| | |
|---|
| | | |
|---|
| | | # export encoder1 |
|---|
| | | self.export_config["model_name"] = "model" |
|---|
| | | model = get_model( |
|---|
| | | models = get_model( |
|---|
| | | model, |
|---|
| | | self.export_config, |
|---|
| | | ) |
|---|
| | | 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) |
|---|
| | | |
|---|
| | | print("output dir: {}".format(export_dir)) |
|---|
| | | if not isinstance(models, tuple): |
|---|
| | | models = (models,) |
|---|
| | | |
|---|
| | | for i, model in enumerate(models): |
|---|
| | | model.eval() |
|---|
| | | if self.onnx: |
|---|
| | | self._export_onnx(model, verbose, export_dir) |
|---|
| | | else: |
|---|
| | | self._export_torchscripts(model, verbose, export_dir) |
|---|
| | | |
|---|
| | | print("output dir: {}".format(export_dir)) |
|---|
| | | |
|---|
| | | |
|---|
| | | def _torch_quantize(self, model): |
|---|
