","number","number","number","number"])}function Vle(e){let{backend:t,inputs:n}=e,{indices:a,values:r,denseShape:s,defaultValue:i}=n,o=a.shape[0],l=a.shape[1],u=t.readSync(s.dataId)[0],p=[o+u,l],d=t.dataIdMap.get(a.dataId).id,c=t.dataIdMap.get(r.dataId).id,h=t.dataIdMap.get(i.dataId).id,m=t.makeOutput(p,a.dtype),f=t.dataIdMap.get(m.dataId).id,g=t.makeOutput(p.slice(0,1),r.dtype),y=t.dataIdMap.get(g.dataId).id,b=t.makeOutput([u],"bool"),x=t.dataIdMap.get(b.dataId).id,v=t.makeOutput([o],a.dtype),k=t.dataIdMap.get(v.dataId).id,T=t.makeOutput([4],"int32"),_=t.dataIdMap.get(T.dataId).id,E=QE(d,c,$t[r.dtype],o,u,l,h,f,y,x,k,_),A=t.readSync(T.dataId),M;switch(A[0]){case 1:{M=C.getSparseFillEmptyRowsIndicesDenseShapeMismatch(A[1]);break}case 2:{M=C.getSparseFillEmptyRowsNegativeIndexErrorMessage(A[1],A[2]);break}case 3:M=C.getSparseFillEmptyRowsOutOfRangeIndexErrorMessage(A[1],A[2],A[3]);break;default:M=""}if(t.disposeData(T.dataId),M)throw t.disposeData(m.dataId),t.disposeData(g.dataId),t.disposeData(b.dataId),t.disposeData(v.dataId),new Error(M);let $=m,S=g;return E!==p[0]&&($=bi({inputs:{x:m},attrs:{begin:0,size:[E,l]},backend:t}),S=bi({inputs:{x:g},attrs:{begin:0,size:E},backend:t}),t.disposeData(m.dataId),t.disposeData(g.dataId)),[$,S,b,v]}var Ule={kernelName:Ec,backendName:"wasm",setupFunc:Wle,kernelFunc:Vle},JE;function Gle(e){JE=e.wasm.cwrap(Du,null,["number","number","number","number","number","number","number"])}function Hle(e){let{backend:t,inputs:n}=e,{inputIndices:a,inputShape:r,newShape:s}=n;if(a.shape.length!==2)throw new Error(`Input indices should be a matrix but received shape
${a.shape}`);if(r.shape.length!==1)throw new Error(`Input shape should be a vector but received shape
${r.shape}`);if(s.shape.length!==1)throw new Error(`Target shape should be a vector but received shape ${s.shape}`);let i=t.dataIdMap.get(a.dataId).id,o=t.dataIdMap.get(r.dataId).id,l=t.dataIdMap.get(s.dataId).id,u=a.shape[0],p=w.sizeFromShape(s.shape),d=t.makeOutput([u,p],a.dtype),c=t.dataIdMap.get(d.dataId).id,h=t.makeOutput([p],s.dtype),m=t.dataIdMap.get(h.dataId).id,f=t.makeOutput([3],"int32"),g=t.dataIdMap.get(f.dataId).id;JE(i,o,l,u,c,m,g);let y=t.readSync(f.dataId),b;switch(y[0]){case 0:{b=C.getSparseReshapeMultipleNegativeOneOutputDimErrorMessage(y[1],y[2]);break}case 1:{b=C.getSparseReshapeNegativeOutputDimErrorMessage(y[1],y[2]);break}case 2:b=C.getSparseReshapeEmptyTensorZeroOutputDimErrorMessage();break;case 3:{let x=Array.from(t.readSync(r.dataId)),v=Array.from(t.readSync(h.dataId));b=C.getSparseReshapeInputOutputMultipleErrorMessage(x,v);break}case 4:{let x=Array.from(t.readSync(r.dataId)),v=Array.from(t.readSync(h.dataId));b=C.getSparseReshapeInputOutputMismatchErrorMessage(x,v);break}default:b=""}if(t.disposeData(f.dataId),b)throw t.disposeData(d.dataId),t.disposeData(h.dataId),new Error(b);return[d,h]}var jle={kernelName:Du,backendName:"wasm",setupFunc:Gle,kernelFunc:Hle},ZE;function eF(e){ZE=e.wasm.cwrap("SparseSegmentReduction",null,["number","number","number","number","number","number","number","number","number"])}function tF(e,t){let{backend:n,inputs:a}=e,{data:r,indices:s,segmentIds:i}=a,o=s.shape[0],l=n.readSync(i.dataId,o-1,o)[0],u=o>0?l+1:0;if(u<0)throw new Error(C.getSparseSegmentReductionNegativeSegmentIdsErrorMessage());let p=r.shape.slice();p[0]=u;let d=n.dataIdMap.get(r.dataId).id,c=n.dataIdMap.get(s.dataId).id,h=n.dataIdMap.get(i.dataId).id,m=n.makeOutput(p,r.dtype),f=n.dataIdMap.get(m.dataId).id,g=n.makeOutput([4],"int32"),y=n.dataIdMap.get(g.dataId).id;ZE(d,$t[r.dtype],r.shape[0],c,h,f,y,t,0);let b=n.readSync(g.dataId),x;switch(b[0]){case 0:{x=C.getSparseSegmentReductionNegativeSegmentIdsErrorMessage();break}case 1:{x=C.getSparseSegmentReductionNonIncreasingSegmentIdsErrorMessage();break}case 2:x=C.getSparseSegmentReductionSegmentIdOutOfRangeErrorMessage(b[1],b[2]);break;case 3:x=C.getSparseSegmentReductionIndicesOutOfRangeErrorMessage(b[1],b[2],b[3]);break;default:x=""}if(n.disposeData(g.dataId),x)throw n.disposeData(m.dataId),new Error(x);return m}function qle(e){return tF(e,!0)}var Kle={kernelName:Fc,backendName:"wasm",setupFunc:eF,kernelFunc:qle};function Xle(e){return tF(e,!1)}var Yle={kernelName:Ac,backendName:"wasm",setupFunc:eF,kernelFunc:Xle};function Qle(e){let{inputs:t,attrs:n,backend:a}=e,{x:r}=t,{numOrSizeSplits:s,axis:i}=n,o=w.parseAxisParam(i,r.shape)[0],l=C.prepareSplitSize(r,s,o),u=new Array(r.shape.length).fill(0),p=r.shape.slice();return l.map(d=>{let c=[...p];c[o]=d;let h=bi({inputs:{x:r},attrs:{begin:u,size:c},backend:a});return u[o]+=d,h})}var Jle={kernelName:$u,backendName:"wasm",kernelFunc:Qle},Zle=dn(po),eue=dn($c),tue=!0,nue=Fn(mo,tue),nF;function aue(e){nF=e.wasm.cwrap(ws,null,["number","number","number","number"])}function rue(e){let{backend:t,inputs:n,attrs:a}=e,{alpha:r}=a,{x:s}=n,i=t.dataIdMap.get(s.dataId).id,o=t.makeOutput(s.shape,s.dtype),l=t.dataIdMap.get(o.dataId).id;return nF(i,r,$t[s.dtype],l),o}var sue={kernelName:ws,backendName:"wasm",setupFunc:aue,kernelFunc:rue},aF;function iue(e){aF=e.wasm.cwrap(Ru,null,["number","array","number","array","array","array","array","array","number","number"])}function oue(e){let{backend:t,inputs:n,attrs:a}=e,{x:r}=n,{begin:s,end:i,strides:o,beginMask:l,endMask:u,ellipsisMask:p,newAxisMask:d,shrinkAxisMask:c}=a,{finalShapeSparse:h,finalShape:m,isIdentity:f,sliceDim0:g,isSimpleSlice:y,begin:b,end:x,strides:v}=Kt.sliceInfo(r.shape,s,i,o,l,u,p,d,c),k;if(f)k=Bn({inputs:{x:r},backend:t,attrs:{shape:m}});else if(g||y){w.assert(r.shape.length>=1,()=>`Input must have rank at least 1, got: ${r.shape.length}`);let T=Kt.computeOutShape(b,x,v),_=bi({inputs:{x:r},backend:t,attrs:{begin:b,size:T}});k=Bn({inputs:{x:_},backend:t,attrs:{shape:m}}),t.disposeData(_.dataId)}else{let T=t.makeOutput(h,"float32"),_=t.dataIdMap.get(r.dataId).id,E=new Uint8Array(new Int32Array(w.computeStrides(r.shape)).buffer),A=new Uint8Array(new Int32Array(b).buffer),M=new Uint8Array(new Int32Array(x).buffer),$=new Uint8Array(new Int32Array(v).buffer),S=new Uint8Array(new Int32Array(h).buffer),P=new Uint8Array(new Int32Array(w.computeStrides(h)).buffer),V=t.dataIdMap.get(T.dataId).id;aF(_,E,r.shape.length,A,M,$,S,P,h.length,V),k=Bn({inputs:{x:T},backend:t,attrs:{shape:m}}),t.disposeData(T.dataId)}return k}var lue={kernelName:Ru,backendName:"wasm",setupFunc:iue,kernelFunc:oue},uue=!0,pue=Fn(fo,uue),rF;function cue(e){rF=e.wasm.cwrap(co,null,["number","number","number","number"])}function due(e){let{backend:t,inputs:n,attrs:a}=e,{axis:r,keepDims:s}=a,{x:i}=n,o=t.dataIdMap.get(i.dataId).id,l=o,u=i,{transposed:p,axes:d,originalAxes:c,inputWasTransposed:h}=Es(i,r,t),m=d;if(h){let x=t.dataIdMap.get(p.dataId).id;x!==o&&(u=p,l=x,m=C.getInnerMostAxes(m.length,u.shape.length))}C.assertAxesAreInnerMostDims("sum",m,u.shape.length);let[f,g]=C.computeOutAndReduceShapes(u.shape,m),y=w.sizeFromShape(g),b=t.makeOutput(f,u.dtype);if(w.sizeFromShape(u.shape)!==0){let x=t.dataIdMap.get(b.dataId).id;rF(l,y,$t[b.dtype],x)}if(h&&t.disposeData(p.dataId),s){let x=C.expandShapeToKeepDim(b.shape,c);b.shape=x}return b}var hue={kernelName:co,backendName:"wasm",setupFunc:cue,kernelFunc:due},mue=dn(go),fue=dn(yo),sF;function gue(e){sF=e.wasm.cwrap(vs,null,["number","array","number","array","number","number"])}function yue(e){let{inputs:t,backend:n,attrs:a}=e,{x:r}=t,s=n.dataIdMap.get(r.dataId).id,{reps:i}=a,o=new Array(r.shape.length);for(let c=0;c<o.length;c++)o[c]=r.shape[c]*i[c];let l=new Uint8Array(new Int32Array(r.shape).buffer),u=new Uint8Array(new Int32Array(o).buffer),p=n.makeOutput(o,r.dtype),d=n.dataIdMap.get(p.dataId).id;return sF(s,l,r.shape.length,u,o.length,$t[p.dtype],d),p}var bue={kernelName:vs,backendName:"wasm",setupFunc:gue,kernelFunc:yue},iF;function xue(e){iF=e.wasm.cwrap(Mu,null,["number","array","number","number","number","bool","number","number"])}var vue=({inputs:e,backend:t,attrs:n})=>{let{x:a}=e,{k:r,sorted:s}=n,i=t.dataIdMap.get(a.dataId).id,o=new Uint8Array(new Int32Array(a.shape).buffer),l=a.shape.slice();l[l.length-1]=r;let u=t.makeOutput(l,a.dtype),p=t.dataIdMap.get(u.dataId).id,d=t.makeOutput(l,"int32"),c=t.dataIdMap.get(d.dataId).id;return iF(i,o,a.shape.length,$t[a.dtype],r,s,p,c),[u,d]},wue={kernelName:Mu,backendName:"wasm",setupFunc:xue,kernelFunc:vue},oF;function kue(e){oF=e.wasm.cwrap(Pu,null,["number","number","bool","number","number","number","number","number","number","array","number","number","number","number","number"])}function Iue(e){let{backend:t,inputs:n,attrs:a}=e,{image:r,transforms:s}=n,{interpolation:i,fillMode:o,fillValue:l,outputShape:u}=a,[p,d,c,h]=r.shape,[m,f]=u!=null?u:[d,c],g=[p,m,f,h],y=new Uint8Array(new Int32Array(w.computeStrides(r.shape)).buffer),b=t.makeOutput(g,r.dtype),x=t.dataIdMap.get(b.dataId).id,v=t.dataIdMap.get(r.dataId).id,k=t.dataIdMap.get(s.dataId).id,T=i==="nearest"?1:2,_;switch(o){case"constant":_=1;break;case"reflect":_=2;break;case"wrap":_=3;break;case"nearest":_=4;break;default:_=1;break}return oF(v,k,s.shape[0]>1,p,m,f,h,c,d,y,r.shape.length-1,T,_,l,x),b}var Sue={kernelName:Pu,backendName:"wasm",setupFunc:kue,kernelFunc:Iue};function Nue(e){let{inputs:t,backend:n,attrs:a}=e,{value:r}=t,{axis:s}=a;s<0&&(s+=r.shape.length);let i=r.shape[s],o=r.shape.length,l=new Array(o-1),u=0;for(let h=0;h<o;h++)h!==s&&(l[u++]=r.shape[h]);let p=new Array(i),d=new Array(o).fill(0),c=r.shape.slice();c[s]=1;for(let h=0;h<p.length;h++)d[s]=h,p[h]=bi({inputs:{x:r},attrs:{begin:d,size:c},backend:n});return p.map(({dataId:h,dtype:m})=>({dataId:h,dtype:m,shape:l}))}var Tue={kernelName:Ou,backendName:"wasm",kernelFunc:Nue};function _ue(e){let{inputs:{x:t},backend:n}=e,a=n.makeOutput(t.shape,t.dtype);return n.typedArrayFromHeap(a).fill(0),a}var Cue={kernelName:Lu,backendName:"wasm",kernelFunc:_ue},Eue=[yse,bse,vse,Ise,Ase,Rse,Ose,Bse,Gse,Yse,Qse,Jse,tie,nie,sie,lie,uie,pie,hie,gie,xie,kie,Nie,Tie,Cie,Eie,Fie,Aie,Rie,Mie,Oie,Bie,Uie,jie,Xie,Jie,eoe,noe,Sse,soe,ooe,uoe,poe,doe,foe,yoe,voe,Ioe,Toe,Coe,Aoe,Doe,Roe,Ooe,Boe,Uoe,Hoe,Koe,Yoe,Joe,WE,nle,sle,lle,ple,dle,hle,mle,Wse,yle,vle,Ile,Sle,Nle,Cle,Ale,Rle,Mle,Kse,Lle,Ble,Ule,jle,Kle,Yle,Jle,Zle,eue,nue,sue,lue,pue,hue,mue,fue,bue,wue,Sue,Cse,Tue,Cue];for(let e of Eue)Rc(e);var _x=X();_x.registerFlag("WASM_HAS_SIMD_SUPPORT",async()=>WebAssembly.validate(new Uint8Array([0,97,115,109,1,0,0,0,1,4,1,96,0,0,3,2,1,0,10,9,1,7,0,65,0,253,15,26,11])));_x.registerFlag("WASM_HAS_MULTITHREAD_SUPPORT",async()=>{if(_x.get("IS_NODE"))return!1;try{return new MessageChannel().port1.postMessage(new SharedArrayBuffer(1)),WebAssembly.validate(new Uint8Array([0,97,115,109,1,0,0,0,1,4,1,96,0,0,3,2,1,0,5,4,1,3,1,1,10,11,1,9,0,65,0,254,16,2,0,26,11]))}catch(e){return!1}});var fI=xi(w$()),Fue=`"use strict";var Module={};var ENVIRONMENT_IS_NODE=typeof process==="object"&&typeof process.versions==="object"&&typeof process.versions.node==="string";if(ENVIRONMENT_IS_NODE){var nodeWorkerThreads=require("worker_threads");var parentPort=nodeWorkerThreads.parentPort;parentPort.on("message",function(data){onmessage({data:data})});var fs=require("fs");Object.assign(global,{self:global,require:require,Module:Module,location:{href:__filename},Worker:nodeWorkerThreads.Worker,importScripts:function(f){(0,eval)(fs.readFileSync(f,"utf8"))},postMessage:function(msg){parentPort.postMessage(msg)},performance:global.performance||{now:function(){return Date.now()}}})}function threadPrintErr(){var text=Array.prototype.slice.call(arguments).join(" ");if(ENVIRONMENT_IS_NODE){fs.writeSync(2,text+"
");return}console.error(text)}function threadAlert(){var text=Array.prototype.slice.call(arguments).join(" ");postMessage({cmd:"alert",text:text,threadId:Module["_pthread_self"]()})}var err=threadPrintErr;self.alert=threadAlert;Module["instantiateWasm"]=((info,receiveInstance)=>{var instance=new WebAssembly.Instance(Module["wasmModule"],info);receiveInstance(instance);Module["wasmModule"]=null;return instance.exports});self.onmessage=(e=>{try{if(e.data.cmd==="load"){Module["wasmModule"]=e.data.wasmModule;Module["wasmMemory"]=e.data.wasmMemory;Module["buffer"]=Module["wasmMemory"].buffer;Module["ENVIRONMENT_IS_PTHREAD"]=true;if(typeof e.data.urlOrBlob==="string"){importScripts(e.data.urlOrBlob)}else{var objectUrl=URL.createObjectURL(e.data.urlOrBlob);importScripts(objectUrl);URL.revokeObjectURL(objectUrl)}WasmBackendModuleThreadedSimd(Module).then(function(instance){Module=instance})}else if(e.data.cmd==="run"){Module["__performance_now_clock_drift"]=performance.now()-e.data.time;Module["__emscripten_thread_init"](e.data.threadInfoStruct,0,0,1);Module["establishStackSpace"]();Module["PThread"].receiveObjectTransfer(e.data);Module["PThread"].threadInit();try{var result=Module["invokeEntryPoint"](e.data.start_routine,e.data.arg);if(Module["keepRuntimeAlive"]()){Module["PThread"].setExitStatus(result)}else{Module["__emscripten_thread_exit"](result)}}catch(ex){if(ex!="unwind"){if(ex instanceof Module["ExitStatus"]){if(Module["keepRuntimeAlive"]()){}else{Module["__emscripten_thread_exit"](ex.status)}}else{throw ex}}}}else if(e.data.cmd==="cancel"){if(Module["_pthread_self"]()){Module["__emscripten_thread_exit"](-1)}}else if(e.data.target==="setimmediate"){}else if(e.data.cmd==="processThreadQueue"){if(Module["_pthread_self"]()){Module["_emscripten_current_thread_process_queued_calls"]()}}else if(e.data.cmd==="processProxyingQueue"){if(Module["_pthread_self"]()){Module["_emscripten_proxy_execute_queue"](e.data.queue)}}else{err("worker.js received unknown command "+e.data.cmd);err(e.data)}}catch(ex){err("worker.js onmessage() captured an uncaught exception: "+ex);if(ex&&ex.stack)err(ex.stack);if(Module["__emscripten_thread_crashed"]){Module["__emscripten_thread_crashed"]()}throw ex}});`,Aue=xi(k$()),lF=class extends gc{constructor(e){super(),this.wasm=e,this.dataIdNextNumber=1,this.wasm.tfjs.initWithThreadsCount(uF),Cx=this.wasm.tfjs.getThreadsCount(),this.dataIdMap=new xm(this,sr())}write(e,t,n){let a={id:this.dataIdNextNumber++};return this.move(a,e,t,n,1),a}numDataIds(){return this.dataIdMap.numDataIds()}async time(e){let t=w.now();return e(),{kernelMs:w.now()-t}}move(e,t,n,a,r){let s=this.dataIdNextNumber++;if(a==="string"){let u=t;this.dataIdMap.set(e,{id:s,stringBytes:u,shape:n,dtype:a,memoryOffset:null,refCount:r});return}let i=w.sizeFromShape(n),o=i*w.bytesPerElement(a),l=this.wasm._malloc(o);this.dataIdMap.set(e,{id:s,memoryOffset:l,shape:n,dtype:a,refCount:r}),this.wasm.tfjs.registerTensor(s,i,l),t!=null&&this.wasm.HEAPU8.set(new Uint8Array(t.buffer,t.byteOffset,o),l)}async read(e){return this.readSync(e)}readSync(e,t,n){let{memoryOffset:a,dtype:r,shape:s,stringBytes:i}=this.dataIdMap.get(e);if(r==="string")return(t==null||t===0)&&(n==null||n>=i.length)?i:i.slice(t,n);t=t||0,n=n||w.sizeFromShape(s);let o=w.bytesPerElement(r),l=this.wasm.HEAPU8.slice(a+t*o,a+n*o);return Rue(l.buffer,r)}disposeData(e,t=!1){if(this.dataIdMap.has(e)){let n=this.dataIdMap.get(e);if(n.refCount--,!t&&n.refCount>0)return!1;this.wasm._free(n.memoryOffset),this.wasm.tfjs.disposeData(n.id),this.dataIdMap.delete(e)}return!0}refCount(e){return this.dataIdMap.has(e)?this.dataIdMap.get(e).refCount:0}incRef(e){let t=this.dataIdMap.get(e);t!=null&&t.refCount++}floatPrecision(){return 32}getMemoryOffset(e){return this.dataIdMap.get(e).memoryOffset}dispose(){this.wasm.tfjs.dispose(),"PThread"in this.wasm&&this.wasm.PThread.terminateAllThreads(),this.wasm=null}memory(){return{unreliable:!1}}makeOutput(e,t,n){let a;if(n==null)a=this.write(null,e,t);else{let r=this.dataIdNextNumber++;a={id:r},this.dataIdMap.set(a,{id:r,memoryOffset:n,shape:e,dtype:t,refCount:1});let s=w.sizeFromShape(e);this.wasm.tfjs.registerTensor(r,s,n)}return{dataId:a,shape:e,dtype:t}}typedArrayFromHeap({shape:e,dtype:t,dataId:n}){let a=this.wasm.HEAPU8.buffer,{memoryOffset:r}=this.dataIdMap.get(n),s=w.sizeFromShape(e);switch(t){case"float32":return new Float32Array(a,r,s);case"int32":return new Int32Array(a,r,s);case"bool":return new Uint8Array(a,r,s);default:throw new Error(`Unknown dtype ${t}`)}}};function $ue(e){return(t,n)=>(w.fetch(e,{credentials:"same-origin"}).then(a=>{a.ok||t.env.a(`failed to load wasm binary file at '${e}'`),a.arrayBuffer().then(r=>{WebAssembly.instantiate(r,t).then(s=>{n(s.instance,s.module)})})}),{})}function gI(e,t,n){if(bm!=null)return bm;let a="tfjs-backend-wasm.wasm";return e&&t?a="tfjs-backend-wasm-threaded-simd.wasm":e&&(a="tfjs-backend-wasm-simd.wasm"),Qp!=null&&Qp[a]!=null?Qp[a]:n+a}async function Due(){let[e,t]=await Promise.all([X().getAsync("WASM_HAS_SIMD_SUPPORT"),X().getAsync("WASM_HAS_MULTITHREAD_SUPPORT")]);return new Promise((n,a)=>{let r={};r.locateFile=(o,l)=>{if(o.endsWith(".worker.js")){let u=Fue.replace(/\n/g,"\\n"),p=new Blob([u],{type:"application/javascript"});return URL.createObjectURL(p)}return o.endsWith(".wasm")?gI(e,t,Kp!=null?Kp:l):l+o},H0&&(r.instantiateWasm=$ue(gI(e,t,Kp!=null?Kp:"")));let s=!1;r.onAbort=()=>{s||Jp||(Jp=!0,a({message:"Make sure the server can serve the `.wasm` file relative to the bundled js file. For more details see https://github.com/tensorflow/tfjs/blob/master/tfjs-backend-wasm/README.md#using-bundlers"}))};let i;t&&e&&bm==null?(r.mainScriptUrlOrBlob=new Blob(["var WasmBackendModuleThreadedSimd = "+fI.default.toString()],{type:"text/javascript"}),i=(0,fI.default)(r)):i=(0,Aue.default)(r),i.then(o=>{s=!0,Jp=!1;let l=null;o.tfjs={init:o.cwrap("init",null,[]),initWithThreadsCount:o.cwrap("init_with_threads_count",null,["number"]),getThreadsCount:o.cwrap("get_threads_count","number",[]),registerTensor:o.cwrap("register_tensor",null,["number","number","number"]),disposeData:o.cwrap("dispose_data",l,["number"]),dispose:o.cwrap("dispose",l,[])},n({wasm:o})})})}function Rue(e,t){switch(t){case"float32":return new Float32Array(e);case"int32":return new Int32Array(e);case"bool":return new Uint8Array(e);default:throw new Error(`Unknown dtype ${t}`)}}var Mue=["tfjs-backend-wasm.wasm","tfjs-backend-wasm-simd.wasm","tfjs-backend-wasm-threaded-simd.wasm"],bm=null,Kp=null,Qp={},Jp=!1,H0=!1;function Pue(e,t=!1){if(Ux("setWasmPath has been deprecated in favor of setWasmPaths and will be removed in a future release."),Jp)throw new Error("The WASM backend was already initialized. Make sure you call `setWasmPath()` before you call `tf.setBackend()` or `tf.ready()`");bm=e,H0=t}function Oue(e,t=!1){if(Jp)throw new Error("The WASM backend was already initialized. Make sure you call `setWasmPaths()` before you call `tf.setBackend()` or `tf.ready()`");if(typeof e=="string")Kp=e;else{Qp=e;let n=Mue.filter(a=>Qp[a]==null);if(n.length>0)throw new Error(`There were no entries found for the following binaries: ${n.join(",")}. Please either call setWasmPaths with a map providing a path for each binary, or with a string indicating the directory where all the binaries can be found.`)}H0=t}var uF=-1,Cx=-1;function Lue(e){uF=e}function zue(){if(Cx===-1)throw new Error("WASM backend not initialized.");return Cx}var Bue="3.18.0",Wue=2;nf("wasm",async()=>{let{wasm:e}=await Due();return new lF(e)},Wue);var Vue="3.18.0",Uue="3.18.0",Gue="3.18.0",Hue="3.18.0",jue="3.18.0",que="3.18.0",Kue="3.18.0",Xue="3.18.0",Yue={tfjs:Vue,"tfjs-core":Uue,"tfjs-data":Gue,"tfjs-layers":Hue,"tfjs-converter":jue,"tfjs-backend-cpu":que,"tfjs-backend-webgl":Kue,"tfjs-backend-wasm":Xue};var g1={};yh(g1,{AnchorPosition:()=>o1,DrawBox:()=>bd,DrawBoxOptions:()=>fg,DrawFaceLandmarks:()=>Tg,DrawFaceLandmarksOptions:()=>Ng,DrawTextField:()=>Br,DrawTextFieldOptions:()=>up,drawContour:()=>Or,drawDetections:()=>spe,drawFaceExpressions:()=>ppe,drawFaceLandmarks:()=>dpe});function Or(e,t,n=!1){if(e.beginPath(),t.slice(1).forEach(({x:a,y:r},s)=>{let i=t[s];e.moveTo(i.x,i.y),e.lineTo(a,r)}),n){let a=t[t.length-1],r=t[0];if(!a||!r)return;e.moveTo(a.x,a.y),e.lineTo(r.x,r.y)}e.stroke()}var X0={};yh(X0,{computeReshapedDimensions:()=>K0,getCenterPoint:()=>Eo,isDimensions:()=>hg,isEven:()=>dg,isFloat:()=>q0,isTensor:()=>_o,isTensor1D:()=>Que,isTensor2D:()=>j0,isTensor3D:()=>Lr,isTensor4D:()=>xa,isValidNumber:()=>er,isValidProbablitiy:()=>ip,range:()=>br,round:()=>Co});var yn=class{constructor(t,n){if(!er(t)||!er(n))throw new Error(`Dimensions.constructor - expected width and height to be valid numbers, instead have ${JSON.stringify({width:t,height:n})}`);this._width=t,this._height=n}get width(){return this._width}get height(){return this._height}reverse(){return new yn(1/this.width,1/this.height)}};function _o(e,t){return e instanceof Fe&&e.shape.length===t}function Que(e){return _o(e,1)}function j0(e){return _o(e,2)}function Lr(e){return _o(e,3)}function xa(e){return _o(e,4)}function q0(e){return e%1!==0}function dg(e){return e%2===0}function Co(e,t=2){let n=10**t;return Math.floor(e*n)/n}function hg(e){return e&&e.width&&e.height}function K0({width:e,height:t},n){let a=n/Math.max(t,e);return new yn(Math.round(e*a),Math.round(t*a))}function Eo(e){return e.reduce((t,n)=>t.add(n),new Pe(0,0)).div(new Pe(e.length,e.length))}function br(e,t,n){return Array(e).fill(0).map((a,r)=>t+r*n)}function er(e){return!!e&&e!==1/0&&e!==-1/0&&!Number.isNaN(e)||e===0}function ip(e){return er(e)&&e>=0&&e<=1}var Pe=class{constructor(t,n){this._x=t,this._y=n}get x(){return this._x}get y(){return this._y}add(t){return new Pe(this.x+t.x,this.y+t.y)}sub(t){return new Pe(this.x-t.x,this.y-t.y)}mul(t){return new Pe(this.x*t.x,this.y*t.y)}div(t){return new Pe(this.x/t.x,this.y/t.y)}abs(){return new Pe(Math.abs(this.x),Math.abs(this.y))}magnitude(){return Math.sqrt(this.x**2+this.y**2)}floor(){return new Pe(Math.floor(this.x),Math.floor(this.y))}};var lt=class{static isRect(t){return!!t&&[t.x,t.y,t.width,t.height].every(er)}static assertIsValidBox(t,n,a=!1){if(!lt.isRect(t))throw new Error(`${n} - invalid box: ${JSON.stringify(t)}, expected object with properties x, y, width, height`);if(!a&&(t.width<0||t.height<0))throw new Error(`${n} - width (${t.width}) and height (${t.height}) must be positive numbers`)}constructor(t,n=!0){let a=t||{},r=[a.left,a.top,a.right,a.bottom].every(er),s=[a.x,a.y,a.width,a.height].every(er);if(!s&&!r)throw new Error(`Box.constructor - expected box to be IBoundingBox | IRect, instead have ${JSON.stringify(a)}`);let[i,o,l,u]=s?[a.x,a.y,a.width,a.height]:[a.left,a.top,a.right-a.left,a.bottom-a.top];lt.assertIsValidBox({x:i,y:o,width:l,height:u},"Box.constructor",n),this._x=i,this._y=o,this._width=l,this._height=u}get x(){return this._x}get y(){return this._y}get width(){return this._width}get height(){return this._height}get left(){return this.x}get top(){return this.y}get right(){return this.x+this.width}get bottom(){return this.y+this.height}get area(){return this.width*this.height}get topLeft(){return new Pe(this.left,this.top)}get topRight(){return new Pe(this.right,this.top)}get bottomLeft(){return new Pe(this.left,this.bottom)}get bottomRight(){return new Pe(this.right,this.bottom)}round(){let[t,n,a,r]=[this.x,this.y,this.width,this.height].map(s=>Math.round(s));return new lt({x:t,y:n,width:a,height:r})}floor(){let[t,n,a,r]=[this.x,this.y,this.width,this.height].map(s=>Math.floor(s));return new lt({x:t,y:n,width:a,height:r})}toSquare(){let{x:t,y:n,width:a,height:r}=this,s=Math.abs(a-r);return a<r&&(t-=s/2,a+=s),r<a&&(n-=s/2,r+=s),new lt({x:t,y:n,width:a,height:r})}rescale(t){let n=hg(t)?t.width:t,a=hg(t)?t.height:t;return new lt({x:this.x*n,y:this.y*a,width:this.width*n,height:this.height*a})}pad(t,n){let[a,r,s,i]=[this.x-t/2,this.y-n/2,this.width+t,this.height+n];return new lt({x:a,y:r,width:s,height:i})}clipAtImageBorders(t,n){let{x:a,y:r,right:s,bottom:i}=this,o=Math.max(a,0),l=Math.max(r,0),u=s-o,p=i-l,d=Math.min(u,t-o),c=Math.min(p,n-l);return new lt({x:o,y:l,width:d,height:c}).floor()}shift(t,n){let{width:a,height:r}=this,s=this.x+t,i=this.y+n;return new lt({x:s,y:i,width:a,height:r})}padAtBorders(t,n){let a=this.width+1,r=this.height+1,s=1,i=1,o=a,l=r,u=this.left,p=this.top,d=this.right,c=this.bottom;return d>n&&(o=-d+n+a,d=n),c>t&&(l=-c+t+r,c=t),u<1&&(l=2-u,u=1),p<1&&(l=2-p,p=1),{dy:i,edy:l,dx:s,edx:o,y:p,ey:c,x:u,ex:d,w:a,h:r}}calibrate(t){return new lt({left:this.left+t.left*this.width,top:this.top+t.top*this.height,right:this.right+t.right*this.width,bottom:this.bottom+t.bottom*this.height}).toSquare().round()}};var Fo=class extends lt{constructor(t,n,a,r,s=!1){super({left:t,top:n,right:a,bottom:r},s)}};var zr=class{constructor(t,n,a,r,s){this._imageDims=new yn(s.width,s.height),this._score=t,this._classScore=n,this._className=a,this._box=new lt(r).rescale(this._imageDims)}get score(){return this._score}get classScore(){return this._classScore}get className(){return this._className}get box(){return this._box}get imageDims(){return this._imageDims}get imageWidth(){return this.imageDims.width}get imageHeight(){return this.imageDims.height}get relativeBox(){return new lt(this._box).rescale(this.imageDims.reverse())}forSize(t,n){return new zr(this.score,this.classScore,this.className,this.relativeBox,{width:t,height:n})}};var vt=class extends zr{constructor(t,n,a){super(t,t,"",n,a)}forSize(t,n){let{score:a,relativeBox:r,imageDims:s}=super.forSize(t,n);return new vt(a,r,s)}};function Y0(e,t,n=!0){let a=Math.max(0,Math.min(e.right,t.right)-Math.max(e.left,t.left)),r=Math.max(0,Math.min(e.bottom,t.bottom)-Math.max(e.top,t.top)),s=a*r;return n?s/(e.area+t.area-s):s/Math.min(e.area,t.area)}function Q0(e){let t=e.map(o=>o.x),n=e.map(o=>o.y),a=t.reduce((o,l)=>l<o?l:o,1/0),r=n.reduce((o,l)=>l<o?l:o,1/0),s=t.reduce((o,l)=>o<l?l:o,0),i=n.reduce((o,l)=>o<l?l:o,0);return new Fo(a,r,s,i)}function J0(e,t,n,a=!0){let r=t.map((i,o)=>({score:i,boxIndex:o})).sort((i,o)=>i.score-o.score).map(i=>i.boxIndex),s=[];for(;r.length>0;){let i=r.pop();s.push(i);let o=r,l=[];for(let u=0;u<o.length;u++){let p=o[u],d=e[i],c=e[p];l.push(Y0(d,c,a))}r=r.filter((u,p)=>l[p]<=n)}return s}function tr(e,t){return O(()=>{let[n,a,r]=t,s=Cn([...e.shape.slice(0,3),1],n,"float32"),i=Cn([...e.shape.slice(0,3),1],a,"float32"),o=Cn([...e.shape.slice(0,3),1],r,"float32"),l=Ze([s,i,o],3);return ce(e,l)})}function Z0(e,t=!1){return O(()=>{let[n,a]=e.shape.slice(1);if(n===a)return e;let r=Math.abs(n-a),s=Math.round(r*(t?.5:1)),i=n>a?2:1,o=c=>{let h=e.shape.slice();return h[i]=c,Cn(h,0,"float32")},l=o(s),u=r-l.shape[i],d=[t&&u?o(u):null,e,l].filter(c=>!!c).map(c=>oe(c,"float32"));return Ze(d,i)})}function Jue(e){let t=e.slice();for(let n=t.length-1;n>0;n--){let a=Math.floor(Math.random()*(n+1)),r=t[n];t[n]=t[a],t[a]=r}return t}function gd(e){return 1/(1+Math.exp(-e))}function Zue(e){return Math.log(e/(1-e))}var Ao=class extends lt{constructor(t,n,a,r,s=!1){super({x:t,y:n,width:a,height:r},s)}};var epe=.5,tpe=.43,npe=.45,ra=class{constructor(t,n,a=new Pe(0,0)){let{width:r,height:s}=n;this._imgDims=new yn(r,s),this._shift=a,this._positions=t.map(i=>i.mul(new Pe(r,s)).add(a))}get shift(){return new Pe(this._shift.x,this._shift.y)}get imageWidth(){return this._imgDims.width}get imageHeight(){return this._imgDims.height}get positions(){return this._positions}get relativePositions(){return this._positions.map(t=>t.sub(this._shift).div(new Pe(this.imageWidth,this.imageHeight)))}forSize(t,n){return new this.constructor(this.relativePositions,{width:t,height:n})}shiftBy(t,n){return new this.constructor(this.relativePositions,this._imgDims,new Pe(t,n))}shiftByPoint(t){return this.shiftBy(t.x,t.y)}align(t,n={}){if(t){let s=t instanceof vt?t.box.floor():new lt(t);return this.shiftBy(s.x,s.y).align(null,n)}let{useDlibAlignment:a,minBoxPadding:r}={useDlibAlignment:!1,minBoxPadding:.2,...n};return a?this.alignDlib():this.alignMinBbox(r)}alignDlib(){let t=this.getRefPointsForAlignment(),[n,a,r]=t,s=d=>r.sub(d).magnitude(),i=(s(n)+s(a))/2,o=Math.floor(i/npe),l=Eo(t),u=Math.floor(Math.max(0,l.x-epe*o)),p=Math.floor(Math.max(0,l.y-tpe*o));return new Ao(u,p,Math.min(o,this.imageWidth+u),Math.min(o,this.imageHeight+p))}alignMinBbox(t){let n=Q0(this.positions);return n.pad(n.width*t,n.height*t)}getRefPointsForAlignment(){throw new Error("getRefPointsForAlignment not implemented by base class")}};var e1=class extends ra{getRefPointsForAlignment(){let t=this.positions;return[t[0],t[1],Eo([t[3],t[4]])]}};var $o=class extends ra{getJawOutline(){return this.positions.slice(0,17)}getLeftEyeBrow(){return this.positions.slice(17,22)}getRightEyeBrow(){return this.positions.slice(22,27)}getNose(){return this.positions.slice(27,36)}getLeftEye(){return this.positions.slice(36,42)}getRightEye(){return this.positions.slice(42,48)}getMouth(){return this.positions.slice(48,68)}getRefPointsForAlignment(){return[this.getLeftEye(),this.getRightEye(),this.getMouth()].map(Eo)}};var op=class{constructor(t,n){this._label=t,this._distance=n}get label(){return this._label}get distance(){return this._distance}toString(t=!0){return`${this.label}${t?` (${Co(this.distance)})`:""}`}};var lp=class extends lt{constructor(n,a){super(n);this._label=a}static assertIsValidLabeledBox(n,a){if(lt.assertIsValidBox(n,a),!er(n.label))throw new Error(`${a} - expected property label (${n.label}) to be a number`)}get label(){return this._label}};var xr=class{constructor(t,n){if(typeof t!="string")throw new Error("LabeledFaceDescriptors - constructor expected label to be a string");if(!Array.isArray(n)||n.some(a=>!(a instanceof Float32Array)))throw new Error("LabeledFaceDescriptors - constructor expected descriptors to be an array of Float32Array");this._label=t,this._descriptors=n}get label(){return this._label}get descriptors(){return this._descriptors}toJSON(){return{label:this.label,descriptors:this.descriptors.map(t=>Array.from(t))}}static fromJSON(t){let n=t.descriptors.map(a=>new Float32Array(a));return new xr(t.label,n)}};var t1=class extends lp{constructor(n,a,r,s){super(n,a);this._score=r,this._classScore=s}static assertIsValidPredictedBox(n,a){if(lp.assertIsValidLabeledBox(n,a),!ip(n.score)||!ip(n.classScore))throw new Error(`${a} - expected properties score (${n.score}) and (${n.classScore}) to be a number between [0, 1]`)}get score(){return this._score}get classScore(){return this._classScore}};function vr(e){return e.detection instanceof vt}function Do(e,t){return{...e,...{detection:t}}}function n1(){let e=window.fetch;if(!e)throw new Error("fetch - missing fetch implementation for browser environment");return{Canvas:HTMLCanvasElement,CanvasRenderingContext2D,Image:HTMLImageElement,ImageData,Video:HTMLVideoElement,createCanvasElement:()=>document.createElement("canvas"),createImageElement:()=>document.createElement("img"),createVideoElement:()=>document.createElement("video"),fetch:e,readFile:()=>{throw new Error("readFile - filesystem not available for browser environment")}}}function yd(){return typeof global=="object"&&typeof process!="undefined"&&process.versions!=null&&process.versions.node!=null}function mg(e){let t="";if(!e&&yd())try{e=jA("fs")}catch(a){t=a.toString()}return{readFile:e?a=>new Promise((r,s)=>{e.readFile(a,(i,o)=>i?s(i):r(o))}):()=>{throw new Error(`readFile - failed to require fs in nodejs environment with error: ${t}`)}}}function a1(){let e=global.Canvas||global.HTMLCanvasElement,t=global.Image||global.HTMLImageElement,n=global.Video||global.HTMLVideoElement,a=()=>{if(e)return new e;throw new Error("createCanvasElement - missing Canvas implementation for nodejs environment")},r=()=>{if(t)return new t;throw new Error("createImageElement - missing Image implementation for nodejs environment")},s=()=>{if(n)return new n;throw new Error("createVideoElement - missing Video implementation for nodejs environment")},i=global.fetch,o=mg();return{Canvas:e||class{},CanvasRenderingContext2D:global.CanvasRenderingContext2D||class{},Image:t||class{},ImageData:global.ImageData||class{},Video:global.HTMLVideoElement||class{},createCanvasElement:a,createImageElement:r,createVideoElement:s,fetch:i,...o}}function r1(){return typeof window=="object"&&typeof document!="undefined"&&typeof HTMLImageElement!="undefined"&&typeof HTMLCanvasElement!="undefined"&&typeof HTMLVideoElement!="undefined"&&typeof ImageData!="undefined"&&typeof CanvasRenderingContext2D!="undefined"}var on;function ape(){if(!on)throw new Error("getEnv - environment is not defined, check isNodejs() and isBrowser()");return on}function s1(e){on=e}function i1(){return r1()?s1(n1()):yd()?s1(a1()):null}function rpe(e){if(on||i1(),!on)throw new Error("monkeyPatch - environment is not defined, check isNodejs() and isBrowser()");let{Canvas:t=on.Canvas,Image:n=on.Image}=e;on.Canvas=t,on.Image=n,on.createCanvasElement=e.createCanvasElement||(()=>new t),on.createImageElement=e.createImageElement||(()=>new n),on.ImageData=e.ImageData||on.ImageData,on.Video=e.Video||on.Video,on.fetch=e.fetch||on.fetch,on.readFile=e.readFile||on.readFile}var et={getEnv:ape,setEnv:s1,initialize:i1,createBrowserEnv:n1,createFileSystem:mg,createNodejsEnv:a1,monkeyPatch:rpe,isBrowser:r1,isNodejs:yd};i1();function Ro(e){return!et.isNodejs()&&typeof e=="string"?document.getElementById(e):e}function jn(e){let{Canvas:t,CanvasRenderingContext2D:n}=et.getEnv();if(e instanceof n)return e;let a=Ro(e);if(!(a instanceof t))throw new Error("resolveContext2d - expected canvas to be of instance of Canvas");let r=a.getContext("2d");if(!r)throw new Error("resolveContext2d - canvas 2d context is null");return r}var o1=(r=>(r.TOP_LEFT="TOP_LEFT",r.TOP_RIGHT="TOP_RIGHT",r.BOTTOM_LEFT="BOTTOM_LEFT",r.BOTTOM_RIGHT="BOTTOM_RIGHT",r))(o1||{}),up=class{constructor(t={}){let{anchorPosition:n,backgroundColor:a,fontColor:r,fontSize:s,fontStyle:i,padding:o}=t;this.anchorPosition=n||"TOP_LEFT",this.backgroundColor=a||"rgba(0, 0, 0, 0.5)",this.fontColor=r||"rgba(255, 255, 255, 1)",this.fontSize=s||14,this.fontStyle=i||"Georgia",this.padding=o||4}},Br=class{constructor(t,n,a={}){this.text=typeof t=="string"?[t]:t instanceof Br?t.text:t,this.anchor=n,this.options=new up(a)}measureWidth(t){let{padding:n}=this.options;return this.text.map(a=>t.measureText(a).width).reduce((a,r)=>a<r?r:a,0)+2*n}measureHeight(){let{fontSize:t,padding:n}=this.options;return this.text.length*t+2*n}getUpperLeft(t,n){let{anchorPosition:a}=this.options,r=a==="BOTTOM_RIGHT"||a==="TOP_RIGHT",s=a==="BOTTOM_LEFT"||a==="BOTTOM_RIGHT",i=this.measureWidth(t),o=this.measureHeight(),l=r?this.anchor.x-i:this.anchor.x,u=s?this.anchor.y-o:this.anchor.y;if(n){let{width:p,height:d}=n,c=Math.max(Math.min(l,p-i),0),h=Math.max(Math.min(u,d-o),0);return{x:c,y:h}}return{x:l,y:u}}draw(t){let n=Ro(t),a=jn(n),{backgroundColor:r,fontColor:s,fontSize:i,fontStyle:o,padding:l}=this.options;a.font=`${i}px ${o}`;let u=this.measureWidth(a),p=this.measureHeight();a.fillStyle=r;let d=this.getUpperLeft(a,n);a.fillRect(d.x,d.y,u,p),a.fillStyle=s,this.text.forEach((c,h)=>{let m=l+d.x,f=l+d.y+(h+1)*i;a.fillText(c,m,f)})}};var fg=class{constructor(t={}){let{boxColor:n,lineWidth:a,label:r,drawLabelOptions:s}=t;this.boxColor=n||"rgba(0, 0, 255, 1)",this.lineWidth=a||2,this.label=r;let i={anchorPosition:"BOTTOM_LEFT",backgroundColor:this.boxColor};this.drawLabelOptions=new up({...i,...s})}},bd=class{constructor(t,n={}){this.box=new lt(t),this.options=new fg(n)}draw(t){let n=jn(t),{boxColor:a,lineWidth:r}=this.options,{x:s,y:i,width:o,height:l}=this.box;n.strokeStyle=a,n.lineWidth=r,n.strokeRect(s,i,o,l);let{label:u}=this.options;u&&new Br([u],{x:s-r/2,y:i},this.options.drawLabelOptions).draw(t)}};function spe(e,t){(Array.isArray(t)?t:[t]).forEach(a=>{let r=a instanceof vt?a.score:vr(a)?a.detection.score:void 0,s=a instanceof vt?a.box:vr(a)?a.detection.box:new lt(a),i=r?`${Co(r)}`:void 0;new bd(s,{label:i}).draw(e)})}function xd(e){let{Image:t,Video:n}=et.getEnv();return e instanceof t&&e.complete||e instanceof n&&e.readyState>=3}function l1(e){return new Promise((t,n)=>{(e instanceof et.getEnv().Canvas||xd(e))&&t(null);function a(s){!s.currentTarget||(s.currentTarget.removeEventListener("load",r),s.currentTarget.removeEventListener("error",a),n(s))}function r(s){!s.currentTarget||(s.currentTarget.removeEventListener("load",r),s.currentTarget.removeEventListener("error",a),t(s))}e.addEventListener("load",r),e.addEventListener("error",a)})}function u1(e){return new Promise((t,n)=>{e instanceof Blob||n(new Error("bufferToImage - expected buf to be of type: Blob"));let a=new FileReader;a.onload=()=>{typeof a.result!="string"&&n(new Error("bufferToImage - expected reader.result to be a string, in onload"));let r=et.getEnv().createImageElement();r.onload=()=>t(r),r.onerror=n,r.src=a.result},a.onerror=n,a.readAsDataURL(e)})}function Mo(e){let{Image:t,Video:n}=et.getEnv();return e instanceof t?new yn(e.naturalWidth,e.naturalHeight):e instanceof n?new yn(e.videoWidth,e.videoHeight):new yn(e.width,e.height)}function Po({width:e,height:t}){let{createCanvasElement:n}=et.getEnv(),a=n();return a.width=e,a.height=t,a}function vd(e,t){let{ImageData:n}=et.getEnv();if(!(e instanceof n)&&!xd(e))throw new Error("createCanvasFromMedia - media has not finished loading yet");let{width:a,height:r}=t||Mo(e),s=Po({width:a,height:r});return e instanceof n?jn(s).putImageData(e,0,0):jn(s).drawImage(e,0,0,a,r),s}async function p1(e,t){let n=t||et.getEnv().createCanvasElement(),[a,r,s]=e.shape.slice(xa(e)?1:0),i=O(()=>e.as3D(a,r,s).toInt());return await bo.toPixels(i,n),i.dispose(),n}function gg(e){let{Image:t,Canvas:n,Video:a}=et.getEnv();return e instanceof t||e instanceof n||e instanceof a}function c1(e,t,n=!1){let{Image:a,Canvas:r}=et.getEnv();if(!(e instanceof a||e instanceof r))throw new Error("imageToSquare - expected arg0 to be HTMLImageElement | HTMLCanvasElement");if(t<=0)return Po({width:1,height:1});let s=Mo(e),i=t/Math.max(s.height,s.width),o=i*s.width,l=i*s.height,u=Po({width:t,height:t}),p=e instanceof r?e:vd(e),d=Math.abs(o-l)/2,c=n&&o<l?d:0,h=n&&l<o?d:0;return p.width>0&&p.height>0&&jn(u).drawImage(p,c,h,o,l),u}var wr=class{constructor(t,n=!1){this._imageTensors=[];this._canvases=[];this._treatAsBatchInput=!1;this._inputDimensions=[];this._inputSize=0;if(!Array.isArray(t))throw new Error(`NetInput.constructor - expected inputs to be an Array of TResolvedNetInput or to be instanceof tf.Tensor4D, instead have ${t}`);this._treatAsBatchInput=n,this._batchSize=t.length,t.forEach((a,r)=>{if(Lr(a)){this._imageTensors[r]=a,this._inputDimensions[r]=a.shape;return}if(xa(a)){let i=a.shape[0];if(i!==1)throw new Error(`NetInput - tf.Tensor4D with batchSize ${i} passed, but not supported in input array`);this._imageTensors[r]=a,this._inputDimensions[r]=a.shape.slice(1);return}let s=a instanceof et.getEnv().Canvas?a:vd(a);this._canvases[r]=s,this._inputDimensions[r]=[s.height,s.width,3]})}get imageTensors(){return this._imageTensors}get canvases(){return this._canvases}get isBatchInput(){return this.batchSize>1||this._treatAsBatchInput}get batchSize(){return this._batchSize}get inputDimensions(){return this._inputDimensions}get inputSize(){return this._inputSize}get reshapedInputDimensions(){return br(this.batchSize,0,1).map((t,n)=>this.getReshapedInputDimensions(n))}getInput(t){return this.canvases[t]||this.imageTensors[t]}getInputDimensions(t){return this._inputDimensions[t]}getInputHeight(t){return this._inputDimensions[t][0]}getInputWidth(t){return this._inputDimensions[t][1]}getReshapedInputDimensions(t){if(typeof this.inputSize!="number")throw new Error("getReshapedInputDimensions - inputSize not set, toBatchTensor has not been called yet");let n=this.getInputWidth(t),a=this.getInputHeight(t);return K0({width:n,height:a},this.inputSize)}toBatchTensor(t,n=!0){return this._inputSize=t,O(()=>{let a=br(this.batchSize,0,1).map(s=>{let i=this.getInput(s);if(i instanceof Fe){let o=xa(i)?i:mn(i);return o=Z0(o,n),(o.shape[1]!==t||o.shape[2]!==t)&&(o=Ln.resizeBilinear(o,[t,t],!1,!1)),o.as3D(t,t,3)}if(i instanceof et.getEnv().Canvas)return bo.fromPixels(c1(i,t,n));throw new Error(`toBatchTensor - at batchIdx ${s}, expected input to be instanceof tf.Tensor or instanceof HTMLCanvasElement, instead have ${i}`)});return Mt(a.map(s=>oe(s,"float32"))).as4D(this.batchSize,t,t,3)})}};async function wt(e){if(e instanceof wr)return e;let t=Array.isArray(e)?e:[e];if(!t.length)throw new Error("toNetInput - empty array passed as input");let n=r=>Array.isArray(e)?` at input index ${r}:`:"",a=t.map(Ro);return a.forEach((r,s)=>{if(!gg(r)&&!Lr(r)&&!xa(r))throw typeof t[s]=="string"?new Error(`toNetInput -${n(s)} string passed, but could not resolve HTMLElement for element id ${t[s]}`):new Error(`toNetInput -${n(s)} expected media to be of type HTMLImageElement | HTMLVideoElement | HTMLCanvasElement | tf.Tensor3D, or to be an element id`);if(xa(r)){let i=r.shape[0];if(i!==1)throw new Error(`toNetInput -${n(s)} tf.Tensor4D with batchSize ${i} passed, but not supported in input array`)}}),await Promise.all(a.map(r=>gg(r)&&l1(r))),new wr(a,Array.isArray(e))}async function pp(e,t){let{Canvas:n}=et.getEnv(),a=e;if(!(e instanceof n)){let i=await wt(e);if(i.batchSize>1)throw new Error("extractFaces - batchSize > 1 not supported");let o=i.getInput(0);a=o instanceof n?o:await p1(o)}let r=jn(a);return t.map(i=>i instanceof vt?i.forSize(a.width,a.height).box.floor():i).map(i=>i.clipAtImageBorders(a.width,a.height)).map(({x:i,y:o,width:l,height:u})=>{let p=Po({width:l,height:u});return l>0&&u>0&&jn(p).putImageData(r.getImageData(i,o,l,u),0,0),p})}async function cp(e,t){if(!Lr(e)&&!xa(e))throw new Error("extractFaceTensors - expected image tensor to be 3D or 4D");if(xa(e)&&e.shape[0]>1)throw new Error("extractFaceTensors - batchSize > 1 not supported");return O(()=>{let[n,a,r]=e.shape.slice(xa(e)?1:0);return t.map(o=>o instanceof vt?o.forSize(a,n).box:o).map(o=>o.clipAtImageBorders(a,n)).filter(o=>o.width>0&&o.height>0).map(({x:o,y:l,width:u,height:p})=>Hu(e.as3D(n,a,r),[l,o,0],[p,u,r]))})}async function Wr(e,t){let{fetch:n}=et.getEnv(),a=await n(e,t);if(!(a.status<400))throw new Error(`failed to fetch: (${a.status}) ${a.statusText}, from url: ${a.url}`);return a}async function ipe(e){let t=await Wr(e),n=await t.blob();if(!n.type.startsWith("image/"))throw new Error(`fetchImage - expected blob type to be of type image/*, instead have: ${n.type}, for url: ${t.url}`);return u1(n)}async function d1(e){return(await Wr(e)).json()}async function ope(e){return new Float32Array(await(await Wr(e)).arrayBuffer())}function pF(e){return new Promise((t,n)=>{e instanceof Blob||n(new Error("bufferToVideo - expected buf to be of type: Blob"));let a=et.getEnv().createVideoElement();a.oncanplay=()=>t(a),a.onerror=n,a.playsInline=!0,a.muted=!0,a.src=URL.createObjectURL(e),a.play()})}async function lpe(e){let t=await Wr(e),n=await t.blob();if(!n.type.startsWith("video/"))throw new Error(`fetchVideo - expected blob type to be of type video/*, instead have: ${n.type}, for url: ${t.url}`);return pF(n)}function yg(e,t){let n=`${t}-weights_manifest.json`;if(!e)return{modelBaseUri:"",manifestUri:n};if(e==="/")return{modelBaseUri:"/",manifestUri:`/${n}`};let a=e.startsWith("http://")?"http://":e.startsWith("https://")?"https://":"";e=e.replace(a,"");let r=e.split("/").filter(o=>o),s=e.endsWith(".json")?r[r.length-1]:n,i=a+(e.endsWith(".json")?r.slice(0,r.length-1):r).join("/");return i=e.startsWith("/")?`/${i}`:i,{modelBaseUri:i,manifestUri:i==="/"?`/${s}`:`${i}/${s}`}}async function h1(e,t){let{manifestUri:n,modelBaseUri:a}=yg(e,t),r=await d1(n);return en.loadWeights(r,a)}function upe(e,t,n=!1){let{width:a,height:r}=n?Mo(t):t;return e.width=a,e.height=r,{width:a,height:r}}var ln=class{constructor(t){this._params=void 0;this._paramMappings=[];this._name=t}get params(){return this._params}get paramMappings(){return this._paramMappings}get isLoaded(){return!!this.params}getParamFromPath(t){let{obj:n,objProp:a}=this.traversePropertyPath(t);return n[a]}reassignParamFromPath(t,n){let{obj:a,objProp:r}=this.traversePropertyPath(t);a[r].dispose(),a[r]=n}getParamList(){return this._paramMappings.map(({paramPath:t})=>({path:t,tensor:this.getParamFromPath(t)}))}getTrainableParams(){return this.getParamList().filter(t=>t.tensor instanceof ls)}getFrozenParams(){return this.getParamList().filter(t=>!(t.tensor instanceof ls))}variable(){this.getFrozenParams().forEach(({path:t,tensor:n})=>{this.reassignParamFromPath(t,n.variable())})}freeze(){this.getTrainableParams().forEach(({path:t,tensor:n})=>{let a=Jn(n.dataSync());n.dispose(),this.reassignParamFromPath(t,a)})}dispose(t=!0){this.getParamList().forEach(n=>{if(t&&n.tensor.isDisposed)throw new Error(`param tensor has already been disposed for path ${n.path}`);n.tensor.dispose()}),this._params=void 0}serializeParams(){return new Float32Array(this.getParamList().map(({tensor:t})=>Array.from(t.dataSync())).reduce((t,n)=>t.concat(n)))}async load(t){if(t instanceof Float32Array){this.extractWeights(t);return}await this.loadFromUri(t)}async loadFromUri(t){if(t&&typeof t!="string")throw new Error(`${this._name}.loadFromUri - expected model uri`);let n=await h1(t,this.getDefaultModelName());this.loadFromWeightMap(n)}async loadFromDisk(t){if(t&&typeof t!="string")throw new Error(`${this._name}.loadFromDisk - expected model file path`);let{readFile:n}=et.getEnv(),{manifestUri:a,modelBaseUri:r}=yg(t,this.getDefaultModelName()),s=u=>Promise.all(u.map(p=>n(p).then(d=>d.buffer))),i=en.weightsLoaderFactory(s),o=JSON.parse((await n(a)).toString()),l=await i(o,r);this.loadFromWeightMap(l)}loadFromWeightMap(t){let{paramMappings:n,params:a}=this.extractParamsFromWeightMap(t);this._paramMappings=n,this._params=a}extractWeights(t){let{paramMappings:n,params:a}=this.extractParams(t);this._paramMappings=n,this._params=a}traversePropertyPath(t){if(!this.params)throw new Error("traversePropertyPath - model has no loaded params");let n=t.split("/").reduce((s,i)=>{if(!s.nextObj.hasOwnProperty(i))throw new Error(`traversePropertyPath - object does not have property ${i}, for path ${t}`);return{obj:s.nextObj,objProp:i,nextObj:s.nextObj[i]}},{nextObj:this.params}),{obj:a,objProp:r}=n;if(!a||!r||!(a[r]instanceof Fe))throw new Error(`traversePropertyPath - parameter is not a tensor, for path ${t}`);return{obj:a,objProp:r}}};function qn(e,t,n){return O(()=>{let a=vo(e,t.depthwise_filter,t.pointwise_filter,n,"same");return a=J(a,t.bias),a})}function bg(e,t,n=!1){return O(()=>{let a=Xe(n?J(Rt(e,t.conv0.filters,[2,2],"same"),t.conv0.bias):qn(e,t.conv0,[2,2])),r=qn(a,t.conv1,[1,1]),s=Xe(J(a,r)),i=qn(s,t.conv2,[1,1]);return Xe(J(a,J(r,i)))})}function wd(e,t,n=!1,a=!0){return O(()=>{let r=Xe(n?J(Rt(e,t.conv0.filters,a?[2,2]:[1,1],"same"),t.conv0.bias):qn(e,t.conv0,a?[2,2]:[1,1])),s=qn(r,t.conv1,[1,1]),i=Xe(J(r,s)),o=qn(i,t.conv2,[1,1]),l=Xe(J(r,J(s,o))),u=qn(l,t.conv3,[1,1]);return Xe(J(r,J(s,J(o,u))))})}function Oo(e,t,n="same",a=!1){return O(()=>{let r=J(Rt(e,t.filters,[1,1],n),t.bias);return a?Xe(r):r})}function An(e,t){Object.keys(e).forEach(n=>{t.some(a=>a.originalPath===n)||e[n].dispose()})}function dp(e,t){return(n,a,r,s)=>{let i=Ja(e(n*a*r*r),[r,r,n,a]),o=qe(e(a));return t.push({paramPath:`${s}/filters`},{paramPath:`${s}/bias`}),{filters:i,bias:o}}}function xg(e,t){return(n,a,r)=>{let s=Ha(e(n*a),[n,a]),i=qe(e(a));return t.push({paramPath:`${r}/weights`},{paramPath:`${r}/bias`}),{weights:s,bias:i}}}var kd=class{constructor(t,n,a){this.depthwise_filter=t;this.pointwise_filter=n;this.bias=a}};function hp(e,t){return(n,a,r)=>{let s=Ja(e(9*n),[3,3,n,1]),i=Ja(e(n*a),[1,1,n,a]),o=qe(e(a));return t.push({paramPath:`${r}/depthwise_filter`},{paramPath:`${r}/pointwise_filter`},{paramPath:`${r}/bias`}),new kd(s,i,o)}}function mp(e){return t=>{let n=e(`${t}/depthwise_filter`,4),a=e(`${t}/pointwise_filter`,4),r=e(`${t}/bias`,1);return new kd(n,a,r)}}function sa(e,t){return(n,a,r)=>{let s=e[n];if(!_o(s,a))throw new Error(`expected weightMap[${n}] to be a Tensor${a}D, instead have ${s}`);return t.push({originalPath:n,paramPath:r||n}),s}}function $n(e){let t=e;function n(r){let s=t.slice(0,r);return t=t.slice(r),s}function a(){return t}return{extractWeights:n,getRemainingWeights:a}}function vg(e,t){let n=dp(e,t),a=hp(e,t);function r(i,o,l,u=!1){let p=u?n(i,o,3,`${l}/conv0`):a(i,o,`${l}/conv0`),d=a(o,o,`${l}/conv1`),c=a(o,o,`${l}/conv2`);return{conv0:p,conv1:d,conv2:c}}function s(i,o,l,u=!1){let{conv0:p,conv1:d,conv2:c}=r(i,o,l,u),h=a(o,o,`${l}/conv3`);return{conv0:p,conv1:d,conv2:c,conv3:h}}return{extractDenseBlock3Params:r,extractDenseBlock4Params:s}}function cF(e){let t=[],{extractWeights:n,getRemainingWeights:a}=$n(e),{extractDenseBlock4Params:r}=vg(n,t),s=r(3,32,"dense0",!0),i=r(32,64,"dense1"),o=r(64,128,"dense2"),l=r(128,256,"dense3");if(a().length!==0)throw new Error(`weights remaing after extract: ${a().length}`);return{paramMappings:t,params:{dense0:s,dense1:i,dense2:o,dense3:l}}}function wg(e){return t=>{let n=e(`${t}/filters`,4),a=e(`${t}/bias`,1);return{filters:n,bias:a}}}function kg(e,t){let n=sa(e,t),a=wg(n),r=mp(n);function s(o,l=!1){let u=l?a(`${o}/conv0`):r(`${o}/conv0`),p=r(`${o}/conv1`),d=r(`${o}/conv2`);return{conv0:u,conv1:p,conv2:d}}function i(o,l=!1){let u=l?a(`${o}/conv0`):r(`${o}/conv0`),p=r(`${o}/conv1`),d=r(`${o}/conv2`),c=r(`${o}/conv3`);return{conv0:u,conv1:p,conv2:d,conv3:c}}return{extractDenseBlock3Params:s,extractDenseBlock4Params:i}}function dF(e){let t=[],{extractDenseBlock4Params:n}=kg(e,t),a={dense0:n("dense0",!0),dense1:n("dense1"),dense2:n("dense2"),dense3:n("dense3")};return An(e,t),{params:a,paramMappings:t}}var fp=class extends ln{constructor(){super("FaceFeatureExtractor")}forwardInput(t){let{params:n}=this;if(!n)throw new Error("FaceFeatureExtractor - load model before inference");return O(()=>{let a=oe(t.toBatchTensor(112,!0),"float32"),s=tr(a,[122.782,117.001,104.298]).div(255),i=wd(s,n.dense0,!0);return i=wd(i,n.dense1),i=wd(i,n.dense2),i=wd(i,n.dense3),i=ga(i,[7,7],[2,2],"valid"),i})}async forward(t){return this.forwardInput(await wt(t))}getDefaultModelName(){return"face_feature_extractor_model"}extractParamsFromWeightMap(t){return dF(t)}extractParams(t){return cF(t)}};function Id(e,t){return O(()=>J(De(e,t.weights),t.bias))}function hF(e,t,n){let a=[],{extractWeights:r,getRemainingWeights:s}=$n(e),o=xg(r,a)(t,n,"fc");if(s().length!==0)throw new Error(`weights remaing after extract: ${s().length}`);return{paramMappings:a,params:{fc:o}}}function mF(e){let t=[],n=sa(e,t);function a(s){let i=n(`${s}/weights`,2),o=n(`${s}/bias`,1);return{weights:i,bias:o}}let r={fc:a("fc")};return An(e,t),{params:r,paramMappings:t}}function Ig(e){let t={},n={};return Object.keys(e).forEach(a=>{let r=a.startsWith("fc")?n:t;r[a]=e[a]}),{featureExtractorMap:t,classifierMap:n}}var gp=class extends ln{constructor(n,a){super(n);this._faceFeatureExtractor=a}get faceFeatureExtractor(){return this._faceFeatureExtractor}runNet(n){let{params:a}=this;if(!a)throw new Error(`${this._name} - load model before inference`);return O(()=>{let r=n instanceof wr?this.faceFeatureExtractor.forwardInput(n):n;return Id(r.as2D(r.shape[0],-1),a.fc)})}dispose(n=!0){this.faceFeatureExtractor.dispose(n),super.dispose(n)}loadClassifierParams(n){let{params:a,paramMappings:r}=this.extractClassifierParams(n);this._params=a,this._paramMappings=r}extractClassifierParams(n){return hF(n,this.getClassifierChannelsIn(),this.getClassifierChannelsOut())}extractParamsFromWeightMap(n){let{featureExtractorMap:a,classifierMap:r}=Ig(n);return this.faceFeatureExtractor.loadFromWeightMap(a),mF(r)}extractParams(n){let a=this.getClassifierChannelsIn(),r=this.getClassifierChannelsOut(),s=r*a+r,i=n.slice(0,n.length-s),o=n.slice(n.length-s);return this.faceFeatureExtractor.extractWeights(i),this.extractClassifierParams(o)}};var m1=["neutral","happy","sad","angry","fearful","disgusted","surprised"],Vr=class{constructor(t){this.neutral=0;this.happy=0;this.sad=0;this.angry=0;this.fearful=0;this.disgusted=0;this.surprised=0;if(t.length!==7)throw new Error(`FaceExpressions.constructor - expected probabilities.length to be 7, have: ${t.length}`);m1.forEach((n,a)=>{this[n]=t[a]})}asSortedArray(){return m1.map(t=>({expression:t,probability:this[t]})).sort((t,n)=>n.probability-t.probability)}};var Sd=class extends gp{constructor(t=new fp){super("FaceExpressionNet",t)}forwardInput(t){return O(()=>Qa(this.runNet(t)))}async forward(t){return this.forwardInput(await wt(t))}async predictExpressions(t){let n=await wt(t),a=await this.forwardInput(n),r=await Promise.all(ht(a).map(async i=>{let o=i.dataSync();return i.dispose(),o}));a.dispose();let s=r.map(i=>new Vr(i));return n.isBatchInput?s:s[0]}getDefaultModelName(){return"face_expression_model"}getClassifierChannelsIn(){return 256}getClassifierChannelsOut(){return 7}};function f1(e){return e.expressions instanceof Vr}function Sg(e,t){return{...e,...{expressions:t}}}function ppe(e,t,n=.1,a){(Array.isArray(t)?t:[t]).forEach(s=>{let i=s instanceof Vr?s:f1(s)?s.expressions:void 0;if(!i)throw new Error("drawFaceExpressions - expected faceExpressions to be FaceExpressions | WithFaceExpressions<{}> or array thereof");let l=i.asSortedArray().filter(d=>d.probability>n),u=vr(s)?s.detection.box.bottomLeft:a||new Pe(0,0);new Br(l.map(d=>`${d.expression} (${Co(d.probability)})`),u).draw(e)})}function Lo(e){return vr(e)&&e.landmarks instanceof ra&&e.unshiftedLandmarks instanceof ra&&e.alignedRect instanceof vt}function cpe(e){let t=(o,l,u,p)=>Math.atan2(p-l,u-o)%Math.PI,n=o=>o*180/Math.PI,a={roll:void 0,pitch:void 0,yaw:void 0};if(!e||!e._positions||e._positions.length!==68)return a;let r=e._positions;a.roll=-t(r[36]._x,r[36]._y,r[45]._x,r[45]._y),a.pitch=t(0,Math.abs(r[0]._x-r[30]._x)/r[30]._x,Math.PI,Math.abs(r[16]._x-r[30]._x)/r[30]._x);let s=r.reduce((o,l)=>o<l._y?o:l._y,1/0),i=r.reduce((o,l)=>o>l._y?o:l._y,-1/0);return a.yaw=Math.PI*(e._imgDims._height/(i-s)/1.4-1),a}function yp(e,t){let{box:n}=e.detection,a=t.shiftBy(n.x,n.y),r=a.align(),{imageDims:s}=e.detection,i=new vt(e.detection.score,r.rescale(s.reverse()),s),o=cpe(t);return{...e,...{landmarks:a,unshiftedLandmarks:t,alignedRect:i,angle:o}}}var Ng=class{constructor(t={}){let{drawLines:n=!0,drawPoints:a=!0,lineWidth:r,lineColor:s,pointSize:i,pointColor:o}=t;this.drawLines=n,this.drawPoints=a,this.lineWidth=r||1,this.pointSize=i||2,this.lineColor=s||"rgba(0, 255, 255, 1)",this.pointColor=o||"rgba(255, 0, 255, 1)"}},Tg=class{constructor(t,n={}){this.faceLandmarks=t,this.options=new Ng(n)}draw(t){let n=jn(t),{drawLines:a,drawPoints:r,lineWidth:s,lineColor:i,pointSize:o,pointColor:l}=this.options;if(a&&this.faceLandmarks instanceof $o&&(n.strokeStyle=i,n.lineWidth=s,Or(n,this.faceLandmarks.getJawOutline()),Or(n,this.faceLandmarks.getLeftEyeBrow()),Or(n,this.faceLandmarks.getRightEyeBrow()),Or(n,this.faceLandmarks.getNose()),Or(n,this.faceLandmarks.getLeftEye(),!0),Or(n,this.faceLandmarks.getRightEye(),!0),Or(n,this.faceLandmarks.getMouth(),!0)),r){n.strokeStyle=l,n.fillStyle=l;let u=p=>{n.beginPath(),n.arc(p.x,p.y,o,0,2*Math.PI),n.fill()};this.faceLandmarks.positions.forEach(u)}}};function dpe(e,t){(Array.isArray(t)?t:[t]).forEach(a=>{let r=a instanceof ra?a:Lo(a)?a.landmarks:void 0;if(!r)throw new Error("drawFaceLandmarks - expected faceExpressions to be FaceLandmarks | WithFaceLandmarks<WithFaceDetection<{}>> or array thereof");new Tg(r).draw(e)})}var fF="1.6.9";function fpe(e,t){let n=dp(e,t),a=hp(e,t);function r(i,o,l){let u=a(i,o,`${l}/separable_conv0`),p=a(o,o,`${l}/separable_conv1`),d=n(i,o,1,`${l}/expansion_conv`);return{separable_conv0:u,separable_conv1:p,expansion_conv:d}}function s(i,o){let l=a(i,i,`${o}/separable_conv0`),u=a(i,i,`${o}/separable_conv1`),p=a(i,i,`${o}/separable_conv2`);return{separable_conv0:l,separable_conv1:u,separable_conv2:p}}return{extractConvParams:n,extractSeparableConvParams:a,extractReductionBlockParams:r,extractMainBlockParams:s}}function gF(e,t){let n=[],{extractWeights:a,getRemainingWeights:r}=$n(e),{extractConvParams:s,extractSeparableConvParams:i,extractReductionBlockParams:o,extractMainBlockParams:l}=fpe(a,n),u=s(3,32,3,"entry_flow/conv_in"),p=o(32,64,"entry_flow/reduction_block_0"),d=o(64,128,"entry_flow/reduction_block_1"),c={conv_in:u,reduction_block_0:p,reduction_block_1:d},h={};br(t,0,1).forEach(y=>{h[`main_block_${y}`]=l(128,`middle_flow/main_block_${y}`)});let m=o(128,256,"exit_flow/reduction_block"),f=i(256,512,"exit_flow/separable_conv"),g={reduction_block:m,separable_conv:f};if(r().length!==0)throw new Error(`weights remaing after extract: ${r().length}`);return{paramMappings:n,params:{entry_flow:c,middle_flow:h,exit_flow:g}}}function gpe(e,t){let n=sa(e,t),a=wg(n),r=mp(n);function s(o){let l=r(`${o}/separable_conv0`),u=r(`${o}/separable_conv1`),p=a(`${o}/expansion_conv`);return{separable_conv0:l,separable_conv1:u,expansion_conv:p}}function i(o){let l=r(`${o}/separable_conv0`),u=r(`${o}/separable_conv1`),p=r(`${o}/separable_conv2`);return{separable_conv0:l,separable_conv1:u,separable_conv2:p}}return{extractConvParams:a,extractSeparableConvParams:r,extractReductionBlockParams:s,extractMainBlockParams:i}}function yF(e,t){let n=[],{extractConvParams:a,extractSeparableConvParams:r,extractReductionBlockParams:s,extractMainBlockParams:i}=gpe(e,n),o=a("entry_flow/conv_in"),l=s("entry_flow/reduction_block_0"),u=s("entry_flow/reduction_block_1"),p={conv_in:o,reduction_block_0:l,reduction_block_1:u},d={};br(t,0,1).forEach(f=>{d[`main_block_${f}`]=i(`middle_flow/main_block_${f}`)});let c=s("exit_flow/reduction_block"),h=r("exit_flow/separable_conv"),m={reduction_block:c,separable_conv:h};return An(e,n),{params:{entry_flow:p,middle_flow:d,exit_flow:m},paramMappings:n}}function bF(e,t,n){return J(Rt(e,t.filters,n,"same"),t.bias)}function y1(e,t,n=!0){let a=n?Xe(e):e;return a=qn(a,t.separable_conv0,[1,1]),a=qn(Xe(a),t.separable_conv1,[1,1]),a=Pt(a,[3,3],[2,2],"same"),a=J(a,bF(e,t.expansion_conv,[2,2])),a}function ype(e,t){let n=qn(Xe(e),t.separable_conv0,[1,1]);return n=qn(Xe(n),t.separable_conv1,[1,1]),n=qn(Xe(n),t.separable_conv2,[1,1]),n=J(n,e),n}var _g=class extends ln{constructor(n){super("TinyXception");this._numMainBlocks=n}forwardInput(n){let{params:a}=this;if(!a)throw new Error("TinyXception - load model before inference");return O(()=>{let r=oe(n.toBatchTensor(112,!0),"float32"),i=tr(r,[122.782,117.001,104.298]).div(255),o=Xe(bF(i,a.entry_flow.conv_in,[2,2]));return o=y1(o,a.entry_flow.reduction_block_0,!1),o=y1(o,a.entry_flow.reduction_block_1),br(this._numMainBlocks,0,1).forEach(l=>{o=ype(o,a.middle_flow[`main_block_${l}`])}),o=y1(o,a.exit_flow.reduction_block),o=Xe(qn(o,a.exit_flow.separable_conv,[1,1])),o})}async forward(n){return this.forwardInput(await wt(n))}getDefaultModelName(){return"tiny_xception_model"}extractParamsFromWeightMap(n){return yF(n,this._numMainBlocks)}extractParams(n){return gF(n,this._numMainBlocks)}};function xF(e){let t=[],{extractWeights:n,getRemainingWeights:a}=$n(e),r=xg(n,t),s=r(512,1,"fc/age"),i=r(512,2,"fc/gender");if(a().length!==0)throw new Error(`weights remaing after extract: ${a().length}`);return{paramMappings:t,params:{fc:{age:s,gender:i}}}}function vF(e){let t=[],n=sa(e,t);function a(s){let i=n(`${s}/weights`,2),o=n(`${s}/bias`,1);return{weights:i,bias:o}}let r={fc:{age:a("fc/age"),gender:a("fc/gender")}};return An(e,t),{params:r,paramMappings:t}}var Cg=(n=>(n.FEMALE="female",n.MALE="male",n))(Cg||{});var Nd=class extends ln{constructor(n=new _g(2)){super("AgeGenderNet");this._faceFeatureExtractor=n}get faceFeatureExtractor(){return this._faceFeatureExtractor}runNet(n){let{params:a}=this;if(!a)throw new Error(`${this._name} - load model before inference`);return O(()=>{let r=n instanceof wr?this.faceFeatureExtractor.forwardInput(n):n,s=ga(r,[7,7],[2,2],"valid").as2D(r.shape[0],-1),i=Id(s,a.fc.age).as1D(),o=Id(s,a.fc.gender);return{age:i,gender:o}})}forwardInput(n){return O(()=>{let{age:a,gender:r}=this.runNet(n);return{age:a,gender:Qa(r)}})}async forward(n){return this.forwardInput(await wt(n))}async predictAgeAndGender(n){let a=await wt(n),r=await this.forwardInput(a),s=ht(r.age),i=ht(r.gender),o=s.map((u,p)=>({ageTensor:u,genderTensor:i[p]})),l=await Promise.all(o.map(async({ageTensor:u,genderTensor:p})=>{let d=u.dataSync()[0],c=p.dataSync()[0],h=c>.5,m=h?"male":"female",f=h?c:1-c;return u.dispose(),p.dispose(),{age:d,gender:m,genderProbability:f}}));return r.age.dispose(),r.gender.dispose(),a.isBatchInput?l:l[0]}getDefaultModelName(){return"age_gender_model"}dispose(n=!0){this.faceFeatureExtractor.dispose(n),super.dispose(n)}loadClassifierParams(n){let{params:a,paramMappings:r}=this.extractClassifierParams(n);this._params=a,this._paramMappings=r}extractClassifierParams(n){return xF(n)}extractParamsFromWeightMap(n){let{featureExtractorMap:a,classifierMap:r}=Ig(n);return this.faceFeatureExtractor.loadFromWeightMap(a),vF(r)}extractParams(n){let r=n.slice(0,n.length-1539),s=n.slice(n.length-1539);return this.faceFeatureExtractor.extractWeights(r),this.extractClassifierParams(s)}};var bp=class extends gp{postProcess(t,n,a){let r=a.map(({width:i,height:o})=>{let l=n/Math.max(o,i);return{width:i*l,height:o*l}}),s=r.length;return O(()=>{let i=(d,c)=>Mt([Cn([68],d,"float32"),Cn([68],c,"float32")],1).as2D(1,136).as1D(),o=(d,c)=>{let{width:h,height:m}=r[d];return c(h,m)?Math.abs(h-m)/2:0},l=d=>o(d,(c,h)=>c<h),u=d=>o(d,(c,h)=>h<c);return t.mul(Cn([s,136],n,"float32")).sub(Mt(Array.from(Array(s),(d,c)=>i(l(c),u(c))))).div(Mt(Array.from(Array(s),(d,c)=>i(r[c].width,r[c].height))))})}forwardInput(t){return O(()=>{let n=this.runNet(t);return this.postProcess(n,t.inputSize,t.inputDimensions.map(([a,r])=>({height:a,width:r})))})}async forward(t){return this.forwardInput(await wt(t))}async detectLandmarks(t){let n=await wt(t),a=O(()=>ht(this.forwardInput(n))),r=await Promise.all(a.map(async(s,i)=>{let o=Array.from(s.dataSync()),l=o.filter((p,d)=>dg(d)),u=o.filter((p,d)=>!dg(d));return new $o(Array(68).fill(0).map((p,d)=>new Pe(l[d],u[d])),{height:n.getInputHeight(i),width:n.getInputWidth(i)})}));return a.forEach(s=>s.dispose()),n.isBatchInput?r:r[0]}getClassifierChannelsOut(){return 136}};var zo=class extends bp{constructor(t=new fp){super("FaceLandmark68Net",t)}getDefaultModelName(){return"face_landmark_68_model"}getClassifierChannelsIn(){return 256}};function wF(e){let t=[],{extractDenseBlock3Params:n}=kg(e,t),a={dense0:n("dense0",!0),dense1:n("dense1"),dense2:n("dense2")};return An(e,t),{params:a,paramMappings:t}}function kF(e){let t=[],{extractWeights:n,getRemainingWeights:a}=$n(e),{extractDenseBlock3Params:r}=vg(n,t),s=r(3,32,"dense0",!0),i=r(32,64,"dense1"),o=r(64,128,"dense2");if(a().length!==0)throw new Error(`weights remaing after extract: ${a().length}`);return{paramMappings:t,params:{dense0:s,dense1:i,dense2:o}}}var Eg=class extends ln{constructor(){super("TinyFaceFeatureExtractor")}forwardInput(t){let{params:n}=this;if(!n)throw new Error("TinyFaceFeatureExtractor - load model before inference");return O(()=>{let a=oe(t.toBatchTensor(112,!0),"float32"),s=tr(a,[122.782,117.001,104.298]).div(255),i=bg(s,n.dense0,!0);return i=bg(i,n.dense1),i=bg(i,n.dense2),i=ga(i,[14,14],[2,2],"valid"),i})}async forward(t){return this.forwardInput(await wt(t))}getDefaultModelName(){return"face_feature_extractor_tiny_model"}extractParamsFromWeightMap(t){return wF(t)}extractParams(t){return kF(t)}};var Td=class extends bp{constructor(t=new Eg){super("FaceLandmark68TinyNet",t)}getDefaultModelName(){return"face_landmark_68_tiny_model"}getClassifierChannelsIn(){return 128}};var b1=class extends zo{};function IF(e,t){return J(B(e,t.weights),t.biases)}function x1(e,t,n,a,r="same"){let{filters:s,bias:i}=t.conv,o=Rt(e,s,n,r);return o=J(o,i),o=IF(o,t.scale),a?Xe(o):o}function SF(e,t){return x1(e,t,[1,1],!0)}function v1(e,t){return x1(e,t,[1,1],!1)}function Fg(e,t){return x1(e,t,[2,2],!0,"valid")}function bpe(e,t){function n(o,l,u){let p=e(o),d=p.length/(l*u*u);if(q0(d))throw new Error(`depth has to be an integer: ${d}, weights.length: ${p.length}, numFilters: ${l}, filterSize: ${u}`);return O(()=>Ae(Ja(p,[l,d,u,u]),[2,3,1,0]))}function a(o,l,u,p){let d=n(o,l,u),c=qe(e(l));return t.push({paramPath:`${p}/filters`},{paramPath:`${p}/bias`}),{filters:d,bias:c}}function r(o,l){let u=qe(e(o)),p=qe(e(o));return t.push({paramPath:`${l}/weights`},{paramPath:`${l}/biases`}),{weights:u,biases:p}}function s(o,l,u,p){let d=a(o,l,u,`${p}/conv`),c=r(l,`${p}/scale`);return{conv:d,scale:c}}function i(o,l,u,p,d=!1){let c=s((d?.5:1)*o,l,u,`${p}/conv1`),h=s(o,l,u,`${p}/conv2`);return{conv1:c,conv2:h}}return{extractConvLayerParams:s,extractResidualLayerParams:i}}function NF(e){let{extractWeights:t,getRemainingWeights:n}=$n(e),a=[],{extractConvLayerParams:r,extractResidualLayerParams:s}=bpe(t,a),i=r(4704,32,7,"conv32_down"),o=s(9216,32,3,"conv32_1"),l=s(9216,32,3,"conv32_2"),u=s(9216,32,3,"conv32_3"),p=s(36864,64,3,"conv64_down",!0),d=s(36864,64,3,"conv64_1"),c=s(36864,64,3,"conv64_2"),h=s(36864,64,3,"conv64_3"),m=s(147456,128,3,"conv128_down",!0),f=s(147456,128,3,"conv128_1"),g=s(147456,128,3,"conv128_2"),y=s(589824,256,3,"conv256_down",!0),b=s(589824,256,3,"conv256_1"),x=s(589824,256,3,"conv256_2"),v=s(589824,256,3,"conv256_down_out"),k=O(()=>Ae(Ha(t(256*128),[128,256]),[1,0]));if(a.push({paramPath:"fc"}),n().length!==0)throw new Error(`weights remaing after extract: ${n().length}`);return{params:{conv32_down:i,conv32_1:o,conv32_2:l,conv32_3:u,conv64_down:p,conv64_1:d,conv64_2:c,conv64_3:h,conv128_down:m,conv128_1:f,conv128_2:g,conv256_down:y,conv256_1:b,conv256_2:x,conv256_down_out:v,fc:k},paramMappings:a}}function xpe(e,t){let n=sa(e,t);function a(i){let o=n(`${i}/scale/weights`,1),l=n(`${i}/scale/biases`,1);return{weights:o,biases:l}}function r(i){let o=n(`${i}/conv/filters`,4),l=n(`${i}/conv/bias`,1),u=a(i);return{conv:{filters:o,bias:l},scale:u}}function s(i){return{conv1:r(`${i}/conv1`),conv2:r(`${i}/conv2`)}}return{extractConvLayerParams:r,extractResidualLayerParams:s}}function TF(e){let t=[],{extractConvLayerParams:n,extractResidualLayerParams:a}=xpe(e,t),r=n("conv32_down"),s=a("conv32_1"),i=a("conv32_2"),o=a("conv32_3"),l=a("conv64_down"),u=a("conv64_1"),p=a("conv64_2"),d=a("conv64_3"),c=a("conv128_down"),h=a("conv128_1"),m=a("conv128_2"),f=a("conv256_down"),g=a("conv256_1"),y=a("conv256_2"),b=a("conv256_down_out"),{fc:x}=e;if(t.push({originalPath:"fc",paramPath:"fc"}),!j0(x))throw new Error(`expected weightMap[fc] to be a Tensor2D, instead have ${x}`);let v={conv32_down:r,conv32_1:s,conv32_2:i,conv32_3:o,conv64_down:l,conv64_1:u,conv64_2:p,conv64_3:d,conv128_down:c,conv128_1:h,conv128_2:m,conv256_down:f,conv256_1:g,conv256_2:y,conv256_down_out:b,fc:x};return An(e,t),{params:v,paramMappings:t}}function nr(e,t){let n=SF(e,t.conv1);return n=v1(n,t.conv2),n=J(n,e),n=Xe(n),n}function _d(e,t){let n=Fg(e,t.conv1);n=v1(n,t.conv2);let a=ga(e,2,2,"valid"),r=St(a.shape),s=a.shape[3]!==n.shape[3];if(a.shape[1]!==n.shape[1]||a.shape[2]!==n.shape[2]){let o=[...n.shape];o[1]=1;let l=St(o);n=Ze([n,l],1);let u=[...n.shape];u[2]=1;let p=St(u);n=Ze([n,p],2)}return a=s?Ze([a,r],3):a,n=J(a,n),n=Xe(n),n}var Bo=class extends ln{constructor(){super("FaceRecognitionNet")}forwardInput(t){let{params:n}=this;if(!n)throw new Error("FaceRecognitionNet - load model before inference");return O(()=>{let a=oe(t.toBatchTensor(150,!0),"float32"),s=tr(a,[122.782,117.001,104.298]).div(255),i=Fg(s,n.conv32_down);i=Pt(i,3,2,"valid"),i=nr(i,n.conv32_1),i=nr(i,n.conv32_2),i=nr(i,n.conv32_3),i=_d(i,n.conv64_down),i=nr(i,n.conv64_1),i=nr(i,n.conv64_2),i=nr(i,n.conv64_3),i=_d(i,n.conv128_down),i=nr(i,n.conv128_1),i=nr(i,n.conv128_2),i=_d(i,n.conv256_down),i=nr(i,n.conv256_1),i=nr(i,n.conv256_2),i=_d(i,n.conv256_down_out);let o=i.mean([1,2]);return De(o,n.fc)})}async forward(t){return this.forwardInput(await wt(t))}async computeFaceDescriptor(t){var s;if((s=t==null?void 0:t.shape)!=null&&s.some(i=>i<=0))return new Float32Array(128);let n=await wt(t),a=O(()=>ht(this.forwardInput(n))),r=await Promise.all(a.map(i=>i.data()));return a.forEach(i=>i.dispose()),n.isBatchInput?r:r[0]}getDefaultModelName(){return"face_recognition_model"}extractParamsFromWeightMap(t){return TF(t)}extractParams(t){return NF(t)}};function vpe(e){let t=new Bo;return t.extractWeights(e),t}function Ag(e,t){return{...e,...{descriptor:t}}}function wpe(e){return typeof e.age=="number"}function $g(e,t){return{...e,...{age:t}}}function kpe(e){return(e.gender==="male"||e.gender==="female")&&ip(e.genderProbability)}function Dg(e,t,n){return{...e,...{gender:t,genderProbability:n}}}function Ipe(e,t){function n(l,u){let p=Ja(e(9*l),[3,3,l,1]),d=qe(e(l)),c=qe(e(l)),h=qe(e(l)),m=qe(e(l));return t.push({paramPath:`${u}/filters`},{paramPath:`${u}/batch_norm_scale`},{paramPath:`${u}/batch_norm_offset`},{paramPath:`${u}/batch_norm_mean`},{paramPath:`${u}/batch_norm_variance`}),{filters:p,batch_norm_scale:d,batch_norm_offset:c,batch_norm_mean:h,batch_norm_variance:m}}function a(l,u,p,d,c){let h=Ja(e(l*u*p*p),[p,p,l,u]),m=qe(e(u));return t.push({paramPath:`${d}/filters`},{paramPath:`${d}/${c?"batch_norm_offset":"bias"}`}),{filters:h,bias:m}}function r(l,u,p,d){let{filters:c,bias:h}=a(l,u,p,d,!0);return{filters:c,batch_norm_offset:h}}function s(l,u,p){let d=n(l,`${p}/depthwise_conv`),c=r(l,u,1,`${p}/pointwise_conv`);return{depthwise_conv:d,pointwise_conv:c}}function i(){let l=r(3,32,3,"mobilenetv1/conv_0"),u=s(32,64,"mobilenetv1/conv_1"),p=s(64,128,"mobilenetv1/conv_2"),d=s(128,128,"mobilenetv1/conv_3"),c=s(128,256,"mobilenetv1/conv_4"),h=s(256,256,"mobilenetv1/conv_5"),m=s(256,512,"mobilenetv1/conv_6"),f=s(512,512,"mobilenetv1/conv_7"),g=s(512,512,"mobilenetv1/conv_8"),y=s(512,512,"mobilenetv1/conv_9"),b=s(512,512,"mobilenetv1/conv_10"),x=s(512,512,"mobilenetv1/conv_11"),v=s(512,1024,"mobilenetv1/conv_12"),k=s(1024,1024,"mobilenetv1/conv_13");return{conv_0:l,conv_1:u,conv_2:p,conv_3:d,conv_4:c,conv_5:h,conv_6:m,conv_7:f,conv_8:g,conv_9:y,conv_10:b,conv_11:x,conv_12:v,conv_13:k}}function o(){let l=r(1024,256,1,"prediction_layer/conv_0"),u=r(256,512,3,"prediction_layer/conv_1"),p=r(512,128,1,"prediction_layer/conv_2"),d=r(128,256,3,"prediction_layer/conv_3"),c=r(256,128,1,"prediction_layer/conv_4"),h=r(128,256,3,"prediction_layer/conv_5"),m=r(256,64,1,"prediction_layer/conv_6"),f=r(64,128,3,"prediction_layer/conv_7"),g=a(512,12,1,"prediction_layer/box_predictor_0/box_encoding_predictor"),y=a(512,9,1,"prediction_layer/box_predictor_0/class_predictor"),b=a(1024,24,1,"prediction_layer/box_predictor_1/box_encoding_predictor"),x=a(1024,18,1,"prediction_layer/box_predictor_1/class_predictor"),v=a(512,24,1,"prediction_layer/box_predictor_2/box_encoding_predictor"),k=a(512,18,1,"prediction_layer/box_predictor_2/class_predictor"),T=a(256,24,1,"prediction_layer/box_predictor_3/box_encoding_predictor"),_=a(256,18,1,"prediction_layer/box_predictor_3/class_predictor"),E=a(256,24,1,"prediction_layer/box_predictor_4/box_encoding_predictor"),A=a(256,18,1,"prediction_layer/box_predictor_4/class_predictor"),M=a(128,24,1,"prediction_layer/box_predictor_5/box_encoding_predictor"),$=a(128,18,1,"prediction_layer/box_predictor_5/class_predictor");return{conv_0:l,conv_1:u,conv_2:p,conv_3:d,conv_4:c,conv_5:h,conv_6:m,conv_7:f,box_predictor_0:{box_encoding_predictor:g,class_predictor:y},box_predictor_1:{box_encoding_predictor:b,class_predictor:x},box_predictor_2:{box_encoding_predictor:v,class_predictor:k},box_predictor_3:{box_encoding_predictor:T,class_predictor:_},box_predictor_4:{box_encoding_predictor:E,class_predictor:A},box_predictor_5:{box_encoding_predictor:M,class_predictor:$}}}return{extractMobilenetV1Params:i,extractPredictionLayerParams:o}}function _F(e){let t=[],{extractWeights:n,getRemainingWeights:a}=$n(e),{extractMobilenetV1Params:r,extractPredictionLayerParams:s}=Ipe(n,t),i=r(),o=s(),u={extra_dim:af(n(5118*4),[1,5118,4])};if(t.push({paramPath:"output_layer/extra_dim"}),a().length!==0)throw new Error(`weights remaing after extract: ${a().length}`);return{params:{mobilenetv1:i,prediction_layer:o,output_layer:u},paramMappings:t}}function Spe(e,t){let n=sa(e,t);function a(u,p,d){let c=n(`${u}/Conv2d_${p}_pointwise/weights`,4,`${d}/filters`),h=n(`${u}/Conv2d_${p}_pointwise/convolution_bn_offset`,1,`${d}/batch_norm_offset`);return{filters:c,batch_norm_offset:h}}function r(u){let p=`mobilenetv1/conv_${u}`,d=`MobilenetV1/Conv2d_${u}_depthwise`,c=`${p}/depthwise_conv`,h=`${p}/pointwise_conv`,m=n(`${d}/depthwise_weights`,4,`${c}/filters`),f=n(`${d}/BatchNorm/gamma`,1,`${c}/batch_norm_scale`),g=n(`${d}/BatchNorm/beta`,1,`${c}/batch_norm_offset`),y=n(`${d}/BatchNorm/moving_mean`,1,`${c}/batch_norm_mean`),b=n(`${d}/BatchNorm/moving_variance`,1,`${c}/batch_norm_variance`);return{depthwise_conv:{filters:m,batch_norm_scale:f,batch_norm_offset:g,batch_norm_mean:y,batch_norm_variance:b},pointwise_conv:a("MobilenetV1",u,h)}}function s(){return{conv_0:a("MobilenetV1",0,"mobilenetv1/conv_0"),conv_1:r(1),conv_2:r(2),conv_3:r(3),conv_4:r(4),conv_5:r(5),conv_6:r(6),conv_7:r(7),conv_8:r(8),conv_9:r(9),conv_10:r(10),conv_11:r(11),conv_12:r(12),conv_13:r(13)}}function i(u,p){let d=n(`${u}/weights`,4,`${p}/filters`),c=n(`${u}/biases`,1,`${p}/bias`);return{filters:d,bias:c}}function o(u){let p=i(`Prediction/BoxPredictor_${u}/BoxEncodingPredictor`,`prediction_layer/box_predictor_${u}/box_encoding_predictor`),d=i(`Prediction/BoxPredictor_${u}/ClassPredictor`,`prediction_layer/box_predictor_${u}/class_predictor`);return{box_encoding_predictor:p,class_predictor:d}}function l(){return{conv_0:a("Prediction",0,"prediction_layer/conv_0"),conv_1:a("Prediction",1,"prediction_layer/conv_1"),conv_2:a("Prediction",2,"prediction_layer/conv_2"),conv_3:a("Prediction",3,"prediction_layer/conv_3"),conv_4:a("Prediction",4,"prediction_layer/conv_4"),conv_5:a("Prediction",5,"prediction_layer/conv_5"),conv_6:a("Prediction",6,"prediction_layer/conv_6"),conv_7:a("Prediction",7,"prediction_layer/conv_7"),box_predictor_0:o(0),box_predictor_1:o(1),box_predictor_2:o(2),box_predictor_3:o(3),box_predictor_4:o(4),box_predictor_5:o(5)}}return{extractMobilenetV1Params:s,extractPredictionLayerParams:l}}function CF(e){let t=[],{extractMobilenetV1Params:n,extractPredictionLayerParams:a}=Spe(e,t),r=e["Output/extra_dim"];if(t.push({originalPath:"Output/extra_dim",paramPath:"output_layer/extra_dim"}),!Lr(r))throw new Error(`expected weightMap['Output/extra_dim'] to be a Tensor3D, instead have ${r}`);let s={mobilenetv1:n(),prediction_layer:a(),output_layer:{extra_dim:r}};return An(e,t),{params:s,paramMappings:t}}function $a(e,t,n){return O(()=>{let a=Rt(e,t.filters,n,"same");return a=J(a,t.batch_norm_offset),an(a,0,6)})}var Npe=.0010000000474974513;function Tpe(e,t,n){return O(()=>{let a=Ss(e,t.filters,n,"same");return a=$r(a,t.batch_norm_mean,t.batch_norm_variance,t.batch_norm_offset,t.batch_norm_scale,Npe),an(a,0,6)})}function _pe(e){return[2,4,6,12].some(t=>t===e)?[2,2]:[1,1]}function EF(e,t){return O(()=>{let n,a=$a(e,t.conv_0,[2,2]);if([t.conv_1,t.conv_2,t.conv_3,t.conv_4,t.conv_5,t.conv_6,t.conv_7,t.conv_8,t.conv_9,t.conv_10,t.conv_11,t.conv_12,t.conv_13].forEach((s,i)=>{let o=i+1,l=_pe(o);a=Tpe(a,s.depthwise_conv,l),a=$a(a,s.pointwise_conv,[1,1]),o===11&&(n=a)}),n===null)throw new Error("mobileNetV1 - output of conv layer 11 is null");return{out:a,conv11:n}})}function Cpe(e,t,n){let a=e.arraySync(),r=Math.min(a[t][0],a[t][2]),s=Math.min(a[t][1],a[t][3]),i=Math.max(a[t][0],a[t][2]),o=Math.max(a[t][1],a[t][3]),l=Math.min(a[n][0],a[n][2]),u=Math.min(a[n][1],a[n][3]),p=Math.max(a[n][0],a[n][2]),d=Math.max(a[n][1],a[n][3]),c=(i-r)*(o-s),h=(p-l)*(d-u);if(c<=0||h<=0)return 0;let m=Math.max(r,l),f=Math.max(s,u),g=Math.min(i,p),y=Math.min(o,d),b=Math.max(g-m,0)*Math.max(y-f,0);return b/(c+h-b)}function FF(e,t,n,a,r){let s=e.shape[0],i=Math.min(n,s),o=t.map((p,d)=>({score:p,boxIndex:d})).filter(p=>p.score>r).sort((p,d)=>d.score-p.score),l=p=>p<=a?1:0,u=[];return o.forEach(p=>{if(u.length>=i)return;let d=p.score;for(let c=u.length-1;c>=0;--c){let h=Cpe(e,p.boxIndex,u[c]);if(h!==0&&(p.score*=l(h),p.score<=r))break}d===p.score&&u.push(p.boxIndex)}),u}function Epe(e){let t=ht(Ae(e,[1,0])),n=[ce(t[2],t[0]),ce(t[3],t[1])],a=[J(t[0],fe(n[0],2)),J(t[1],fe(n[1],2))];return{sizes:n,centers:a}}function Fpe(e,t){let{sizes:n,centers:a}=Epe(e),r=ht(Ae(t,[1,0])),s=fe(B(gn(fe(r[2],5)),n[0]),2),i=J(B(fe(r[0],10),n[0]),a[0]),o=fe(B(gn(fe(r[3],5)),n[1]),2),l=J(B(fe(r[1],10),n[1]),a[1]);return Ae(Mt([ce(i,s),ce(l,o),J(i,s),J(l,o)]),[1,0])}function AF(e,t,n){return O(()=>{let a=e.shape[0],r=Fpe(W(On(n.extra_dim,[a,1,1]),[-1,4]),W(e,[-1,4]));r=W(r,[a,r.shape[0]/a,4]);let s=ma(He(t,[0,0,1],[-1,-1,-1])),i=He(s,[0,0,0],[-1,-1,1]);i=W(i,[a,i.shape[1]]);let o=ht(r),l=ht(i);return{boxes:o,scores:l}})}function Wo(e,t){return O(()=>{let n=e.shape[0],a=W(Oo(e,t.box_encoding_predictor),[n,-1,1,4]),r=W(Oo(e,t.class_predictor),[n,-1,3]);return{boxPredictionEncoding:a,classPrediction:r}})}function $F(e,t,n){return O(()=>{let a=$a(e,n.conv_0,[1,1]),r=$a(a,n.conv_1,[2,2]),s=$a(r,n.conv_2,[1,1]),i=$a(s,n.conv_3,[2,2]),o=$a(i,n.conv_4,[1,1]),l=$a(o,n.conv_5,[2,2]),u=$a(l,n.conv_6,[1,1]),p=$a(u,n.conv_7,[2,2]),d=Wo(t,n.box_predictor_0),c=Wo(e,n.box_predictor_1),h=Wo(r,n.box_predictor_2),m=Wo(i,n.box_predictor_3),f=Wo(l,n.box_predictor_4),g=Wo(p,n.box_predictor_5),y=Ze([d.boxPredictionEncoding,c.boxPredictionEncoding,h.boxPredictionEncoding,m.boxPredictionEncoding,f.boxPredictionEncoding,g.boxPredictionEncoding],1),b=Ze([d.classPrediction,c.classPrediction,h.classPrediction,m.classPrediction,f.classPrediction,g.classPrediction],1);return{boxPredictions:y,classPredictions:b}})}var va=class{constructor({minConfidence:t,maxResults:n}={}){this._name="SsdMobilenetv1Options";if(this._minConfidence=t||.5,this._maxResults=n||100,typeof this._minConfidence!="number"||this._minConfidence<=0||this._minConfidence>=1)throw new Error(`${this._name} - expected minConfidence to be a number between 0 and 1`);if(typeof this._maxResults!="number")throw new Error(`${this._name} - expected maxResults to be a number`)}get minConfidence(){return this._minConfidence}get maxResults(){return this._maxResults}};var Fs=class extends ln{constructor(){super("SsdMobilenetv1")}forwardInput(t){let{params:n}=this;if(!n)throw new Error("SsdMobilenetv1 - load model before inference");return O(()=>{let a=oe(t.toBatchTensor(512,!1),"float32"),r=ce(fe(a,127.5),1),s=EF(r,n.mobilenetv1),{boxPredictions:i,classPredictions:o}=$F(s.out,s.conv11,n.prediction_layer);return AF(i,o,n.output_layer)})}async forward(t){return this.forwardInput(await wt(t))}async locateFaces(t,n={}){let{maxResults:a,minConfidence:r}=new va(n),s=await wt(t),{boxes:i,scores:o}=this.forwardInput(s),l=i[0],u=o[0];for(let x=1;x<i.length;x++)i[x].dispose(),o[x].dispose();let p=Array.from(u.dataSync()),c=FF(l,p,a,.5,r),h=s.getReshapedInputDimensions(0),m=s.inputSize,f=m/h.width,g=m/h.height,y=l.arraySync(),b=c.map(x=>{let[v,k]=[Math.max(0,y[x][0]),Math.min(1,y[x][2])].map(E=>E*g),[T,_]=[Math.max(0,y[x][1]),Math.min(1,y[x][3])].map(E=>E*f);return new vt(p[x],new Ao(T,v,_-T,k-v),{height:s.getInputHeight(0),width:s.getInputWidth(0)})});return l.dispose(),u.dispose(),b}getDefaultModelName(){return"ssd_mobilenetv1_model"}extractParamsFromWeightMap(t){return CF(t)}extractParams(t){return _F(t)}};function DF(e){let t=new Fs;return t.extractWeights(e),t}function Ape(e){return DF(e)}var w1=class extends Fs{};var RF=.4,MF=[new Pe(.738768,.874946),new Pe(2.42204,2.65704),new Pe(4.30971,7.04493),new Pe(10.246,4.59428),new Pe(12.6868,11.8741)],PF=[new Pe(1.603231,2.094468),new Pe(6.041143,7.080126),new Pe(2.882459,3.518061),new Pe(4.266906,5.178857),new Pe(9.041765,10.66308)],OF=[117.001,114.697,97.404],LF="tiny_yolov2_model",zF="tiny_yolov2_separable_conv_model";var Rg=e=>typeof e=="number";function k1(e){if(!e)throw new Error(`invalid config: ${e}`);if(typeof e.withSeparableConvs!="boolean")throw new Error(`config.withSeparableConvs has to be a boolean, have: ${e.withSeparableConvs}`);if(!Rg(e.iouThreshold)||e.iouThreshold<0||e.iouThreshold>1)throw new Error(`config.iouThreshold has to be a number between [0, 1], have: ${e.iouThreshold}`);if(!Array.isArray(e.classes)||!e.classes.length||!e.classes.every(t=>typeof t=="string"))throw new Error(`config.classes has to be an array class names: string[], have: ${JSON.stringify(e.classes)}`);if(!Array.isArray(e.anchors)||!e.anchors.length||!e.anchors.map(t=>t||{}).every(t=>Rg(t.x)&&Rg(t.y)))throw new Error(`config.anchors has to be an array of { x: number, y: number }, have: ${JSON.stringify(e.anchors)}`);if(e.meanRgb&&(!Array.isArray(e.meanRgb)||e.meanRgb.length!==3||!e.meanRgb.every(Rg)))throw new Error(`config.meanRgb has to be an array of shape [number, number, number], have: ${JSON.stringify(e.meanRgb)}`)}function xp(e){return O(()=>{let t=B(e,we(.10000000149011612));return J(Xe(ce(e,t)),t)})}function Ur(e,t){return O(()=>{let n=ya(e,[[0,0],[1,1],[1,1],[0,0]]);return n=Rt(n,t.conv.filters,[1,1],"valid"),n=ce(n,t.bn.sub),n=B(n,t.bn.truediv),n=J(n,t.conv.bias),xp(n)})}function Gr(e,t){return O(()=>{let n=ya(e,[[0,0],[1,1],[1,1],[0,0]]);return n=vo(n,t.depthwise_filter,t.pointwise_filter,[1,1],"valid"),n=J(n,t.bias),xp(n)})}function $pe(e,t){let n=dp(e,t);function a(i,o){let l=qe(e(i)),u=qe(e(i));return t.push({paramPath:`${o}/sub`},{paramPath:`${o}/truediv`}),{sub:l,truediv:u}}function r(i,o,l){let u=n(i,o,3,`${l}/conv`),p=a(o,`${l}/bn`);return{conv:u,bn:p}}let s=hp(e,t);return{extractConvParams:n,extractConvWithBatchNormParams:r,extractSeparableConvParams:s}}function BF(e,t,n,a){let{extractWeights:r,getRemainingWeights:s}=$n(e),i=[],{extractConvParams:o,extractConvWithBatchNormParams:l,extractSeparableConvParams:u}=$pe(r,i),p;if(t.withSeparableConvs){let[d,c,h,m,f,g,y,b,x]=a,v=t.isFirstLayerConv2d?o(d,c,3,"conv0"):u(d,c,"conv0"),k=u(c,h,"conv1"),T=u(h,m,"conv2"),_=u(m,f,"conv3"),E=u(f,g,"conv4"),A=u(g,y,"conv5"),M=b?u(y,b,"conv6"):void 0,$=x?u(b,x,"conv7"):void 0,S=o(x||b||y,5*n,1,"conv8");p={conv0:v,conv1:k,conv2:T,conv3:_,conv4:E,conv5:A,conv6:M,conv7:$,conv8:S}}else{let[d,c,h,m,f,g,y,b,x]=a,v=l(d,c,"conv0"),k=l(c,h,"conv1"),T=l(h,m,"conv2"),_=l(m,f,"conv3"),E=l(f,g,"conv4"),A=l(g,y,"conv5"),M=l(y,b,"conv6"),$=l(b,x,"conv7"),S=o(x,5*n,1,"conv8");p={conv0:v,conv1:k,conv2:T,conv3:_,conv4:E,conv5:A,conv6:M,conv7:$,conv8:S}}if(s().length!==0)throw new Error(`weights remaing after extract: ${s().length}`);return{params:p,paramMappings:i}}function Dpe(e,t){let n=sa(e,t);function a(o){let l=n(`${o}/sub`,1),u=n(`${o}/truediv`,1);return{sub:l,truediv:u}}function r(o){let l=n(`${o}/filters`,4),u=n(`${o}/bias`,1);return{filters:l,bias:u}}function s(o){let l=r(`${o}/conv`),u=a(`${o}/bn`);return{conv:l,bn:u}}let i=mp(n);return{extractConvParams:r,extractConvWithBatchNormParams:s,extractSeparableConvParams:i}}function WF(e,t){let n=[],{extractConvParams:a,extractConvWithBatchNormParams:r,extractSeparableConvParams:s}=Dpe(e,n),i;if(t.withSeparableConvs){let o=t.filterSizes&&t.filterSizes.length||9;i={conv0:t.isFirstLayerConv2d?a("conv0"):s("conv0"),conv1:s("conv1"),conv2:s("conv2"),conv3:s("conv3"),conv4:s("conv4"),conv5:s("conv5"),conv6:o>7?s("conv6"):void 0,conv7:o>8?s("conv7"):void 0,conv8:a("conv8")}}else i={conv0:r("conv0"),conv1:r("conv1"),conv2:r("conv2"),conv3:r("conv3"),conv4:r("conv4"),conv5:r("conv5"),conv6:r("conv6"),conv7:r("conv7"),conv8:a("conv8")};return An(e,n),{params:i,paramMappings:n}}var ar=class{constructor({inputSize:t,scoreThreshold:n}={}){this._name="TinyYolov2Options";if(this._inputSize=t||416,this._scoreThreshold=n||.5,typeof this._inputSize!="number"||this._inputSize%32!==0)throw new Error(`${this._name} - expected inputSize to be a number divisible by 32`);if(typeof this._scoreThreshold!="number"||this._scoreThreshold<=0||this._scoreThreshold>=1)throw new Error(`${this._name} - expected scoreThreshold to be a number between 0 and 1`)}get inputSize(){return this._inputSize}get scoreThreshold(){return this._scoreThreshold}};var I1=class extends ln{constructor(n){super("TinyYolov2");k1(n),this._config=n}get config(){return this._config}get withClassScores(){return this.config.withClassScores||this.config.classes.length>1}get boxEncodingSize(){return 5+(this.withClassScores?this.config.classes.length:0)}runTinyYolov2(n,a){let r=Ur(n,a.conv0);return r=Pt(r,[2,2],[2,2],"same"),r=Ur(r,a.conv1),r=Pt(r,[2,2],[2,2],"same"),r=Ur(r,a.conv2),r=Pt(r,[2,2],[2,2],"same"),r=Ur(r,a.conv3),r=Pt(r,[2,2],[2,2],"same"),r=Ur(r,a.conv4),r=Pt(r,[2,2],[2,2],"same"),r=Ur(r,a.conv5),r=Pt(r,[2,2],[1,1],"same"),r=Ur(r,a.conv6),r=Ur(r,a.conv7),Oo(r,a.conv8,"valid",!1)}runMobilenet(n,a){let r=this.config.isFirstLayerConv2d?xp(Oo(n,a.conv0,"valid",!1)):Gr(n,a.conv0);return r=Pt(r,[2,2],[2,2],"same"),r=Gr(r,a.conv1),r=Pt(r,[2,2],[2,2],"same"),r=Gr(r,a.conv2),r=Pt(r,[2,2],[2,2],"same"),r=Gr(r,a.conv3),r=Pt(r,[2,2],[2,2],"same"),r=Gr(r,a.conv4),r=Pt(r,[2,2],[2,2],"same"),r=Gr(r,a.conv5),r=Pt(r,[2,2],[1,1],"same"),r=a.conv6?Gr(r,a.conv6):r,r=a.conv7?Gr(r,a.conv7):r,Oo(r,a.conv8,"valid",!1)}forwardInput(n,a){let{params:r}=this;if(!r)throw new Error("TinyYolov2 - load model before inference");return O(()=>{let s=oe(n.toBatchTensor(a,!1),"float32");return s=this.config.meanRgb?tr(s,this.config.meanRgb):s,s=s.div(255),this.config.withSeparableConvs?this.runMobilenet(s,r):this.runTinyYolov2(s,r)})}async forward(n,a){return this.forwardInput(await wt(n),a)}async detect(n,a={}){let{inputSize:r,scoreThreshold:s}=new ar(a),i=await wt(n),o=await this.forwardInput(i,r),l=O(()=>ht(o)[0].expandDims()),u={width:i.getInputWidth(0),height:i.getInputHeight(0)},p=await this.extractBoxes(l,i.getReshapedInputDimensions(0),s);o.dispose(),l.dispose();let d=p.map(y=>y.box),c=p.map(y=>y.score),h=p.map(y=>y.classScore),m=p.map(y=>this.config.classes[y.label]);return J0(d.map(y=>y.rescale(r)),c,this.config.iouThreshold,!0).map(y=>new zr(c[y],h[y],m[y],d[y],u))}getDefaultModelName(){return""}extractParamsFromWeightMap(n){return WF(n,this.config)}extractParams(n){let a=this.config.filterSizes||I1.DEFAULT_FILTER_SIZES,r=a?a.length:void 0;if(r!==7&&r!==8&&r!==9)throw new Error(`TinyYolov2 - expected 7 | 8 | 9 convolutional filters, but found ${r} filterSizes in config`);return BF(n,this.config,this.boxEncodingSize,a)}async extractBoxes(n,a,r){let{width:s,height:i}=a,o=Math.max(s,i),l=o/s,u=o/i,p=n.shape[1],d=this.config.anchors.length,[c,h,m]=O(()=>{let b=n.reshape([p,p,d,this.boxEncodingSize]),x=b.slice([0,0,0,0],[p,p,d,4]),v=b.slice([0,0,0,4],[p,p,d,1]),k=this.withClassScores?Qa(b.slice([0,0,0,5],[p,p,d,this.config.classes.length]),3):we(0);return[x,v,k]}),f=[],g=await h.array(),y=await c.array();for(let b=0;b<p;b++)for(let x=0;x<p;x++)for(let v=0;v<d;v++){let k=gd(g[b][x][v][0]);if(!r||k>r){let T=(x+gd(y[b][x][v][0]))/p*l,_=(b+gd(y[b][x][v][1]))/p*u,E=Math.exp(y[b][x][v][2])*this.config.anchors[v].x/p*l,A=Math.exp(y[b][x][v][3])*this.config.anchors[v].y/p*u,M=T-E/2,$=_-A/2,S={row:b,col:x,anchor:v},{classScore:P,label:V}=this.withClassScores?await this.extractPredictedClass(m,S):{classScore:1,label:0};f.push({box:new Fo(M,$,M+E,$+A),score:k,classScore:k*P,label:V,...S})}}return c.dispose(),h.dispose(),m.dispose(),f}async extractPredictedClass(n,a){let{row:r,col:s,anchor:i}=a,o=await n.array();return Array(this.config.classes.length).fill(0).map((l,u)=>o[r][s][i][u]).map((l,u)=>({classScore:l,label:u})).reduce((l,u)=>l.classScore>u.classScore?l:u)}},Vo=I1;Vo.DEFAULT_FILTER_SIZES=[3,16,32,64,128,256,512,1024,1024];var Uo=class extends Vo{constructor(t=!0){let n={withSeparableConvs:t,iouThreshold:RF,classes:["face"],...t?{anchors:PF,meanRgb:OF}:{anchors:MF,withClassScores:!0}};super(n)}get withSeparableConvs(){return this.config.withSeparableConvs}get anchors(){return this.config.anchors}async locateFaces(t,n){return(await this.detect(t,n)).map(r=>new vt(r.score,r.relativeBox,{width:r.imageWidth,height:r.imageHeight}))}getDefaultModelName(){return this.withSeparableConvs?zF:LF}extractParamsFromWeightMap(t){return super.extractParamsFromWeightMap(t)}};function Rpe(e,t=!0){let n=new Uo(t);return n.extractWeights(e),n}var Cd=class extends ar{constructor(){super(...arguments);this._name="TinyFaceDetectorOptions"}};var wa=class{async then(t){return t(await this.run())}async run(){throw new Error("ComposableTask - run is not implemented")}};async function Go(e,t,n,a,r=({alignedRect:s})=>s){let s=e.map(l=>Lo(l)?r(l):l.detection),i=a||(t instanceof Fe?await cp(t,s):await pp(t,s)),o=await n(i);return i.forEach(l=>l instanceof Fe&&l.dispose()),o}async function vp(e,t,n,a,r){return Go([e],t,async s=>n(s[0]),a,r)}var VF=.4,UF=[new Pe(1.603231,2.094468),new Pe(6.041143,7.080126),new Pe(2.882459,3.518061),new Pe(4.266906,5.178857),new Pe(9.041765,10.66308)],GF=[117.001,114.697,97.404];var Ho=class extends Vo{constructor(){let t={withSeparableConvs:!0,iouThreshold:VF,classes:["face"],anchors:UF,meanRgb:GF,isFirstLayerConv2d:!0,filterSizes:[3,16,32,64,128,256,512]};super(t)}get anchors(){return this.config.anchors}async locateFaces(t,n){return(await this.detect(t,n)).map(r=>new vt(r.score,r.relativeBox,{width:r.imageWidth,height:r.imageHeight}))}getDefaultModelName(){return"tiny_face_detector_model"}extractParamsFromWeightMap(t){return super.extractParamsFromWeightMap(t)}};var tt={ssdMobilenetv1:new Fs,tinyFaceDetector:new Ho,tinyYolov2:new Uo,faceLandmark68Net:new zo,faceLandmark68TinyNet:new Td,faceRecognitionNet:new Bo,faceExpressionNet:new Sd,ageGenderNet:new Nd},HF=(e,t)=>tt.ssdMobilenetv1.locateFaces(e,t),Mpe=(e,t)=>tt.tinyFaceDetector.locateFaces(e,t),Ppe=(e,t)=>tt.tinyYolov2.locateFaces(e,t),jF=e=>tt.faceLandmark68Net.detectLandmarks(e),Ope=e=>tt.faceLandmark68TinyNet.detectLandmarks(e),Lpe=e=>tt.faceRecognitionNet.computeFaceDescriptor(e),zpe=e=>tt.faceExpressionNet.predictExpressions(e),Bpe=e=>tt.ageGenderNet.predictAgeAndGender(e),qF=e=>tt.ssdMobilenetv1.load(e),Wpe=e=>tt.tinyFaceDetector.load(e),Vpe=e=>tt.tinyYolov2.load(e),Upe=e=>tt.faceLandmark68Net.load(e),Gpe=e=>tt.faceLandmark68TinyNet.load(e),Hpe=e=>tt.faceRecognitionNet.load(e),jpe=e=>tt.faceExpressionNet.load(e),qpe=e=>tt.ageGenderNet.load(e),Kpe=qF,Xpe=HF,Ype=jF;var Mg=class extends wa{constructor(n,a,r){super();this.parentTask=n;this.input=a;this.extractedFaces=r}},jo=class extends Mg{async run(){let t=await this.parentTask,n=await Go(t,this.input,async a=>Promise.all(a.map(r=>tt.faceExpressionNet.predictExpressions(r))),this.extractedFaces);return t.map((a,r)=>Sg(a,n[r]))}withAgeAndGender(){return new Ko(this,this.input)}},qo=class extends Mg{async run(){let t=await this.parentTask;if(!t)return;let n=await vp(t,this.input,a=>tt.faceExpressionNet.predictExpressions(a),this.extractedFaces);return Sg(t,n)}withAgeAndGender(){return new Xo(this,this.input)}},As=class extends jo{withAgeAndGender(){return new Ds(this,this.input)}withFaceDescriptors(){return new Hr(this,this.input)}},$s=class extends qo{withAgeAndGender(){return new Rs(this,this.input)}withFaceDescriptor(){return new jr(this,this.input)}};var Pg=class extends wa{constructor(n,a,r){super();this.parentTask=n;this.input=a;this.extractedFaces=r}},Ko=class extends Pg{async run(){let t=await this.parentTask,n=await Go(t,this.input,async a=>Promise.all(a.map(r=>tt.ageGenderNet.predictAgeAndGender(r))),this.extractedFaces);return t.map((a,r)=>{let{age:s,gender:i,genderProbability:o}=n[r];return $g(Dg(a,i,o),s)})}withFaceExpressions(){return new jo(this,this.input)}},Xo=class extends Pg{async run(){let t=await this.parentTask;if(!t)return;let{age:n,gender:a,genderProbability:r}=await vp(t,this.input,s=>tt.ageGenderNet.predictAgeAndGender(s),this.extractedFaces);return $g(Dg(t,a,r),n)}withFaceExpressions(){return new qo(this,this.input)}},Ds=class extends Ko{withFaceExpressions(){return new As(this,this.input)}withFaceDescriptors(){return new Hr(this,this.input)}},Rs=class extends Xo{withFaceExpressions(){return new $s(this,this.input)}withFaceDescriptor(){return new jr(this,this.input)}};var Ed=class extends wa{constructor(n,a){super();this.parentTask=n;this.input=a}},Hr=class extends Ed{async run(){let t=await this.parentTask;return(await Go(t,this.input,a=>Promise.all(a.map(r=>tt.faceRecognitionNet.computeFaceDescriptor(r))),null,a=>a.landmarks.align(null,{useDlibAlignment:!0}))).map((a,r)=>Ag(t[r],a))}withFaceExpressions(){return new As(this,this.input)}withAgeAndGender(){return new Ds(this,this.input)}},jr=class extends Ed{async run(){let t=await this.parentTask;if(!t)return;let n=await vp(t,this.input,a=>tt.faceRecognitionNet.computeFaceDescriptor(a),null,a=>a.landmarks.align(null,{useDlibAlignment:!0}));return Ag(t,n)}withFaceExpressions(){return new $s(this,this.input)}withAgeAndGender(){return new Rs(this,this.input)}};var Fd=class extends wa{constructor(n,a,r){super();this.parentTask=n;this.input=a;this.useTinyLandmarkNet=r}get landmarkNet(){return this.useTinyLandmarkNet?tt.faceLandmark68TinyNet:tt.faceLandmark68Net}},Ad=class extends Fd{async run(){let t=await this.parentTask,n=t.map(i=>i.detection),a=this.input instanceof Fe?await cp(this.input,n):await pp(this.input,n),r=await Promise.all(a.map(i=>this.landmarkNet.detectLandmarks(i)));return a.forEach(i=>i instanceof Fe&&i.dispose()),t.filter((i,o)=>r[o]).map((i,o)=>yp(i,r[o]))}withFaceExpressions(){return new As(this,this.input)}withAgeAndGender(){return new Ds(this,this.input)}withFaceDescriptors(){return new Hr(this,this.input)}},$d=class extends Fd{async run(){let t=await this.parentTask;if(!t)return;let{detection:n}=t,a=this.input instanceof Fe?await cp(this.input,[n]):await pp(this.input,[n]),r=await this.landmarkNet.detectLandmarks(a[0]);return a.forEach(s=>s instanceof Fe&&s.dispose()),yp(t,r)}withFaceExpressions(){return new $s(this,this.input)}withAgeAndGender(){return new Rs(this,this.input)}withFaceDescriptor(){return new jr(this,this.input)}};var Dd=class extends wa{constructor(n,a=new va){super();this.input=n;this.options=a}},wp=class extends Dd{async run(){let{input:t,options:n}=this,a;if(n instanceof Cd)a=tt.tinyFaceDetector.locateFaces(t,n);else if(n instanceof va)a=tt.ssdMobilenetv1.locateFaces(t,n);else if(n instanceof ar)a=tt.tinyYolov2.locateFaces(t,n);else throw new Error("detectFaces - expected options to be instance of TinyFaceDetectorOptions | SsdMobilenetv1Options | TinyYolov2Options");return a}runAndExtendWithFaceDetections(){return new Promise((t,n)=>{this.run().then(a=>t(a.map(r=>Do({},r)))).catch(a=>n(a))})}withFaceLandmarks(t=!1){return new Ad(this.runAndExtendWithFaceDetections(),this.input,t)}withFaceExpressions(){return new jo(this.runAndExtendWithFaceDetections(),this.input)}withAgeAndGender(){return new Ko(this.runAndExtendWithFaceDetections(),this.input)}},Rd=class extends Dd{async run(){let t=await new wp(this.input,this.options),n=t[0];return t.forEach(a=>{a.score>n.score&&(n=a)}),n}runAndExtendWithFaceDetection(){return new Promise(async t=>{let n=await this.run();t(n?Do({},n):void 0)})}withFaceLandmarks(t=!1){return new $d(this.runAndExtendWithFaceDetection(),this.input,t)}withFaceExpressions(){return new qo(this.runAndExtendWithFaceDetection(),this.input)}withAgeAndGender(){return new Xo(this.runAndExtendWithFaceDetection(),this.input)}};function Qpe(e,t=new va){return new Rd(e,t)}function Og(e,t=new va){return new wp(e,t)}async function KF(e,t){return Og(e,new va(t?{minConfidence:t}:{})).withFaceLandmarks().withFaceDescriptors()}async function Jpe(e,t={}){return Og(e,new ar(t)).withFaceLandmarks().withFaceDescriptors()}var Zpe=KF;function S1(e,t){if(e.length!==t.length)throw new Error("euclideanDistance: arr1.length !== arr2.length");let n=Array.from(e),a=Array.from(t);return Math.sqrt(n.map((r,s)=>r-a[s]).reduce((r,s)=>r+s**2,0))}var Md=class{constructor(t,n=.6){this._distanceThreshold=n;let a=Array.isArray(t)?t:[t];if(!a.length)throw new Error("FaceRecognizer.constructor - expected atleast one input");let r=1,s=()=>`person ${r++}`;this._labeledDescriptors=a.map(i=>{if(i instanceof xr)return i;if(i instanceof Float32Array)return new xr(s(),[i]);if(i.descriptor&&i.descriptor instanceof Float32Array)return new xr(s(),[i.descriptor]);throw new Error("FaceRecognizer.constructor - expected inputs to be of type LabeledFaceDescriptors | WithFaceDescriptor<any> | Float32Array | Array<LabeledFaceDescriptors | WithFaceDescriptor<any> | Float32Array>")})}get labeledDescriptors(){return this._labeledDescriptors}get distanceThreshold(){return this._distanceThreshold}computeMeanDistance(t,n){return n.map(a=>S1(a,t)).reduce((a,r)=>a+r,0)/(n.length||1)}matchDescriptor(t){return this.labeledDescriptors.map(({descriptors:n,label:a})=>new op(a,this.computeMeanDistance(t,n))).reduce((n,a)=>n.distance<a.distance?n:a)}findBestMatch(t){let n=this.matchDescriptor(t);return n.distance<this._distanceThreshold?n:new op("unknown",n.distance)}toJSON(){return{distanceThreshold:this._distanceThreshold,labeledDescriptors:this._labeledDescriptors.map(t=>t.toJSON())}}static fromJSON(t){let n=t.labeledDescriptors.map(a=>xr.fromJSON(a));return new Md(n,t.distanceThreshold)}};function ece(e){let t=new Ho;return t.extractWeights(e),t}function XF(e,t){let{width:n,height:a}=new yn(t.width,t.height);if(n<=0||a<=0)throw new Error(`resizeResults - invalid dimensions: ${JSON.stringify({width:n,height:a})}`);if(Array.isArray(e))return e.map(r=>XF(r,{width:n,height:a}));if(Lo(e)){let r=e.detection.forSize(n,a),s=e.unshiftedLandmarks.forSize(r.box.width,r.box.height);return yp(Do(e,r),s)}return vr(e)?Do(e,e.detection.forSize(n,a)):e instanceof ra||e instanceof vt?e.forSize(n,a):e}var tce=fF;return KA(nce);})();
/**
* @license
* Copyright 2017 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
* =============================================================================
*/
/**
* @license
* Copyright 2018 Google LLC
*
* Use of this source code is governed by an MIT-style
* license that can be found in the LICENSE file or at
* https://opensource.org/licenses/MIT.
* =============================================================================
*/
/**
* @license
* Copyright 2018 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*
* =============================================================================
*/
/**
* @license
* Copyright 2018 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
* =============================================================================
*/
/**
* @license
* Copyright 2019 Google LLC
*
* Use of this source code is governed by an MIT-style
* license that can be found in the LICENSE file or at
* https://opensource.org/licenses/MIT.
* =============================================================================
*/
/**
* @license
* Copyright 2019 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*
* =============================================================================
*/
/**
* @license
* Copyright 2019 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
* =============================================================================
*/
/**
* @license
* Copyright 2020 Google Inc. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
* =============================================================================
*/
/**
* @license
* Copyright 2020 Google LLC
*
* Use of this source code is governed by an MIT-style
* license that can be found in the LICENSE file or at
* https://opensource.org/licenses/MIT.
* =============================================================================
*/
/**
* @license
* Copyright 2020 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
* =============================================================================
*/
/**
* @license
* Copyright 2020 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the License);
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
* =============================================================================
*/
/**
* @license
* Copyright 2021 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
* =============================================================================
*/
/**
* @license
* Copyright 2021 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* https://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
* =============================================================================
*/
/**
* @license
* Copyright 2022 Google Inc. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
* =============================================================================
*/
/**
* @license
* Copyright 2022 Google LLC
*
* Use of this source code is governed by an MIT-style
* license that can be found in the LICENSE file or at
* https://opensource.org/licenses/MIT.
* =============================================================================
*/
/**
* @license
* Copyright 2022 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
* =============================================================================
*/
/**
* @license
* Copyright 2022 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the 'License');
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an 'AS IS' BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
* =============================================================================
*/
/**
* @license
* Copyright 2018 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
* =============================================================================
*/
/** @license See the LICENSE file. */
Reference in New Issue View Git Blame Copy Permalink