Unable to fine-tune biogpt for ner task
Hi I was using run_ner.py script It showed an error that it doesn't support model which have slow tokenizer and then they suggested that try to run old run_ner.py script which is this (https://github.com/huggingface/transformers/tree/main/examples/legacy/token-classification), Now I am getting an error
/content/transformers/examples/legacy/token-classification/run_ner.py:327 in │
│ │
│ │
│ 324 │
│ 325 │
│ 326 if name == "main": │
│ ❱ 327 │ main() │
│ 328 │
│ │
│ /content/transformers/examples/legacy/token-classification/run_ner.py:262 in │
│ main │
│ │
│ 259 │ │
│ 260 │ # Training │
│ 261 │ if training_args.do_train: │
│ ❱ 262 │ │ trainer.train( │
│ 263 │ │ │ model_path=model_args.model_name_or_path if os.path.isdir( │
│ 264 │ │ ) │
│ 265 │ │ trainer.save_model() │
│ │
│ /usr/local/lib/python3.10/dist-packages/transformers/trainer.py:1536 in │
│ train │
│ │
│ 1533 │ │ inner_training_loop = find_executable_batch_size( │
│ 1534 │ │ │ self._inner_training_loop, self._train_batch_size, args.a │
│ 1535 │ │ ) │
│ ❱ 1536 │ │ return inner_training_loop( │
│ 1537 │ │ │ args=args, │
│ 1538 │ │ │ resume_from_checkpoint=resume_from_checkpoint, │
│ 1539 │ │ │ trial=trial, │
│ │
│ /usr/local/lib/python3.10/dist-packages/transformers/trainer.py:1779 in │
│ _inner_training_loop │
│ │
│ 1776 │ │ │ │ rng_to_sync = True │
│ 1777 │ │ │ │
│ 1778 │ │ │ step = -1 │
│ ❱ 1779 │ │ │ for step, inputs in enumerate(epoch_iterator): │
│ 1780 │ │ │ │ total_batched_samples += 1 │
│ 1781 │ │ │ │ if rng_to_sync: │
│ 1782 │ │ │ │ │ self._load_rng_state(resume_from_checkpoint) │
│ │
│ /usr/local/lib/python3.10/dist-packages/accelerate/data_loader.py:377 in │
│ iter │
│ │
│ 374 │ │ dataloader_iter = super().iter() │
│ 375 │ │ # We iterate one batch ahead to check when we are at the end │
│ 376 │ │ try: │
│ ❱ 377 │ │ │ current_batch = next(dataloader_iter) │
│ 378 │ │ except StopIteration: │
│ 379 │ │ │ yield │
│ 380 │
│ │
│ /usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:633 │
│ in next │
│ │
│ 630 │ │ │ if self._sampler_iter is None: │
│ 631 │ │ │ │ # TODO(https://github.com/pytorch/pytorch/issues/7675 │
│ 632 │ │ │ │ self._reset() # type: ignore[call-arg] │
│ ❱ 633 │ │ │ data = self._next_data() │
│ 634 │ │ │ self._num_yielded += 1 │
│ 635 │ │ │ if self._dataset_kind == _DatasetKind.Iterable and \ │
│ 636 │ │ │ │ │ self._IterableDataset_len_called is not None and │
│ │
│ /usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:677 │
│ in _next_data │
│ │
│ 674 │ │
│ 675 │ def _next_data(self): │
│ 676 │ │ index = self._next_index() # may raise StopIteration │
│ ❱ 677 │ │ data = self._dataset_fetcher.fetch(index) # may raise StopIt │
│ 678 │ │ if self._pin_memory: │
│ 679 │ │ │ data = _utils.pin_memory.pin_memory(data, self._pin_memor │
│ 680 │ │ return data │
│ │
│ /usr/local/lib/python3.10/dist-packages/torch/utils/data/_utils/fetch.py:54 │
│ in fetch │
│ │
│ 51 │ │ │ │ data = [self.dataset[idx] for idx in possibly_batched_i │
│ 52 │ │ else: │
│ 53 │ │ │ data = self.dataset[possibly_batched_index] │
│ ❱ 54 │ │ return self.collate_fn(data) │
│ 55 │
│ │
│ /usr/local/lib/python3.10/dist-packages/transformers/trainer_utils.py:698 in │
│ call │
│ │
│ 695 │ │
│ 696 │ def call(self, features: List[dict]): │
│ 697 │ │ features = [self._remove_columns(feature) for feature in featu │
│ ❱ 698 │ │ return self.data_collator(features) │
│ 699 │
│ │
│ /usr/local/lib/python3.10/dist-packages/transformers/data/data_collator.py:7 │
│ 0 in default_data_collator │
│ │
│ 67 │ # on the whole batch. │
│ 68 │ │
│ 69 │ if return_tensors == "pt": │
│ ❱ 70 │ │ return torch_default_data_collator(features) │
│ 71 │ elif return_tensors == "tf": │
│ 72 │ │ return tf_default_data_collator(features) │
│ 73 │ elif return_tensors == "np": │
│ │
│ /usr/local/lib/python3.10/dist-packages/transformers/data/data_collator.py:1 │
│ 36 in torch_default_data_collator │
│ │
│ 133 │ │ │ elif isinstance(v, np.ndarray): │
│ 134 │ │ │ │ batch[k] = torch.tensor(np.stack([f[k] for f in featu │
│ 135 │ │ │ else: │
│ ❱ 136 │ │ │ │ batch[k] = torch.tensor([f[k] for f in features]) │
│ 137 │ │
│ 138 │ return batch │
│ 139 │
╰──────────────────────────────────────────────────────────────────────────────╯
RuntimeError: Could not infer dtype of NoneType