Pytorch 索引错误（'index out of range in self'）：如何解决？

Question

我最近在 deep learning tutorial on youtube (entire code can be found here) 之后遇到了障碍。我在第 4.4 部分遇到问题。目标是 return 某些股票的文章摘要字典（它们的代码在列表中：monitered_tickers）。

def summarize(articles):
summaries = []
for article in articles:
    input_ids = tokenizer.encode(article, return_tensors='pt')
    output = model.generate(input_ids, max_length=55, num_beams=5, early_stopping=True)
    summary = tokenizer.decode(output[0], skip_special_tokens=True)
    summaries.append(summary)
return summaries

summaries = {ticker:summarize(articles[ticker]) for ticker in monitered_tickers}
summaries

当我运行上面的代码时，我得到以下错误：

    ---------------------------------------------------------------------------
IndexError                                Traceback (most recent call last)
~\AppData\Local\Temp/ipykernel_10688/3329134555.py in <module>
----> 1 summaries = {ticker:summarize(articles[ticker]) for ticker in monitered_tickers}
      2 summaries

~\AppData\Local\Temp/ipykernel_10688/3329134555.py in <dictcomp>(.0)
----> 1 summaries = {ticker:summarize(articles[ticker]) for ticker in monitered_tickers}
      2 summaries


~\AppData\Local\Temp/ipykernel_10688/3177436574.py in summarize(articles)
      3     for article in articles:
      4         input_ids = tokenizer.encode(article, return_tensors='pt')
----> 5         output = model.generate(input_ids, max_length=40, num_beams=5, early_stopping = True)
      6         summary = tokenizer.decode(output[0], skip_special_tokens=True)
      7         summaries.append(summary)

~\anaconda3\lib\site-packages\torch\autograd\grad_mode.py in decorate_context(*args, **kwargs)
     26         def decorate_context(*args, **kwargs):
     27             with self.__class__():
---> 28                 return func(*args, **kwargs)
     29         return cast(F, decorate_context)
     30 

~\anaconda3\lib\site-packages\transformers\generation_utils.py in generate(self, inputs, max_length, min_length, do_sample, early_stopping, num_beams, temperature, top_k, top_p, repetition_penalty, bad_words_ids, bos_token_id, pad_token_id, eos_token_id, length_penalty, no_repeat_ngram_size, encoder_no_repeat_ngram_size, num_return_sequences, max_time, max_new_tokens, decoder_start_token_id, use_cache, num_beam_groups, diversity_penalty, prefix_allowed_tokens_fn, logits_processor, stopping_criteria, output_attentions, output_hidden_states, output_scores, return_dict_in_generate, forced_bos_token_id, forced_eos_token_id, remove_invalid_values, synced_gpus, **model_kwargs)
   1022             # if model is encoder decoder encoder_outputs are created
   1023             # and added to `model_kwargs`
-> 1024             model_kwargs = self._prepare_encoder_decoder_kwargs_for_generation(
   1025                 inputs_tensor, model_kwargs, model_input_name
   1026             )

~\anaconda3\lib\site-packages\transformers\generation_utils.py in _prepare_encoder_decoder_kwargs_for_generation(self, inputs_tensor, model_kwargs, model_input_name)
    484                 encoder_args = ()
    485 
--> 486             model_kwargs["encoder_outputs"]: ModelOutput = encoder(*encoder_args, **encoder_kwargs)
    487 
    488         return model_kwargs

~\anaconda3\lib\site-packages\torch\nn\modules\module.py in _call_impl(self, *input, **kwargs)
   1100         if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks
   1101                 or _global_forward_hooks or _global_forward_pre_hooks):
-> 1102             return forward_call(*input, **kwargs)
   1103         # Do not call functions when jit is used
   1104         full_backward_hooks, non_full_backward_hooks = [], []

~\anaconda3\lib\site-packages\transformers\models\pegasus\modeling_pegasus.py in forward(self, input_ids, attention_mask, head_mask, inputs_embeds, output_attentions, output_hidden_states, return_dict)
    754             inputs_embeds = self.embed_tokens(input_ids) * self.embed_scale
    755 
--> 756         embed_pos = self.embed_positions(input_shape)
    757 
    758         hidden_states = inputs_embeds + embed_pos

~\anaconda3\lib\site-packages\torch\nn\modules\module.py in _call_impl(self, *input, **kwargs)
   1100         if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks
   1101                 or _global_forward_hooks or _global_forward_pre_hooks):
-> 1102             return forward_call(*input, **kwargs)
   1103         # Do not call functions when jit is used
   1104         full_backward_hooks, non_full_backward_hooks = [], []

~\anaconda3\lib\site-packages\torch\autograd\grad_mode.py in decorate_context(*args, **kwargs)
     26         def decorate_context(*args, **kwargs):
     27             with self.__class__():
---> 28                 return func(*args, **kwargs)
     29         return cast(F, decorate_context)
     30 

~\anaconda3\lib\site-packages\transformers\models\pegasus\modeling_pegasus.py in forward(self, input_ids_shape, past_key_values_length)
    138             past_key_values_length, past_key_values_length + seq_len, dtype=torch.long, device=self.weight.device
    139         )
--> 140         return super().forward(positions)
    141 
    142 

~\anaconda3\lib\site-packages\torch\nn\modules\sparse.py in forward(self, input)
    156 
    157     def forward(self, input: Tensor) -> Tensor:
--> 158         return F.embedding(
    159             input, self.weight, self.padding_idx, self.max_norm,
    160             self.norm_type, self.scale_grad_by_freq, self.sparse)

~\anaconda3\lib\site-packages\torch\nn\functional.py in embedding(input, weight, padding_idx, max_norm, norm_type, scale_grad_by_freq, sparse)
   2042         # remove once script supports set_grad_enabled
   2043         _no_grad_embedding_renorm_(weight, input, max_norm, norm_type)
-> 2044     return torch.embedding(weight, input, padding_idx, scale_grad_by_freq, sparse)
   2045 
   2046 

IndexError: index out of range in self

想知道如何解决这个问题。有点新手，所以对此的任何帮助将不胜感激。谢谢。

Answer 1

您的文章长度可能超过了模型的最大输入长度。使用：

tokenizer.encode(article, return_tensors='pt', max_length=512, truncation=True)

Pytorch 索引错误（'index out of range in self'）：如何解决？

Pytorch Index Error ('index out of range in self'): How to Solve?

python

pytorch