Classes
| AvgPool<T extends TNumber> | Performs average pooling on the input. |
| AvgPool.Options | Optional attributes for AvgPool
|
| AvgPool3d<T extends TNumber> | Performs 3D average pooling on the input. |
| AvgPool3d.Options | Optional attributes for AvgPool3d
|
| AvgPool3dGrad<T extends TNumber> | Computes gradients of average pooling function. |
| AvgPool3dGrad.Options | Optional attributes for AvgPool3dGrad
|
| AvgPoolGrad<T extends TNumber> | Computes gradients of the average pooling function. |
| AvgPoolGrad.Options | Optional attributes for AvgPoolGrad
|
| BatchNormWithGlobalNormalization<T extends TType> | Batch normalization. |
| BatchNormWithGlobalNormalizationGrad<T extends TType> | Gradients for batch normalization. |
| BiasAdd<T extends TType> | Adds `bias` to `value`. |
| BiasAdd.Options | Optional attributes for BiasAdd
|
| BiasAddGrad<T extends TType> | The backward operation for "BiasAdd" on the "bias" tensor. |
| BiasAddGrad.Options | Optional attributes for BiasAddGrad
|
| BlockLSTM<T extends TNumber> | Computes the LSTM cell forward propagation for all the time steps. |
| BlockLSTM.Options | Optional attributes for BlockLSTM
|
| BlockLSTMGrad<T extends TNumber> | Computes the LSTM cell backward propagation for the entire time sequence. |
| ComputeAccidentalHits | Computes the ids of the positions in sampled_candidates that match true_labels. |
| ComputeAccidentalHits.Options | Optional attributes for ComputeAccidentalHits
|
| Conv2d<T extends TNumber> | Computes a 2-D convolution given 4-D `input` and `filter` tensors. |
| Conv2d.Options | Optional attributes for Conv2d
|
| Conv2dBackpropFilter<T extends TNumber> | Computes the gradients of convolution with respect to the filter. |
| Conv2dBackpropFilter.Options | Optional attributes for Conv2dBackpropFilter
|
| Conv2dBackpropInput<T extends TNumber> | Computes the gradients of convolution with respect to the input. |
| Conv2dBackpropInput.Options | Optional attributes for Conv2dBackpropInput
|
| Conv3d<T extends TNumber> | Computes a 3-D convolution given 5-D `input` and `filter` tensors. |
| Conv3d.Options | Optional attributes for Conv3d
|
| Conv3dBackpropFilter<T extends TNumber> | Computes the gradients of 3-D convolution with respect to the filter. |
| Conv3dBackpropFilter.Options | Optional attributes for Conv3dBackpropFilter
|
| Conv3dBackpropInput<U extends TNumber> | Computes the gradients of 3-D convolution with respect to the input. |
| Conv3dBackpropInput.Options | Optional attributes for Conv3dBackpropInput
|
| CtcBeamSearchDecoder<T extends TNumber> | Performs beam search decoding on the logits given in input. |
| CtcBeamSearchDecoder.Options | Optional attributes for CtcBeamSearchDecoder
|
| CtcGreedyDecoder<T extends TNumber> | Performs greedy decoding on the logits given in inputs. |
| CtcGreedyDecoder.Options | Optional attributes for CtcGreedyDecoder
|
| CtcLoss<T extends TNumber> | Calculates the CTC Loss (log probability) for each batch entry. |
| CtcLoss.Options | Optional attributes for CtcLoss
|
| CTCLossV2 | Calculates the CTC Loss (log probability) for each batch entry. |
| CTCLossV2.Options | Optional attributes for CTCLossV2
|
| CudnnRNN<T extends TNumber> | A RNN backed by cuDNN. |
| CudnnRNN.Options | Optional attributes for CudnnRNN
|
| CudnnRNNBackprop<T extends TNumber> | Backprop step of CudnnRNNV3. |
| CudnnRNNBackprop.Options | Optional attributes for CudnnRNNBackprop
|
| CudnnRNNCanonicalToParams<T extends TNumber> | Converts CudnnRNN params from canonical form to usable form. |
| CudnnRNNCanonicalToParams.Options | Optional attributes for CudnnRNNCanonicalToParams
|
| CudnnRnnParamsSize<U extends TNumber> | Computes size of weights that can be used by a Cudnn RNN model. |
| CudnnRnnParamsSize.Options | Optional attributes for CudnnRnnParamsSize
|
| CudnnRNNParamsToCanonical<T extends TNumber> | Retrieves CudnnRNN params in canonical form. |
| CudnnRNNParamsToCanonical.Options | Optional attributes for CudnnRNNParamsToCanonical
|
| DataFormatDimMap<T extends TNumber> | Returns the dimension index in the destination data format given the one in
the source data format. |
| DataFormatDimMap.Options | Optional attributes for DataFormatDimMap
|
| DataFormatVecPermute<T extends TNumber> | Permute input tensor from `src_format` to `dst_format`. |
| DataFormatVecPermute.Options | Optional attributes for DataFormatVecPermute
|
| DepthToSpace<T extends TType> | DepthToSpace for tensors of type T. |
| DepthToSpace.Options | Optional attributes for DepthToSpace
|
| DepthwiseConv2dNative<T extends TNumber> | Computes a 2-D depthwise convolution given 4-D `input` and `filter` tensors. |
| DepthwiseConv2dNative.Options | Optional attributes for DepthwiseConv2dNative
|
| DepthwiseConv2dNativeBackpropFilter<T extends TNumber> | Computes the gradients of depthwise convolution with respect to the filter. |
| DepthwiseConv2dNativeBackpropFilter.Options | Optional attributes for DepthwiseConv2dNativeBackpropFilter
|
| DepthwiseConv2dNativeBackpropInput<T extends TNumber> | Computes the gradients of depthwise convolution with respect to the input. |
| DepthwiseConv2dNativeBackpropInput.Options | Optional attributes for DepthwiseConv2dNativeBackpropInput
|
| Dilation2d<T extends TNumber> | Computes the grayscale dilation of 4-D `input` and 3-D `filter` tensors. |
| Dilation2dBackpropFilter<T extends TNumber> | Computes the gradient of morphological 2-D dilation with respect to the filter. |
| Dilation2dBackpropInput<T extends TNumber> | Computes the gradient of morphological 2-D dilation with respect to the input. |
| Elu<T extends TNumber> | Computes exponential linear: `exp(features) - 1` if < 0, `features` otherwise. |
| EluGrad<T extends TNumber> | Computes gradients for the exponential linear (Elu) operation. |
| FixedUnigramCandidateSampler | Generates labels for candidate sampling with a learned unigram distribution. |
| FixedUnigramCandidateSampler.Options | Optional attributes for FixedUnigramCandidateSampler
|
| FractionalAvgPool<T extends TNumber> | Performs fractional average pooling on the input. |
| FractionalAvgPool.Options | Optional attributes for FractionalAvgPool
|
| FractionalAvgPoolGrad<T extends TNumber> | Computes gradient of the FractionalAvgPool function. |
| FractionalAvgPoolGrad.Options | Optional attributes for FractionalAvgPoolGrad
|
| FractionalMaxPool<T extends TNumber> | Performs fractional max pooling on the input. |
| FractionalMaxPool.Options | Optional attributes for FractionalMaxPool
|
| FractionalMaxPoolGrad<T extends TNumber> | Computes gradient of the FractionalMaxPool function. |
| FractionalMaxPoolGrad.Options | Optional attributes for FractionalMaxPoolGrad
|
| FusedBatchNorm<T extends TNumber, U extends TNumber> | Batch normalization. |
| FusedBatchNorm.Options | Optional attributes for FusedBatchNorm
|
| FusedBatchNormGrad<T extends TNumber, U extends TNumber> | Gradient for batch normalization. |
| FusedBatchNormGrad.Options | Optional attributes for FusedBatchNormGrad
|
| FusedPadConv2d<T extends TNumber> | Performs a padding as a preprocess during a convolution. |
| FusedResizeAndPadConv2d<T extends TNumber> | Performs a resize and padding as a preprocess during a convolution. |
| FusedResizeAndPadConv2d.Options | Optional attributes for FusedResizeAndPadConv2d
|
| GRUBlockCell<T extends TNumber> | Computes the GRU cell forward propagation for 1 time step. |
| GRUBlockCellGrad<T extends TNumber> | Computes the GRU cell back-propagation for 1 time step. |
| InTopK | Says whether the targets are in the top `K` predictions. |
| InvGrad<T extends TType> | Computes the gradient for the inverse of `x` wrt its input. |
| IsotonicRegression<U extends TNumber> | Solves a batch of isotonic regression problems. |
| L2Loss<T extends TNumber> | L2 Loss. |
| LeakyRelu<T extends TNumber> | Computes rectified linear: `max(features, features * alpha)`. |
| LeakyRelu.Options | Optional attributes for LeakyRelu
|
| LearnedUnigramCandidateSampler | Generates labels for candidate sampling with a learned unigram distribution. |
| LearnedUnigramCandidateSampler.Options | Optional attributes for LearnedUnigramCandidateSampler
|
| LocalResponseNormalization<T extends TNumber> | Local Response Normalization. |
| LocalResponseNormalization.Options | Optional attributes for LocalResponseNormalization
|
| LocalResponseNormalizationGrad<T extends TNumber> | Gradients for Local Response Normalization. |
| LocalResponseNormalizationGrad.Options | Optional attributes for LocalResponseNormalizationGrad
|
| LogSoftmax<T extends TNumber> | Computes log softmax activations. |
| LSTMBlockCell<T extends TNumber> | Computes the LSTM cell forward propagation for 1 time step. |
| LSTMBlockCell.Options | Optional attributes for LSTMBlockCell
|
| LSTMBlockCellGrad<T extends TNumber> | Computes the LSTM cell backward propagation for 1 timestep. |
| MaxPool<T extends TType> | Performs max pooling on the input. |
| MaxPool.Options | Optional attributes for MaxPool
|
| MaxPool3d<T extends TNumber> | Performs 3D max pooling on the input. |
| MaxPool3d.Options | Optional attributes for MaxPool3d
|
| MaxPool3dGrad<U extends TNumber> | Computes gradients of 3D max pooling function. |
| MaxPool3dGrad.Options | Optional attributes for MaxPool3dGrad
|
| MaxPool3dGradGrad<T extends TNumber> | Computes second-order gradients of the maxpooling function. |
| MaxPool3dGradGrad.Options | Optional attributes for MaxPool3dGradGrad
|
| MaxPoolGrad<T extends TNumber> | Computes gradients of the maxpooling function. |
| MaxPoolGrad.Options | Optional attributes for MaxPoolGrad
|
| MaxPoolGradGrad<T extends TNumber> | Computes second-order gradients of the maxpooling function. |
| MaxPoolGradGrad.Options | Optional attributes for MaxPoolGradGrad
|
| MaxPoolGradGradWithArgmax<T extends TNumber> | Computes second-order gradients of the maxpooling function. |
| MaxPoolGradGradWithArgmax.Options | Optional attributes for MaxPoolGradGradWithArgmax
|
| MaxPoolGradWithArgmax<T extends TNumber> | Computes gradients of the maxpooling function. |
| MaxPoolGradWithArgmax.Options | Optional attributes for MaxPoolGradWithArgmax
|
| MaxPoolWithArgmax<T extends TNumber, U extends TNumber> | Performs max pooling on the input and outputs both max values and indices. |
| MaxPoolWithArgmax.Options | Optional attributes for MaxPoolWithArgmax
|
| NthElement<T extends TNumber> | Finds values of the `n`-th order statistic for the last dimension. |
| NthElement.Options | Optional attributes for NthElement
|
| QuantizedAvgPool<T extends TType> | Produces the average pool of the input tensor for quantized types. |
| QuantizedBatchNormWithGlobalNormalization<U extends TType> | Quantized Batch normalization. |
| QuantizedBiasAdd<V extends TType> | Adds Tensor 'bias' to Tensor 'input' for Quantized types. |
| QuantizedConv2d<V extends TType> | Computes a 2D convolution given quantized 4D input and filter tensors. |
| QuantizedConv2d.Options | Optional attributes for QuantizedConv2d
|
| QuantizedConv2DAndRelu<V extends TType> | |
| QuantizedConv2DAndRelu.Options | Optional attributes for QuantizedConv2DAndRelu
|
| QuantizedConv2DAndReluAndRequantize<V extends TType> | |
| QuantizedConv2DAndReluAndRequantize.Options | Optional attributes for QuantizedConv2DAndReluAndRequantize
|
| QuantizedConv2DAndRequantize<V extends TType> | |
| QuantizedConv2DAndRequantize.Options | Optional attributes for QuantizedConv2DAndRequantize
|
| QuantizedConv2DPerChannel<V extends TType> | Computes QuantizedConv2D per channel. |
| QuantizedConv2DPerChannel.Options | Optional attributes for QuantizedConv2DPerChannel
|
| QuantizedConv2DWithBias<V extends TType> | |
| QuantizedConv2DWithBias.Options | Optional attributes for QuantizedConv2DWithBias
|
| QuantizedConv2DWithBiasAndRelu<V extends TType> | |
| QuantizedConv2DWithBiasAndRelu.Options | Optional attributes for QuantizedConv2DWithBiasAndRelu
|
| QuantizedConv2DWithBiasAndReluAndRequantize<W extends TType> | |
| QuantizedConv2DWithBiasAndReluAndRequantize.Options | Optional attributes for QuantizedConv2DWithBiasAndReluAndRequantize
|
| QuantizedConv2DWithBiasAndRequantize<W extends TType> | |
| QuantizedConv2DWithBiasAndRequantize.Options | Optional attributes for QuantizedConv2DWithBiasAndRequantize
|
| QuantizedConv2DWithBiasSignedSumAndReluAndRequantize<X extends TType> | |
| QuantizedConv2DWithBiasSignedSumAndReluAndRequantize.Options | Optional attributes for QuantizedConv2DWithBiasSignedSumAndReluAndRequantize
|
| QuantizedConv2DWithBiasSumAndRelu<V extends TType> | |
| QuantizedConv2DWithBiasSumAndRelu.Options | Optional attributes for QuantizedConv2DWithBiasSumAndRelu
|
| QuantizedConv2DWithBiasSumAndReluAndRequantize<X extends TType> | |
| QuantizedConv2DWithBiasSumAndReluAndRequantize.Options | Optional attributes for QuantizedConv2DWithBiasSumAndReluAndRequantize
|
| QuantizedDepthwiseConv2D<V extends TType> | Computes quantized depthwise Conv2D. |
| QuantizedDepthwiseConv2D.Options | Optional attributes for QuantizedDepthwiseConv2D
|
| QuantizedDepthwiseConv2DWithBias<V extends TType> | Computes quantized depthwise Conv2D with Bias. |
| QuantizedDepthwiseConv2DWithBias.Options | Optional attributes for QuantizedDepthwiseConv2DWithBias
|
| QuantizedDepthwiseConv2DWithBiasAndRelu<V extends TType> | Computes quantized depthwise Conv2D with Bias and Relu. |
| QuantizedDepthwiseConv2DWithBiasAndRelu.Options | Optional attributes for QuantizedDepthwiseConv2DWithBiasAndRelu
|
| QuantizedDepthwiseConv2DWithBiasAndReluAndRequantize<W extends TType> | Computes quantized depthwise Conv2D with Bias, Relu and Requantize. |
| QuantizedDepthwiseConv2DWithBiasAndReluAndRequantize.Options | Optional attributes for QuantizedDepthwiseConv2DWithBiasAndReluAndRequantize
|
| QuantizedInstanceNorm<T extends TType> | Quantized Instance normalization. |
| QuantizedInstanceNorm.Options | Optional attributes for QuantizedInstanceNorm
|
| QuantizedMaxPool<T extends TType> | Produces the max pool of the input tensor for quantized types. |
| QuantizedRelu<U extends TType> | Computes Quantized Rectified Linear: `max(features, 0)` |
| QuantizedRelu6<U extends TType> | Computes Quantized Rectified Linear 6: `min(max(features, 0), 6)` |
| QuantizedReluX<U extends TType> | Computes Quantized Rectified Linear X: `min(max(features, 0), max_value)` |
| Relu<T extends TType> | Computes rectified linear: `max(features, 0)`. |
| Relu6<T extends TNumber> | Computes rectified linear 6: `min(max(features, 0), 6)`. |
| Relu6Grad<T extends TNumber> | Computes rectified linear 6 gradients for a Relu6 operation. |
| ReluGrad<T extends TNumber> | Computes rectified linear gradients for a Relu operation. |
| Selu<T extends TNumber> | Computes scaled exponential linear: `scale * alpha * (exp(features) - 1)`
if < 0, `scale * features` otherwise. |
| SeluGrad<T extends TNumber> | Computes gradients for the scaled exponential linear (Selu) operation. |
| SigmoidCrossEntropyWithLogits | |
| Softmax<T extends TNumber> | Computes softmax activations. |
| SoftmaxCrossEntropyWithLogits | |
| Softsign<T extends TNumber> | Computes softsign: `features / (abs(features) + 1)`. |
| SoftsignGrad<T extends TNumber> | Computes softsign gradients for a softsign operation. |
| SpaceToBatch<T extends TType> | SpaceToBatch for 4-D tensors of type T. |
| SpaceToDepth<T extends TType> | SpaceToDepth for tensors of type T. |
| SpaceToDepth.Options | Optional attributes for SpaceToDepth
|
| SparseSoftmaxCrossEntropyWithLogits | |
| TopK<T extends TNumber> | Finds values and indices of the `k` largest elements for the last dimension. |
| TopK.Options | Optional attributes for TopK
|