[Quantization] model quantization for inference

wanchen.swc

2023-03-06 69ccdd35cda4c8482e189fa350fbcb83997872f2

 funasr/export/export_model.py

@@ -7,28 +7,35 @@
import logging
import torch

from funasr.bin.asr_inference_paraformer import Speech2Text
from funasr.export.models import get_model
import numpy as np
import random


# torch_version = float(".".join(torch.__version__.split(".")[:2]))
# assert torch_version > 1.9

class ASRModelExportParaformer:
    def __init__(self, cache_dir: Union[Path, str] = None, onnx: bool = True):
    def __init__(
        self, cache_dir: Union[Path, str] = None, onnx: bool = True, quant: bool = True
    ):
        assert check_argument_types()
        self.set_all_random_seed(0)
        if cache_dir is None:
            cache_dir = Path.home() / "cache" / "export"
            cache_dir = Path.home() / ".cache" / "export"

        self.cache_dir = Path(cache_dir)
        self.export_config = dict(
            feats_dim=560,
            onnx=False,
        )
        logging.info("output dir: {}".format(self.cache_dir))
        print("output dir: {}".format(self.cache_dir))
        self.onnx = onnx
        self.quant = quant
        

    def _export(
        self,
        model: Speech2Text,
        model,
        tag_name: str = None,
        verbose: bool = False,
    ):
@@ -42,25 +49,58 @@
            model,
            self.export_config,
        )
        self._export_onnx(model, verbose, export_dir)
        model.eval()
        # self._export_onnx(model, verbose, export_dir)
        if self.onnx:
            self._export_onnx(model, verbose, export_dir)
        else:
            self._export_torchscripts(model, verbose, export_dir)

        logging.info("output dir: {}".format(export_dir))
        print("output dir: {}".format(export_dir))


    def _torch_quantize(self, model):
        from torch_quant.module import ModuleFilter
        from torch_quant.observer import HistogramObserver
        from torch_quant.quantizer import Backend, Quantizer
        from funasr.export.models.modules.decoder_layer import DecoderLayerSANM
        from funasr.export.models.modules.encoder_layer import EncoderLayerSANM
        module_filter = ModuleFilter(include_classes=[EncoderLayerSANM, DecoderLayerSANM])
        module_filter.exclude_op_types = [torch.nn.Conv1d]
        quantizer = Quantizer(
            module_filter=module_filter,
            backend=Backend.FBGEMM,
            act_ob_ctr=HistogramObserver,
        )
        model.eval()
        calib_model = quantizer.calib(model)
        # run calibration data
        # using dummy inputs for a example
        dummy_input = model.get_dummy_inputs()
        _ = calib_model(*dummy_input)
        quant_model = quantizer.quantize(model)
        return quant_model

    def _export_torchscripts(self, model, verbose, path, enc_size=None):
        if enc_size:
            dummy_input = model.get_dummy_inputs(enc_size)
        else:
            dummy_input = model.get_dummy_inputs_txt()
            dummy_input = model.get_dummy_inputs()

        # model_script = torch.jit.script(model)
        model_script = torch.jit.trace(model, dummy_input)
        model_script.save(os.path.join(path, f'{model.model_name}.torchscripts'))

        if self.quant:
            quant_model = self._torch_quantize(model)
            model_script = torch.jit.trace(quant_model, dummy_input)
            model_script.save(os.path.join(path, f'{model.model_name}_quant.torchscripts'))


    def set_all_random_seed(self, seed: int):
        random.seed(seed)
        np.random.seed(seed)
        torch.random.manual_seed(seed)
    def export(self,
               tag_name: str = 'damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch',
               mode: str = 'paraformer',
@@ -79,9 +119,9 @@
            with open(json_file, 'r') as f:
                config_data = json.load(f)
                mode = config_data['model']['model_config']['mode']
        if mode == 'paraformer':
        if mode.startswith('paraformer'):
            from funasr.tasks.asr import ASRTaskParaformer as ASRTask
        elif mode == 'uniasr':
        elif mode.startswith('uniasr'):
            from funasr.tasks.asr import ASRTaskUniASR as ASRTask
            
        model, asr_train_args = ASRTask.build_model_from_file(
@@ -89,38 +129,6 @@
        )
        self._export(model, tag_name)
            
    # def export_from_modelscope(
    #     self,
    #     tag_name: str = 'damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch',
    # ):
    #
    #     from funasr.tasks.asr import ASRTaskParaformer as ASRTask
    #     from modelscope.hub.snapshot_download import snapshot_download
    #
    #     model_dir = snapshot_download(tag_name, cache_dir=self.cache_dir)
    #     asr_train_config = os.path.join(model_dir, 'config.yaml')
    #     asr_model_file = os.path.join(model_dir, 'model.pb')
    #     cmvn_file = os.path.join(model_dir, 'am.mvn')
    #     model, asr_train_args = ASRTask.build_model_from_file(
    #         asr_train_config, asr_model_file, cmvn_file, 'cpu'
    #     )
    #     self.export(model, tag_name)
    #
    # def export_from_local(
    #     self,
    #     tag_name: str = '/root/cache/export/damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch',
    # ):
    #
    #     from funasr.tasks.asr import ASRTaskParaformer as ASRTask
    #
    #     model_dir = tag_name
    #     asr_train_config = os.path.join(model_dir, 'config.yaml')
    #     asr_model_file = os.path.join(model_dir, 'model.pb')
    #     cmvn_file = os.path.join(model_dir, 'am.mvn')
    #     model, asr_train_args = ASRTask.build_model_from_file(
    #         asr_train_config, asr_model_file, cmvn_file, 'cpu'
    #     )
    #     self.export(model, tag_name)

    def _export_onnx(self, model, verbose, path, enc_size=None):
        if enc_size:
@@ -130,20 +138,41 @@

        # model_script = torch.jit.script(model)
        model_script = model #torch.jit.trace(model)
        model_path = os.path.join(path, f'{model.model_name}.onnx')

        torch.onnx.export(
            model_script,
            dummy_input,
            os.path.join(path, f'{model.model_name}.onnx'),
            model_path,
            verbose=verbose,
            opset_version=12,
            opset_version=14,
            input_names=model.get_input_names(),
            output_names=model.get_output_names(),
            dynamic_axes=model.get_dynamic_axes()
        )

        if self.quant:
            from onnxruntime.quantization import QuantType, quantize_dynamic
            quant_model_path = os.path.join(path, f'{model.model_name}_quant.onnx')
            quantize_dynamic(
                model_input=model_path,
                model_output=quant_model_path,
                weight_type=QuantType.QUInt8,
            )


if __name__ == '__main__':
    output_dir = "../export"
    export_model = ASRModelExportParaformer(cache_dir=output_dir, onnx=False)
    export_model.export('damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch')
    # export_model.export('/root/cache/export/damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch')
    import sys
    
    model_path = sys.argv[1]
    output_dir = sys.argv[2]
    onnx = sys.argv[3]
    quant = sys.argv[4]
    onnx = onnx.lower()
    onnx = onnx == 'true'
    quant = quant == 'true'
    # model_path = 'damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch'
    # output_dir = "../export"
    export_model = ASRModelExportParaformer(cache_dir=output_dir, onnx=onnx, quant=quant)
    export_model.export(model_path)
    # export_model.export('/root/cache/export/damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch')