Supported Frameworks#
We support almost all the commonly used Machine Learning frameworks & libraries such as ONNX, PyTorch, Keras, TensorFlow, and TensorFlow Lite.
Supported Operations#
Supported Operations#
Layer type |
Comments |
|---|---|
Convolution |
|
DepthwiseConvolution |
|
GroupConvolution |
|
Pooling |
|
TransposeConvolution |
|
AveragePool |
|
MaxPool |
|
Pad |
|
Relu |
|
Erf |
|
LeakyRelu |
|
PRelu |
|
Clip |
|
Softplus |
|
Sigmoid |
|
Gelu |
|
Swish |
|
HardSwish |
|
HardSigmoid |
|
Tanh |
|
Elu |
|
QuickGelu |
|
Celu |
|
Pow |
|
Sqrt |
|
Add |
|
Div |
|
Mul |
|
Sub |
|
ConstantOfShape |
|
Shape |
|
SpaceToDepth |
|
DepthToSpace |
|
Exp |
|
Elu |
|
Log |
|
Neg |
|
Mish |
|
Pow |
|
Sqrt |
|
InstanceNormalization |
|
L2Normalization |
|
LayerNormalization |
|
GroupNormalization |
|
RmsNormalization |
|
ReduceMax |
|
Concatenate |
|
Tile |
Depends on graph structure and node information |
Upsampling |
|
Resize |
|
Cast |
|
Einsum |
Depends on graph structure and node information |
ReduceMean |
|
Slice |
|
Erf |
|
Flatten |
|
Expand |
|
Flip |
Depends on graph structure and node information |
Gather |
Depends on graph structure and node information |
GatherND |
Depends on graph structure and node information |
Gemm |
|
Not |
|
PRelu |
|
Split |
|
Squeeze |
|
TopK |
|
Transpose |
|
Unsqueeze |
|
Reshape |
|
ReduceProd |
|
Reciprocal |
|
ReduceL2 |
|
MatMul |
|
Softmax |
|
Softplus |
|
Abs |
To be supported |
Sin |
To be supported |
Cos |
To be supported |
Crop |
To be supported |
ReduceMin |
To be supported |