- Note
In this document, the numbers in the header of all tables represent the version of onnx opset.
- ✓: onnx specification defined, and supported.
- X: onnx specification defined, but not support yet.
- Empty: Not defined (Support status follows latest).
Total: 95/159
ONNX Operator | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | NNabla Func | Description |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
✓ ✓ ✓ ✓ ✓ ✓ X ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ X ✓ ✓ ✓ X X ✓ ✓ ✓ ✓ X ✓ ✓ X ✓ ✓ ✓ ✓ X X X ✓ ✓ ✓ X ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ X X X X X ✓ X X X X ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ X ✓ X ✓ ✓ X ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ X ✓ ✓ X ✓ |
X X ✓ ✓ |
X |
✓ |
✓ |
✓ ✓ X ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ |
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ X ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ X ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ X ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ X ✓ ✓ ✓ ✓ ✓ ✓ X ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ X ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ X ✓ |
✓ ✓ X ✓ ✓ X X |
✓ ✓ ✓ ✓ X X X ✓ ✓ X ✓ X ✓ ✓ ✓ ✓ ✓ ✓ ✓ X ✓ X ✓ X X X X X X ✓ ✓ X X ✓ ✓ |
X X X X ✓ X ✓ X X X X X X ✓ X X X ✓ X ✓ X X |
✓ ✓ X X ✓ X X X X X X X ✓ X X X ✓ ✓ X X ✓ ✓ ✓ X X X X X X ✓ X X X X X ✓ ✓ ✓ ✓ ✓ ✓ X ✓ X X X X X X X X X X ✓ ✓ X ✓ X X ✓ |
✓ ✓ ✓ ✓ X ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ X X X ✓ X ✓ ✓ ✓ X X ✓ ✓ ✓ ✓ ✓ X ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ X ✓ ✓ X ✓ ✓ ✓ ✓ ✓ X X X ✓ ✓ X X ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ |
Abs ACos ACosh Add2, Reshape LogicalAnd, Reshape Flip, Max, RSubScalar Flip, Min, RSubScalar ASin ASinh ATan ATanh AveragePooling, Pad BatchNormalization Ceil Add2, Constant, Div2, ELU, Exp, MaximumScalar, MinimumScalar, Mul2, MulScalar, Reshape, Sub2 Identity Concatenate Identity Constant Convolution Deconvolution, Pad Cos Cosh Reshape, Transpose DequantizeLinear Div2, Reshape Identity ELU Equal, Reshape Exp Broadcast, Reshape Reshape Floor Concatenate, Slice Add2, BatchMatmul, MulScalar, Reshape GlobalAveragePooling Greater, Reshape Equal, Greater, GreaterEqual, LogicalOr, Reshape AddScalar, HardSigmoid, MaximumScalar, MinimumScalar, MulScalar Max, OneHot, Transpose Identity BatchNormalization, Concatenate, Reshape, Split IsInf IsNaN AddScalar, Div2, MulScalar, PowScalar, SumPooling, Transpose LeakyReLU Less, Reshape Equal, Less, LessEqual, LogicalOr, Reshape Log Div2, Exp, Log, Max, Sub2, Sum BatchMatmul, Reshape Maximum2 MaxPooling, Pad Identity, Mean, Stack Minimum2 Mul2, Reshape MulScalar LogicalNot LogicalOr, Reshape PReLU Pad Pow2, Reshape QuantizeLinear RDivScalar Max Mean Min Prod Sum PowScalar, Sum ReLU Reshape Round SELU Sigmoid Sign Sin Sinh Slice Div2, Exp, Max, Sub2, Sum SoftPlus SoftSign Reshape, Transpose Split, Stack PowScalar Reshape Reshape, Sub2 AddN Tan Tanh Constant, GreaterScalar, Where Tile Transpose Reshape Unpooling Where LogicalXor, Reshape |
Not all features are verified. Those features can be verified by ONNXRuntime when opset > 6. Some feature is not supported by Nnabla such as Pad's edge mode. if opset >= 10, the ceil_mode is not supported. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not all features are verified. Those features can be verified by ONNXRuntime. if opset >= 10, the ceil_mode is not supported, dilations is not equal to 1 is not supported. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Onnx required to support "edge" mode, while nnabla does not support it. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. |
- ✓: Support to export this opset.
- △: Partially support to export this opset (e.g. some cases cannot be supported, or not completely tested).
- X: Supported, but test failed.
- Empty: Not support corresponding opset version.
Total: 124/215
Count 11/18
NNabla Function 7 9 10 11 13 ONNX Op Description Affine RNN LSTM GRU Convolution FusedConvolution DepthwiseConvolution Deconvolution DepthwiseDeconvolution DeformableConvolution AdaptiveSeparableConvolution MaxPooling
✓
✓
✓ ✓ ✓
✓
✓
✓
✓ ✓ ✓
✓
✓
✓
✓ ✓ ✓
✓
✓
✓
✓ ✓ ✓
✓
✓
✓
✓ ✓ ✓
✓
Gemm, Reshape
Conv, Reshape
Conv, Reshape ConvTranspose, Reshape ConvTranspose, Reshape
Constant, MaxPool, Pad, Reshape
Not yet implemented. Not yet implemented. Not yet implemented.
Not yet implemented.
Not yet implemented. Not yet implemented.
AveragePooling GlobalAveragePooling SumPooling Unpooling Embed RoiAlign
△ ✓ ✓ ✓ ✓
△ ✓ ✓ ✓ ✓
△ ✓ ✓ ✓ ✓
△ ✓ ✓ ✓ ✓
△ ✓ ✓ ✓ ✓
AveragePool, Constant, Pad, Reshape GlobalAveragePool AveragePool, Constant, Mul, Pad, Reshape Resize Gather
Currently only supports the cases where both ignore_border and including_pad are True.
Not yet implemented.
Count 20/22
NNabla Function 7 9 10 11 13 ONNX Op Description Sigmoid Swish Tanh ReLU LeakyReLU Softmax LogSoftmax ELU SELU CReLU CELU PReLU GELU Mish ReLU6 HardSigmoid HardTanh LogSigmoid
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
✓ ✓ ✓ ✓
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
✓ ✓ ✓ ✓
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
✓ ✓ ✓ ✓
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
✓ ✓ ✓ ✓
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
✓ ✓ ✓ ✓
Sigmoid Mul, Sigmoid Tanh Relu LeakyRelu Div, Exp, ReduceMax, ReduceSum, Sub Exp, Log, ReduceMax, ReduceSum, Sub Elu Selu Concat, Neg, Relu Concat, Elu, Neg PRelu, Reshape Add, Constant, Div, Mul, Pow, Sqrt, Tanh
Constant, Min, Relu HardSigmoid Constant, Max, Min, Neg Log, Sigmoid
Not yet implemented.
SoftPlus SoftSign TanhShrink Sinc
X ✓ ✓ X
X ✓ ✓ X
X ✓ ✓ X
X ✓ ✓ ✓
X ✓ ✓ ✓
Softplus Softsign Sub, Tanh Constant, Div, Equal, Sin, Where
Not yet implemented.
Count 7/14
NNabla Function 7 9 10 11 13 ONNX Op Description FusedBatchNormalization BatchNormalization GroupNormalization InstanceNormalization LayerNormalization NormNormalization SyncBatchNormalization TensorNormalization WeightNormalization WeightStandardization SpectralNorm MeanSubtraction ClipGradByValue ClipGradByNorm
✓ ✓
✓ ✓
✓ ✓ ✓
✓ ✓
✓ ✓
✓ ✓ ✓
✓ ✓
✓ ✓
✓ ✓ ✓
✓ ✓
✓ ✓
✓ ✓ ✓
✓ ✓
✓ ✓
✓ ✓ ✓
Add, BatchNormalization, Constant, Div, Mul, ReduceMean, ReduceSum, Relu, Reshape, Squeeze, Sub BatchNormalization, Constant, Div, Mul, ReduceMean, ReduceSum, Reshape, Squeeze, Sub
Add, Constant, Div, Mul, Pow, ReduceMean, ReduceSum, Reshape, Sub Add, Constant, Div, Mul, Pow, ReduceMean, ReduceSum, Sub
Add, Constant, Mul, Pow, ReduceSum, Reshape Add, Constant, Div, Mul, Pow, ReduceMean, ReduceSum, Sub Add, Constant, Div, Gemm, Pow, ReduceSum, Reshape, Sqrt, Transpose
Not yet implemented.
Not yet implemented. Not yet implemented. Not yet implemented.
Not yet implemented. Not yet implemented. Not yet implemented.
Count 5/10
NNabla Function 7 9 10 11 13 ONNX Op Description Sum CumSum Mean Max Min Norm Prod CumProd ReduceSum ReduceMean
✓
✓ ✓ ✓
✓
✓
✓ ✓ ✓
✓
✓
✓ ✓ ✓
✓
✓
✓ ✓ ✓
✓
✓
✓ ✓ ✓
✓
ReduceSum
ReduceMean ReduceMax ReduceMin
ReduceProd
Not yet implemented.
Not yet implemented.
Not yet implemented. Not yet implemented. Not yet implemented.
Count 11/14
NNabla Function 7 9 10 11 13 ONNX Op Description Add2 AddN BcAdd2 Sub2 Mul2 MulN Div2 Pow2 AddScalar MulScalar PowScalar RSubScalar RDivScalar RPowScalar
✓
✓ ✓
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
✓
✓ ✓
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
✓
✓ ✓
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
✓
✓ ✓
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
✓
✓ ✓
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
Add
Sub Mul
Div Pow Add, Constant Constant, Mul Constant, Pow Constant, Sub Constant, Div Constant, Pow
Not yet implemented. Not yet implemented.
Not yet implemented.
Count 29/30
NNabla Function 7 9 10 11 13 ONNX Op Description Sign Minimum2 Maximum2 MinimumScalar MaximumScalar LogicalAnd LogicalOr LogicalXor Equal NotEqual GreaterEqual Greater LessEqual Less SearchSorted LogicalAndScalar LogicalOrScalar LogicalXorScalar EqualScalar NotEqualScalar GreaterEqualScalar GreaterScalar LessEqualScalar LessScalar LogicalNot IsNaN IsInf ResetNaN ResetInf Where
X ✓ ✓ ✓ ✓ ✓ ✓ ✓ X X ✓ ✓ ✓ ✓
✓ ✓ ✓ X X ✓ ✓ ✓ ✓ ✓ X X X X X
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ X X ✓ ✓ ✓ ✓
✓ ✓ ✓ X X ✓ ✓ ✓ ✓ ✓ ✓ X ✓ X ✓
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ X X ✓ ✓ ✓ ✓
✓ ✓ ✓ X X ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
Sign Add, Constant, Min Add, Constant, Max Add, Constant, Min Add, Constant, Max And Or Xor Equal Equal, Not Less, Not Greater Greater, Not Less
And, Constant Constant, Or Constant, Xor Constant, Equal Constant, Equal, Not Constant, Less, Not Constant, Greater Constant, Greater, Not Constant, Less Not IsNaN IsInf Constant, IsNaN, Where Constant, IsInf, Where Where
Not yet implemented.
Count 22/22
NNabla Function 7 9 10 11 13 ONNX Op Description Constant Arange Abs Exp Log Identity BatchMatmul Round Ceil Floor Sin Cos Tan Sinh Cosh ASin ACos ATan ATan2 ASinh ACosh ATanh
✓ ✓ ✓ ✓ ✓ ✓ ✓ X ✓ ✓ ✓ ✓ ✓ X X ✓ ✓ ✓ ✓ X X X
✓ ✓ ✓ ✓ ✓ ✓ ✓ X ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
✓ ✓ ✓ ✓ ✓ ✓ ✓ X ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
Constant, Identity Constant, Identity Abs Exp Log Identity MatMul, Transpose Round Ceil Floor Sin Cos Tan Sinh Cosh Asin Acos Atan Atan, Div Asinh Acosh Atanh
Count 12/30
NNabla Function 7 9 10 11 13 ONNX Op Description Concatenate Split Stack
✓ ✓ ✓
✓ ✓ ✓
✓ ✓ ✓
✓ ✓ ✓
✓ ✓ ✓
Concat Split, Squeeze Concat, Unsqueeze
Slice
△ △ ✓ ✓ ✓ Constant, Slice ONNX slice cannot support step != 1 on opset < 10. Pad Transpose Broadcast BroadcastTo Tile OneHot Flip Shift Sort Reshape MatrixDiag MatrixDiagPart Meshgrid BatchDet BatchInv BatchLogdet Assign Gather GatherNd BoolGather ScatterNd ScatterAdd BoolScatter BoolFill PackPaddedSequence PadPackedSequence
△ ✓ X ✓ ✓ X ✓
✓
△ ✓ ✓ ✓ ✓ ✓ ✓
✓
△ ✓ ✓ ✓ ✓ ✓ ✓
✓
△ ✓ ✓ ✓ ✓ ✓ ✓
✓
△ ✓ ✓ ✓ ✓ ✓ ✓
✓
Constant, Pad Transpose
Constant, Reshape, Tile Flatten, Gather, Reshape Gather, Identity, Transpose
Constant, Reshape
When the mode of the pad is reflect, if the size of the pad exceeds the input size, onnxruntime cannot handle it.
Not yet implemented. Not yet implemented.
Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented.
Count 1/5
NNabla Function 7 9 10 11 13 ONNX Op Description Interpolate FFT IFFT STFT ISTFT
X
X
△
✓
✓
Resize
Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented.
Count 0/15
NNabla Function 7 9 10 11 13 ONNX Op Description Dropout TopKData TopKGrad Rand Randint Randn RandBinomial RandBeta RandGamma RandomChoice RandomCrop RandomFlip RandomShift RandomErase ImageAugmentation
X
X
X
X
X
Dropout
The Dropout in nnabla has no test mode and contains random parameters, so the test result is not the same as onnx. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented.
Count 0/9
NNabla Function 7 9 10 11 13 ONNX Op Description SigmoidCrossEntropy BinaryCrossEntropy SoftmaxCrossEntropy CategoricalCrossEntropy SquaredError AbsoluteError HuberLoss EpsilonInsensitiveLoss KLMultinomial
Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented.
Count 0/3
NNabla Function 7 9 10 11 13 ONNX Op Description AffineGrid WarpByGrid WarpByFlow
Not yet implemented. Not yet implemented. Not yet implemented.
Count 6/14
NNabla Function 7 9 10 11 13 ONNX Op Description BinarySigmoid BinaryTanh BinaryConnectAffine BinaryConnectConvolution BinaryWeightAffine BinaryWeightConvolution INQAffine INQConvolution FixedPointQuantize MinMaxQuantize Pow2Quantize Prune QuantizeLinear DequantizeLinear
X X ✓ ✓ ✓ ✓
✓ ✓ ✓ ✓ ✓ ✓
✓ ✓ ✓ ✓ ✓ ✓
✓ ✓
✓ ✓ ✓ ✓ ✓ ✓
✓ ✓ ✓ ✓ ✓ ✓
✓ ✓
Constant, Greater, Where Constant, Greater, Where Gemm, Reshape Conv, Reshape Add, MatMul, Mul, Reshape Add, Conv, Mul, Reshape
Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented.
Count 0/3
NNabla Function 7 9 10 11 13 ONNX Op Description TopNError BinaryError ConfusionMatrix
Not yet implemented. Not yet implemented. Not yet implemented.
Count 0/6
NNabla Function 7 9 10 11 13 ONNX Op Description VATNoise Unlink Sink NmsDetection2d MaxPoolingBackward PatchCorrelation
Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented.
- ✓: Supported
- △: Partially supported
- X: Supported, but test failed.
- Empty: Not support yet.
Total: 106/122
Tensorflow support statusTensorflow Function | Status | NNabla Func | Description |
---|---|---|---|
|
|
Abs ACos ACosh Add2 AddN Greater, Min, Reshape Greater, Reshape, Sum Max Min ASin ASinh ATan ATan, Add2, Div2, Mul2, Reshape, Sign, Sub2 ATanh AveragePooling, Pad, Transpose AveragePooling, Pad, Transpose BatchMatmul, Transpose Add2, Mul2, PowScalar, RDivScalar, Reshape, Sub2 Add2, Reshape |
|
|
|
NA Ceil Maximum2, Minimum2, Reshape Concatenate Concatenate NA Convolution, Pad, Reshape, Transpose Deconvolution, Reshape, Transpose Convolution, Pad, Transpose Deconvolution, Transpose Convolution, Pad, Transpose Deconvolution, Pad, Transpose Cos Cosh Concatenate, MulScalar, ReLU Reshape, Transpose Convolution, Pad, Reshape, Transpose Div2 ELU Equal Exp Reshape Floor Div2, Floor Div2, Floor, Mul2, Sub2 |
Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. |
|
|
Concatenate, Slice Greater Less, LogicalNot Identity IsInf IsNaN LeakyReLU Less Greater, LogicalNot Log |
Not yet implemented. |
|
|
MulScalar, SoftPlus Add2, Exp, Log, Max, Reshape, Sub2, Sum, Transpose LogicalAnd LogicalNot LogicalOr LogicalAnd, LogicalNot, LogicalOr Max MaxPooling, Pad, Reshape, Transpose MaxPooling, Pad, Transpose Maximum2 Mean Min Minimum2 Mul2 MulScalar Equal, LogicalNot Concatenate, Reshape Pad Pow2 Prod Div2 RDivScalar ReLU MaximumScalar, MinimumScalar Reshape Round PowScalar, RDivScalar SELU Sigmoid Sign Sin Sinh Slice Div2, Exp, Max, Reshape, Sub2, Sum, Transpose |
Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. |
|
|
SoftPlus SoftSign Reshape, Transpose Split, Stack Split, Stack PowScalar Mul2 Mul2, Sub2 Reshape Identity Slice Sub2 Sum Mul2, Sigmoid Tan Tanh Tile Transpose |
Not yet implemented. Not yet implemented. |
|
|
Div2 Reshape, Split, Stack Where NA |
Not yet implemented. Not yet implemented. |
- ✓: Supported
- △: Partially supported
- X: Supported, but test failed.
- Empty: Not support yet.
Total: 124/215
Count 11/18
NNabla Function Status Description Affine RNN LSTM GRU
✓
Not yet implemented. Not yet implemented. Not yet implemented.
Convolution FusedConvolution
△
The cases dilations and strides larger than 1 are not supported by tensorflow. Not yet implemented.
DepthwiseConvolution
△ The cases dilations and strides larger than 1 are not supported by tensorflow. Deconvolution
△ The cases dilations larger than 1 are not supported by tensorflow. DepthwiseDeconvolution DeformableConvolution AdaptiveSeparableConvolution MaxPooling
△
✓
The cases dilations larger than 1 are not supported by tensorflow. Not yet implemented. Not yet implemented.
AveragePooling GlobalAveragePooling SumPooling
△ ✓ ✓
Currently only supports the cases both ignore_border and including_pad are True.
Unpooling Embed RoiAlign
△ ✓
The kernel only supports 2d.
Not yet implemented.
Count 20/22
NNabla Function Status Description Sigmoid Swish Tanh ReLU LeakyReLU Softmax LogSoftmax ELU SELU CReLU CELU PReLU GELU Mish ReLU6 HardSigmoid HardTanh LogSigmoid
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ △ ✓ ✓ ✓ ✓
✓ ✓ ✓ ✓
Not yet implemented.
SoftPlus SoftSign TanhShrink Sinc
X ✓ ✓ ✓
Not yet implemented.
Count 7/14
NNabla Function Status Description FusedBatchNormalization BatchNormalization GroupNormalization InstanceNormalization LayerNormalization NormNormalization SyncBatchNormalization TensorNormalization WeightNormalization WeightStandardization SpectralNorm MeanSubtraction ClipGradByValue ClipGradByNorm
✓ ✓
✓ ✓
✓ ✓ ✓
Not yet implemented.
Not yet implemented. Not yet implemented. Not yet implemented.
Not yet implemented. Not yet implemented. Not yet implemented.
Count 5/10
NNabla Function Status Description Sum CumSum Mean Max Min Norm Prod CumProd ReduceSum ReduceMean
✓
✓ ✓ ✓
✓
Not yet implemented.
Not yet implemented.
Not yet implemented. Not yet implemented. Not yet implemented.
Count 11/14
NNabla Function Status Description Add2 AddN BcAdd2 Sub2 Mul2 MulN Div2 Pow2 AddScalar MulScalar PowScalar RSubScalar RDivScalar RPowScalar
✓
✓ ✓
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
Not yet implemented. Not yet implemented.
Not yet implemented.
Count 29/30
NNabla Function Status Description Sign Minimum2 Maximum2 MinimumScalar MaximumScalar LogicalAnd LogicalOr LogicalXor Equal NotEqual GreaterEqual Greater LessEqual Less SearchSorted LogicalAndScalar LogicalOrScalar LogicalXorScalar EqualScalar NotEqualScalar GreaterEqualScalar GreaterScalar LessEqualScalar LessScalar LogicalNot IsNaN IsInf ResetNaN ResetInf Where
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
Not yet implemented.
Count 22/22
NNabla Function Status Description Constant Arange Abs Exp Log Identity BatchMatmul Round Ceil Floor Sin Cos Tan Sinh Cosh ASin ACos ATan ATan2 ASinh ACosh ATanh
✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
Count 12/30
NNabla Function Status Description Concatenate Split Stack Slice
✓ ✓ ✓ ✓
Pad Transpose Broadcast BroadcastTo Tile OneHot Flip Shift Sort Reshape MatrixDiag MatrixDiagPart Meshgrid BatchDet BatchInv BatchLogdet Assign Gather GatherNd BoolGather ScatterNd ScatterAdd BoolScatter BoolFill PackPaddedSequence PadPackedSequence
△ ✓ ✓ ✓ ✓ ✓ ✓
✓
When the mode of the pad is reflect, if the size of the pad exceeds the input size, tensorflow cannot handle it.
Not yet implemented. Not yet implemented.
Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented.
Count 1/5
NNabla Function Status Description Interpolate FFT IFFT STFT ISTFT
△
Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented.
Count 0/15
NNabla Function Status Description Dropout TopKData TopKGrad Rand Randint Randn RandBinomial RandBeta RandGamma RandomChoice RandomCrop RandomFlip RandomShift RandomErase ImageAugmentation
X
The Dropout in nnabla has no test mode and contains random parameters, so the test result is not the same as tensorflow. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented.
Count 0/9
NNabla Function Status Description SigmoidCrossEntropy BinaryCrossEntropy SoftmaxCrossEntropy CategoricalCrossEntropy SquaredError AbsoluteError HuberLoss EpsilonInsensitiveLoss KLMultinomial
Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented.
Count 0/3
NNabla Function Status Description AffineGrid WarpByGrid WarpByFlow
Not yet implemented. Not yet implemented. Not yet implemented.
Count 6/14
NNabla Function Status Description BinarySigmoid BinaryTanh BinaryConnectAffine
✓ ✓ ✓
BinaryConnectConvolution BinaryWeightAffine
△ ✓
The cases dilations and strides larger than 1 are not supported by tensorflow.
BinaryWeightConvolution INQAffine INQConvolution FixedPointQuantize MinMaxQuantize Pow2Quantize Prune QuantizeLinear DequantizeLinear
△
The cases dilations and strides larger than 1 are not supported by tensorflow. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented.
Count 0/3
NNabla Function Status Description TopNError BinaryError ConfusionMatrix
Not yet implemented. Not yet implemented. Not yet implemented.
Count 0/6
NNabla Function Status Description VATNoise Unlink Sink NmsDetection2d MaxPoolingBackward PatchCorrelation
Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented. Not yet implemented.
- ✓: Supported
- △: Partially supported
- X: Supported, but test failed.
- Empty: Not support yet.
Total: 82/215
Count 11/18
NNabla Function Status Affine RNN LSTM GRU
✓
Convolution FusedConvolution
△
DepthwiseConvolution
✓ Deconvolution
△ DepthwiseDeconvolution DeformableConvolution AdaptiveSeparableConvolution
△
MaxPooling
△ AveragePooling
△ GlobalAveragePooling
✓ SumPooling
△ Unpooling
△ Embed RoiAlign
✓
Count 10/22
NNabla Function Status Sigmoid
✓ Swish
✓ Tanh
✓ ReLU
✓ LeakyReLU
✓ Softmax
✓ LogSoftmax
✓ ELU
△ SELU
X CReLU
X CELU
X PReLU
✓ GELU Mish
X
ReLU6
✓ HardSigmoid
X HardTanh
X LogSigmoid
X SoftPlus
X SoftSign
X TanhShrink
X Sinc
X
Count 1/14
NNabla Function Status FusedBatchNormalization
X BatchNormalization GroupNormalization
✓
InstanceNormalization
X LayerNormalization NormNormalization SyncBatchNormalization TensorNormalization
X
WeightNormalization
X WeightStandardization
X SpectralNorm MeanSubtraction ClipGradByValue ClipGradByNorm
X
Count 5/10
NNabla Function Status Sum CumSum
✓
Mean
✓ Max
✓ Min Norm
✓
Prod CumProd ReduceSum ReduceMean
✓
Count 11/14
NNabla Function Status Add2 AddN BcAdd2
✓
Sub2
✓ Mul2 MulN
✓
Div2
✓ Pow2
✓ AddScalar
✓ MulScalar
✓ PowScalar
✓ RSubScalar
✓ RDivScalar
✓ RPowScalar
✓
Count 23/30
NNabla Function Status Sign
X Minimum2
✓ Maximum2
✓ MinimumScalar
✓ MaximumScalar
✓ LogicalAnd
✓ LogicalOr
✓ LogicalXor
✓ Equal
✓ NotEqual
✓ GreaterEqual
✓ Greater
✓ LessEqual
✓ Less SearchSorted
✓
LogicalAndScalar
✓ LogicalOrScalar
✓ LogicalXorScalar
✓ EqualScalar
✓ NotEqualScalar
✓ GreaterEqualScalar
✓ GreaterScalar
✓ LessEqualScalar
✓ LessScalar
✓ LogicalNot
✓ IsNaN
X IsInf
X ResetNaN
X ResetInf
X Where
X
Count 10/22
NNabla Function Status Constant
X Arange
X Abs
✓ Exp
✓ Log
✓ Identity
X BatchMatmul
✓ Round
✓ Ceil
✓ Floor
✓ Sin
✓ Cos
✓ Tan
✓ Sinh
X Cosh
X ASin
X ACos
X ATan
X ATan2
X ASinh
X ACosh
X ATanh
X
Count 10/30
NNabla Function Status Concatenate
✓ Split
✓ Stack
✓ Slice
△ Pad
△ Transpose
△ Broadcast
△ BroadcastTo
X Tile
✓ OneHot
X Flip Shift Sort
✓
Reshape MatrixDiag MatrixDiagPart Meshgrid BatchDet BatchInv BatchLogdet Assign Gather GatherNd BoolGather ScatterNd ScatterAdd BoolScatter BoolFill PackPaddedSequence PadPackedSequence
✓
Count 1/5
NNabla Function Status Interpolate FFT IFFT STFT ISTFT
△
Count 0/15
NNabla Function Status Dropout TopKData TopKGrad Rand Randint Randn RandBinomial RandBeta RandGamma RandomChoice RandomCrop RandomFlip RandomShift RandomErase ImageAugmentation
X
Count 0/9
NNabla Function Status SigmoidCrossEntropy BinaryCrossEntropy SoftmaxCrossEntropy CategoricalCrossEntropy SquaredError AbsoluteError HuberLoss EpsilonInsensitiveLoss KLMultinomial
Count 0/3
NNabla Function Status AffineGrid WarpByGrid WarpByFlow
Count 0/14
NNabla Function Status BinarySigmoid
X BinaryTanh
X BinaryConnectAffine
X BinaryConnectConvolution
X BinaryWeightAffine
X BinaryWeightConvolution INQAffine INQConvolution FixedPointQuantize MinMaxQuantize Pow2Quantize Prune QuantizeLinear DequantizeLinear
X
Count 0/3
NNabla Function Status TopNError BinaryError ConfusionMatrix
Count 0/6
NNabla Function Status VATNoise Unlink Sink NmsDetection2d MaxPoolingBackward PatchCorrelation
NNabla version: None
- ✓: Supported
- △: Partially supported
- X: Supported, but test failed or no test data.
- Empty: Not support yet.
Total: 56/215
Count 8/18
NNabla Function Status Description Affine RNN LSTM GRU Convolution FusedConvolution DepthwiseConvolution Deconvolution DepthwiseDeconvolution DeformableConvolution AdaptiveSeparableConvolution MaxPooling AveragePooling GlobalAveragePooling SumPooling Unpooling Embed RoiAlign
✓
✓
✓ △
△ ✓
✓ ✓
Count 11/22
NNabla Function Status Description Sigmoid Swish Tanh ReLU LeakyReLU Softmax LogSoftmax ELU SELU CReLU CELU PReLU GELU Mish ReLU6 HardSigmoid HardTanh LogSigmoid SoftPlus SoftSign TanhShrink Sinc
✓ ✓ ✓ ✓ ✓ ✓
✓ ✓ ✓ △ ✓
Count 1/14
NNabla Function Status Description FusedBatchNormalization BatchNormalization GroupNormalization InstanceNormalization LayerNormalization NormNormalization SyncBatchNormalization TensorNormalization WeightNormalization WeightStandardization SpectralNorm MeanSubtraction ClipGradByValue ClipGradByNorm
✓
X
X
Count 1/10
NNabla Function Status Description Sum CumSum Mean Max Min Norm Prod CumProd ReduceSum ReduceMean
△
Count 11/14
NNabla Function Status Description Add2 AddN BcAdd2 Sub2 Mul2 MulN Div2 Pow2 AddScalar MulScalar PowScalar RSubScalar RDivScalar RPowScalar
✓ X
✓ ✓ X ✓ ✓ ✓ ✓ ✓ △ ✓ ✓
Count 5/30
NNabla Function Status Description Sign Minimum2 Maximum2 MinimumScalar MaximumScalar LogicalAnd LogicalOr LogicalXor Equal NotEqual GreaterEqual Greater LessEqual Less SearchSorted LogicalAndScalar LogicalOrScalar LogicalXorScalar EqualScalar NotEqualScalar GreaterEqualScalar GreaterScalar LessEqualScalar LessScalar LogicalNot IsNaN IsInf ResetNaN ResetInf Where
✓ ✓ ✓ ✓ ✓
Count 6/22
NNabla Function Status Description Constant Arange Abs Exp Log Identity BatchMatmul Round Ceil Floor Sin Cos Tan Sinh Cosh ASin ACos ATan ATan2 ASinh ACosh ATanh
✓ ✓ ✓ ✓ △ ✓
Count 7/30
NNabla Function Status Description Concatenate Split Stack Slice Pad Transpose Broadcast BroadcastTo Tile OneHot Flip Shift Sort Reshape MatrixDiag MatrixDiagPart Meshgrid BatchDet BatchInv BatchLogdet Assign Gather GatherNd BoolGather ScatterNd ScatterAdd BoolScatter BoolFill PackPaddedSequence PadPackedSequence
✓ ✓ △ △
△
✓ X
✓ X X
Count 0/5
NNabla Function Status Description Interpolate FFT IFFT STFT ISTFT
Count 0/15
NNabla Function Status Description Dropout TopKData TopKGrad Rand Randint Randn RandBinomial RandBeta RandGamma RandomChoice RandomCrop RandomFlip RandomShift RandomErase ImageAugmentation
X
Count 0/9
NNabla Function Status Description SigmoidCrossEntropy BinaryCrossEntropy SoftmaxCrossEntropy CategoricalCrossEntropy SquaredError AbsoluteError HuberLoss EpsilonInsensitiveLoss KLMultinomial
Count 0/3
NNabla Function Status Description AffineGrid WarpByGrid WarpByFlow
Count 6/14
NNabla Function Status Description BinarySigmoid BinaryTanh BinaryConnectAffine BinaryConnectConvolution BinaryWeightAffine BinaryWeightConvolution INQAffine INQConvolution FixedPointQuantize MinMaxQuantize Pow2Quantize Prune QuantizeLinear DequantizeLinear
✓ ✓ ✓ ✓ ✓ △
Count 0/3
NNabla Function Status Description TopNError BinaryError ConfusionMatrix
Count 0/6
NNabla Function Status Description VATNoise Unlink Sink NmsDetection2d MaxPoolingBackward PatchCorrelation