百川2运行demo报错：AttributeError: 'BaichuanModel' object has no attribute 'future_mask'

[Yu-Yang-Li]

修改配置后，在demo_qwen_agent.ipynb之前一切正常，但在运行：

重置对话，清空对话历史
agent.reset()
agent.run('调用插件，查询北京市明天的天气', remote=False)

报错：
File ~/modelscope-agent/demo/../modelscope_agent/agent.py:143, in AgentExecutor.run(self, task, remote, print_info)
141 # generate prompt and call llm
142 prompt = self.prompt_generator.generate(llm_result, exec_result)
--> 143 llm_result = self.llm.generate(prompt)
144 if print_info:
145 print(f'|prompt{idx}: {prompt}')

File ~/modelscope-agent/demo/../modelscope_agent/llm/local_llm.py:72, in LocalLLM.generate(self, prompt)
69 response = self.model.chat(
70 self.tokenizer, prompt, history=[], system='')[0]
71 else:
---> 72 response = self.chat(prompt)
74 end_idx = response.find(self.end_token)
75 if end_idx != -1:

File ~/modelscope-agent/demo/../modelscope_agent/llm/local_llm.py:95, in LocalLLM.chat(self, prompt)
91 input_ids = self.tokenizer(
92 prompt, return_tensors='pt').input_ids.to(device)
93 input_len = input_ids.shape[1]
---> 95 result = self.model.generate(
96 input_ids=input_ids, generation_config=self.generation_cfg)
98 result = result[0].tolist()[input_len:]
99 response = self.tokenizer.decode(result)

File ~/.local/lib/python3.10/site-packages/torch/utils/_contextlib.py:115, in context_decorator..decorate_context(*args, **kwargs)
112 @functools.wraps(func)
113 def decorate_context(*args, **kwargs):
114 with ctx_factory():
--> 115 return func(*args, **kwargs)

File ~/.local/lib/python3.10/site-packages/transformers/generation/utils.py:1648, in GenerationMixin.generate(self, inputs, generation_config, logits_processor, stopping_criteria, prefix_allowed_tokens_fn, synced_gpus, assistant_model, streamer, negative_prompt_ids, negative_prompt_attention_mask, **kwargs)
1640 input_ids, model_kwargs = self._expand_inputs_for_generation(
1641 input_ids=input_ids,
1642 expand_size=generation_config.num_return_sequences,
1643 is_encoder_decoder=self.config.is_encoder_decoder,
1644 **model_kwargs,
1645 )
1647 # 13. run sample
-> 1648 return self.sample(
1649 input_ids,
1650 logits_processor=logits_processor,
1651 logits_warper=logits_warper,
1652 stopping_criteria=stopping_criteria,
1653 pad_token_id=generation_config.pad_token_id,
1654 eos_token_id=generation_config.eos_token_id,
1655 output_scores=generation_config.output_scores,
1656 return_dict_in_generate=generation_config.return_dict_in_generate,
1657 synced_gpus=synced_gpus,
1658 streamer=streamer,
1659 **model_kwargs,
1660 )
1662 elif generation_mode == GenerationMode.BEAM_SEARCH:
1663 # 11. prepare beam search scorer
1664 beam_scorer = BeamSearchScorer(
1665 batch_size=batch_size,
1666 num_beams=generation_config.num_beams,
(...)
1671 max_length=generation_config.max_length,
1672 )

File ~/.local/lib/python3.10/site-packages/transformers/generation/utils.py:2730, in GenerationMixin.sample(self, input_ids, logits_processor, stopping_criteria, logits_warper, max_length, pad_token_id, eos_token_id, output_attentions, output_hidden_states, output_scores, return_dict_in_generate, synced_gpus, streamer, **model_kwargs)
2727 model_inputs = self.prepare_inputs_for_generation(input_ids, **model_kwargs)
2729 # forward pass to get next token
-> 2730 outputs = self(
2731 **model_inputs,
2732 return_dict=True,
2733 output_attentions=output_attentions,
2734 output_hidden_states=output_hidden_states,
2735 )
2737 if synced_gpus and this_peer_finished:
2738 continue # don't waste resources running the code we don't need

File ~/.local/lib/python3.10/site-packages/torch/nn/modules/module.py:1501, in Module._call_impl(self, *args, **kwargs)
1496 # If we don't have any hooks, we want to skip the rest of the logic in
1497 # this function, and just call forward.
1498 if not (self._backward_hooks or self._backward_pre_hooks or self._forward_hooks or self._forward_pre_hooks
1499 or _global_backward_pre_hooks or _global_backward_hooks
1500 or _global_forward_hooks or _global_forward_pre_hooks):
-> 1501 return forward_call(*args, **kwargs)
1502 # Do not call functions when jit is used
1503 full_backward_hooks, non_full_backward_hooks = [], []

File ~/.conda/envs/LLM/lib/python3.10/site-packages/accelerate/hooks.py:165, in add_hook_to_module..new_forward(*args, **kwargs)
163 output = old_forward(*args, **kwargs)
164 else:
--> 165 output = old_forward(*args, **kwargs)
166 return module._hf_hook.post_forward(module, output)

File ~/.cache/huggingface/modules/transformers_modules/Baichuan2-13B-Base/modeling_baichuan.py:688, in BaichuanForCausalLM.forward(self, input_ids, attention_mask, past_key_values, inputs_embeds, labels, use_cache, output_attentions, output_hidden_states, return_dict, **kwargs)
683 return_dict = (
684 return_dict if return_dict is not None else self.config.use_return_dict
685 )
687 # decoder outputs consists of (dec_features, layer_state, dec_hidden, dec_attn)
--> 688 outputs = self.model(
689 input_ids=input_ids,
690 attention_mask=attention_mask,
691 past_key_values=past_key_values,
692 inputs_embeds=inputs_embeds,
693 use_cache=use_cache,
694 output_attentions=output_attentions,
695 output_hidden_states=output_hidden_states,
696 return_dict=return_dict,
697 )
699 hidden_states = outputs[0]
700 logits = self.lm_head(hidden_states)

File ~/.local/lib/python3.10/site-packages/torch/nn/modules/module.py:1501, in Module._call_impl(self, *args, **kwargs)
1496 # If we don't have any hooks, we want to skip the rest of the logic in
1497 # this function, and just call forward.
1498 if not (self._backward_hooks or self._backward_pre_hooks or self._forward_hooks or self._forward_pre_hooks
1499 or _global_backward_pre_hooks or _global_backward_hooks
1500 or _global_forward_hooks or _global_forward_pre_hooks):
-> 1501 return forward_call(*args, **kwargs)
1502 # Do not call functions when jit is used
1503 full_backward_hooks, non_full_backward_hooks = [], []

File ~/.cache/huggingface/modules/transformers_modules/Baichuan2-13B-Base/modeling_baichuan.py:401, in BaichuanModel.forward(self, input_ids, attention_mask, past_key_values, inputs_embeds, use_cache, output_attentions, output_hidden_states, return_dict)
399 alibi_mask = self.alibi_mask
400 else:
--> 401 alibi_mask = self.get_alibi_mask(inputs_embeds, seq_length_with_past)
403 if attention_mask is not None:
404 if len(attention_mask.shape) == 2:

File ~/.cache/huggingface/modules/transformers_modules/Baichuan2-13B-Base/modeling_baichuan.py:351, in BaichuanModel.get_alibi_mask(self, tensor, seq_length_with_past)
343 self.max_cache_pos = seq_length_with_past
344 self.register_buffer(
345 "future_mask",
346 _gen_alibi_mask(tensor, self.n_head, self.max_cache_pos).to(
(...)
349 persistent=False,
350 )
--> 351 mask = self.future_mask[
352 : self.n_head, :seq_length_with_past, :seq_length_with_past
353 ]
354 return mask

File ~/.local/lib/python3.10/site-packages/torch/nn/modules/module.py:1614, in Module.getattr(self, name)
1612 if name in modules:
1613 return modules[name]
-> 1614 raise AttributeError("'{}' object has no attribute '{}'".format(
1615 type(self).name, name))

AttributeError: 'BaichuanModel' object has no attribute 'future_mask'

[Yu-Yang-Li]

model_name = 'Baichuan2-13B-Base'
模型配置：
model_cfg = {
'Baichuan2-13B-Base':{
'model_id': 'baichuan-inc/Baichuan2-13B-Base',
'model_revision': 'v1.0.1',
'use_raw_generation_config': False,
'custom_chat': False,
'generate_cfg':{
'do_sample': True,
'max_new_tokens': 512,
'top_k': 10
}
}
}

[Zhikaiiii]

model_name = 'Baichuan2-13B-Base' 模型配置： model_cfg = { 'Baichuan2-13B-Base':{ 'model_id': 'baichuan-inc/Baichuan2-13B-Base', 'model_revision': 'v1.0.1', 'use_raw_generation_config': False, 'custom_chat': False, 'generate_cfg':{ 'do_sample': True, 'max_new_tokens': 512, 'top_k': 10 } } }

我这边用这个配置是可以跑通的，我的modelscope版本是1.8.4，请问你的是多少呢？