OverflowError: cannot fit 'int' into an index-sized integer

[Tabz05]

This is the stacktrace of the issue:

Num processes: 1
Process index: 0
Local process index: 0
Device: cuda

Mixed precision type: fp16

{'resnet_time_scale_shift', 'dual_cross_attention', 'upcast_attention', 'class_embed_type', 'num_class_embeds', 'only_cross_attention', 'use_linear_projection'} was not found in config. Values will be initialized to default values.
The tokenizer class you load from this checkpoint is not the same type as the class this function is called from. It may result in unexpected tokenization.
The tokenizer class you load from this checkpoint is 'LEDTokenizer'.
The class this function is called from is 'CLIPTokenizer'.
{'prediction_type'} was not found in config. Values will be initialized to default values.
{'config', 'disk_store'} was not found in config. Values will be initialized to default values.
{'config', 'disk_store'} was not found in config. Values will be initialized to default values.
╭─────────────────────────────── Traceback (most recent call last) ────────────────────────────────╮
│ /content/drive/MyDrive/Free-Bloom-main/Free-Bloom-main/main.py:183 in │
│ │
│ 180 │ args = parser.parse_args() │
│ 181 │ │
│ 182 │ conf = OmegaConf.load(args.config) │
│ ❱ 183 │ main(**conf) │
│ 184 │
│ │
│ /content/drive/MyDrive/Free-Bloom-main/Free-Bloom-main/main.py:160 in main │
│ │
│ 157 │ │ │
│ 158 │ │ validation_data.pop('negative_prompt') │
│ 159 │ │ # key frame │
│ ❱ 160 │ │ key_frames, text_embedding = validation_pipeline(prompt, video_length=len(prompt │
│ 161 │ │ │ │ │ │ │ │ │ │ │ │ │ │ latents=x_T.type(weight_dtype), │
│ 162 │ │ │ │ │ │ │ │ │ │ │ │ │ │ negative_prompt=negative_prompt │
│ 163 │ │ │ │ │ │ │ │ │ │ │ │ │ │ output_dir=output_dir, │
│ │
│ /usr/local/lib/python3.10/dist-packages/torch/autograd/grad_mode.py:27 in decorate_context │
│ │
│ 24 │ │ @functools.wraps(func) │
│ 25 │ │ def decorate_context(*args, **kwargs): │
│ 26 │ │ │ with self.clone(): │
│ ❱ 27 │ │ │ │ return func(*args, **kwargs) │
│ 28 │ │ return cast(F, decorate_context) │
│ 29 │ │
│ 30 │ def _wrap_generator(self, func): │
│ │
│ /content/drive/MyDrive/Free-Bloom-main/Free-Bloom-main/freebloom/pipelines/pipeline_spatio_tempo │
│ ral.py:391 in call │
│ │
│ 388 │ │ do_classifier_free_guidance = guidance_scale > 1.0 │
│ 389 │ │ │
│ 390 │ │ # Encode input prompt │
│ ❱ 391 │ │ text_embeddings = self._encode_prompt( │
│ 392 │ │ │ prompt, device, num_videos_per_prompt, do_classifier_free_guidance, negative │
│ 393 │ │ ) │
│ 394 │
│ │
│ /content/drive/MyDrive/Free-Bloom-main/Free-Bloom-main/freebloom/pipelines/pipeline_spatio_tempo │
│ ral.py:164 in _encode_prompt │
│ │
│ 161 │ def _encode_prompt(self, prompt, device, num_videos_per_prompt, do_classifier_free_g │
│ 162 │ │ batch_size = len(prompt) if isinstance(prompt, list) else 1 │
│ 163 │ │ │
│ ❱ 164 │ │ text_inputs = self.tokenizer( │
│ 165 │ │ │ prompt, │
│ 166 │ │ │ padding="max_length", │
│ 167 │ │ │ max_length=self.tokenizer.model_max_length, │
│ │
│ /usr/local/lib/python3.10/dist-packages/transformers/tokenization_utils_base.py:2530 in call │
│ │
│ 2527 │ │ │ # input mode in this case. │
│ 2528 │ │ │ if not self._in_target_context_manager: │
│ 2529 │ │ │ │ self._switch_to_input_mode() │
│ ❱ 2530 │ │ │ encodings = self._call_one(text=text, text_pair=text_pair, **all_kwargs) │
│ 2531 │ │ if text_target is not None: │
│ 2532 │ │ │ self._switch_to_target_mode() │
│ 2533 │ │ │ target_encodings = self._call_one(text=text_target, text_pair=text_pair_targ │
│ │
│ /usr/local/lib/python3.10/dist-packages/transformers/tokenization_utils_base.py:2616 in │
│ _call_one │
│ │
│ 2613 │ │ │ │ │ f" {len(text_pair)}." │
│ 2614 │ │ │ │ ) │
│ 2615 │ │ │ batch_text_or_text_pairs = list(zip(text, text_pair)) if text_pair is not No │
│ ❱ 2616 │ │ │ return self.batch_encode_plus( │
│ 2617 │ │ │ │ batch_text_or_text_pairs=batch_text_or_text_pairs, │
│ 2618 │ │ │ │ add_special_tokens=add_special_tokens, │
│ 2619 │ │ │ │ padding=padding, │
│ │
│ /usr/local/lib/python3.10/dist-packages/transformers/tokenization_utils_base.py:2807 in │
│ batch_encode_plus │
│ │
│ 2804 │ │ │ **kwargs, │
│ 2805 │ │ ) │
│ 2806 │ │ │
│ ❱ 2807 │ │ return self._batch_encode_plus( │
│ 2808 │ │ │ batch_text_or_text_pairs=batch_text_or_text_pairs, │
│ 2809 │ │ │ add_special_tokens=add_special_tokens, │
│ 2810 │ │ │ padding_strategy=padding_strategy, │
│ │
│ /usr/local/lib/python3.10/dist-packages/transformers/tokenization_utils.py:737 in │
│ _batch_encode_plus │
│ │
│ 734 │ │ │ second_ids = get_input_ids(pair_ids) if pair_ids is not None else None │
│ 735 │ │ │ input_ids.append((first_ids, second_ids)) │
│ 736 │ │ │
│ ❱ 737 │ │ batch_outputs = self._batch_prepare_for_model( │
│ 738 │ │ │ input_ids, │
│ 739 │ │ │ add_special_tokens=add_special_tokens, │
│ 740 │ │ │ padding_strategy=padding_strategy, │
│ │
│ /usr/local/lib/python3.10/dist-packages/transformers/tokenization_utils.py:809 in │
│ _batch_prepare_for_model │
│ │
│ 806 │ │ │ │ │ batch_outputs[key] = [] │
│ 807 │ │ │ │ batch_outputs[key].append(value) │
│ 808 │ │ │
│ ❱ 809 │ │ batch_outputs = self.pad( │
│ 810 │ │ │ batch_outputs, │
│ 811 │ │ │ padding=padding_strategy.value, │
│ 812 │ │ │ max_length=max_length, │
│ │
│ /usr/local/lib/python3.10/dist-packages/transformers/tokenization_utils_base.py:3014 in pad │
│ │
│ 3011 │ │ batch_outputs = {} │
│ 3012 │ │ for i in range(batch_size): │
│ 3013 │ │ │ inputs = {k: v[i] for k, v in encoded_inputs.items()} │
│ ❱ 3014 │ │ │ outputs = self._pad( │
│ 3015 │ │ │ │ inputs, │
│ 3016 │ │ │ │ max_length=max_length, │
│ 3017 │ │ │ │ padding_strategy=padding_strategy, │
│ │
│ /usr/local/lib/python3.10/dist-packages/transformers/tokenization_utils_base.py:3378 in _pad │
│ │
│ 3375 │ │ │ │
│ 3376 │ │ │ if self.padding_side == "right": │
│ 3377 │ │ │ │ if return_attention_mask: │
│ ❱ 3378 │ │ │ │ │ encoded_inputs["attention_mask"] = encoded_inputs["attention_mask"] │
│ 3379 │ │ │ │ if "token_type_ids" in encoded_inputs: │
│ 3380 │ │ │ │ │ encoded_inputs["token_type_ids"] = ( │
│ 3381 │ │ │ │ │ │ encoded_inputs["token_type_ids"] + [self.pad_token_type_id] * di │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
OverflowError: cannot fit 'int' into an index-sized integer

Kindly provide a solution.

[JudgementH]

Hi!
I sincerely apologize for not noticing this message sooner due to the way GitHub notifications work.

It seems that the error occurs within the transformers library. Please note that transformers has had significant updates. Could you make sure you are using transformers==4.27.1, as this is the version our codebase currently supports?