Skip to content

Latest commit

 

History

History
161 lines (135 loc) · 5.6 KB

export.md

File metadata and controls

161 lines (135 loc) · 5.6 KB
Raw

Export

  1. Introduction

  2. Supported Framework Model Matrix

  3. Examples

  4. Appendix

Introduction

Open Neural Network Exchange (ONNX) is an open standard format for representing machine learning models. Exporting FP32 PyTorch/Tensorflow models has become popular and easy to use. For Intel Neural Compressor, we hope to export the INT8 model into the ONNX format to achieve higher applicability in multiple frameworks.

Here is the workflow of our export API for PyTorch/Tensorflow FP32/INT8 model.

Architecture

Supported Framework Model Matrix

Framework model type exported ONNX model type
PyTorch FP32 FP32
Post-Training Static Quantized INT8 QOperator/QDQ INT8
Post-Training Dynamic Quantized INT8 QOperator INT8
Quantization-aware Training INT8 QOperator/QDQ INT8
TensorFlow FP32 FP32
Post-Training Static Quantized INT8 QDQ INT8
Quantization-aware Training INT8 QDQ INT8

Examples

PyTorch Model

FP32 Model Export

from neural_compressor.experimental.common import Model
from neural_compressor.config import Torch2ONNXConfig
inc_model = Model(model)
fp32_onnx_config = Torch2ONNXConfig(
    dtype="fp32",
    example_inputs=torch.randn(1, 3, 224, 224),
    input_names=['input'],
    output_names=['output'],
    dynamic_axes={"input": {0: "batch_size"},
                    "output": {0: "batch_size"}},
)
inc_model.export('fp32-model.onnx', fp32_onnx_config)

INT8 Model Export

# q_model is a Neural Compressor model after performing quantization.
from neural_compressor.config import Torch2ONNXConfig
int8_onnx_config = Torch2ONNXConfig(
    dtype="int8",
    opset_version=14,
    quant_format="QOperator", # or QDQ
    example_inputs=torch.randn(1, 3, 224, 224),
    input_names=['input'],
    output_names=['output'],
    dynamic_axes={"input": {0: "batch_size"},
                    "output": {0: "batch_size"}},
)
q_model.export('int8-model.onnx', int8_onnx_config)

Note: Two export examples covering computer vision and natural language processing tasks exist in examples. Users can leverage them to verify the accuracy and performance of the exported ONNX model.

Tensorflow Model

FP32 Model Export

from neural_compressor.experimental.common import Model
from neural_compressor.config import TF2ONNXConfig
inc_model = Model(model)
config = TF2ONNXConfig(dtype='fp32')
inc_model.export('fp32-model.onnx', config)

INT8 Model Export

# q_model is a Neural Compressor model after performing quantization.
from neural_compressor.config import TF2ONNXConfig
config = TF2ONNXConfig(dtype='int8')
q_model.export('int8-model.onnx', config)

Note: Some export examples of computer vision task exist in examples. Users can leverage them to verify the accuracy and performance of the exported ONNX model.

Appendix

Supported quantized ops

This table lists the TorchScript operators that are supported by ONNX export with torch v2.0. Refer to this link for more supported/unsupported ops.

Operator opset_version(s)
quantized::add Since opset 10
quantized::add_relu Since opset 10
quantized::cat Since opset 10
quantized::conv1d_relu Since opset 10
quantized::conv2d Since opset 10
quantized::conv2d_relu Since opset 10
quantized::group_norm Since opset 10
quantized::hardswish Since opset 10
quantized::instance_norm Since opset 10
quantized::layer_norm Since opset 10
quantized::leaky_relu Since opset 10
quantized::linear Since opset 10
quantized::mul Since opset 10
quantized::sigmoid Since opset 10

Note: The export function may fail due to unsupported operations. Please fallback unsupported quantized ops by setting 'op_type_dict' or 'op_name_dict' in 'QuantizationAwareTrainingConfig' or 'PostTrainingQuantConfig' config. Fallback examples please refer to Text classification