| | |
|---|
| | | |
|---|
| | | def _forward( |
|---|
| | | data_path_and_name_and_type: Sequence[Tuple[str, str, str]] = None, |
|---|
| | | raw_inputs: List[List[Union[np.ndarray, torch.Tensor, str]]] = None, |
|---|
| | | raw_inputs: List[List[Union[np.ndarray, torch.Tensor, str, bytes]]] = None, |
|---|
| | | output_dir_v2: Optional[str] = None, |
|---|
| | | param_dict: Optional[dict] = None, |
|---|
| | | ): |
|---|
| | |
|---|
| | | if isinstance(raw_inputs, (list, tuple)): |
|---|
| | | assert all([len(example) >= 2 for example in raw_inputs]), \ |
|---|
| | | "The length of test case in raw_inputs must larger than 1 (>=2)." |
|---|
| | | if not isinstance(raw_inputs, List): |
|---|
| | | raw_inputs = [raw_inputs] |
|---|
| | | |
|---|
| | | def prepare_dataset(): |
|---|
| | | for idx, example in enumerate(raw_inputs): |
|---|
