finetune後的對話模型佈署問題

[0920GX]

我執行了finetune_pt_multiturn.sh、並使用我自己的資料集finetune
之後在Terminal執行
!SET MODEL_PATH=THUDM/chatglm3-6b !SET PT_PATH=./output/kirin_v0.8-20231230-044633-128-2e-2/checkpoint-1000/ !streamlit run ./ChatGLM3/composite_demo/main.py
雖然streamlit的頁面正常啟動，但當我在對話框內輸入任意文字並送出都會得到

完整的error log如下：
== Input ==
hi

==History==
[{'role': 'system', 'content': "You are ChatGLM3, a large language model trained by Zhipu.AI. Follow the user's instructions carefully. Respond using markdown."}, {'role': 'user', 'content': 'hi'}, {'role': 'user', 'content': 'hi'}, {'role': 'user', 'content': 'hi'}]
2024-01-06 15:24:39.197 Uncaught app exception
Traceback (most recent call last):
File "C:\Users\user\anaconda3\envs\kirin\lib\site-packages\streamlit\runtime\scriptrunner\script_runner.py", line 534, in _run_script
exec(code, module.dict)
File "C:\Users\user\Desktop\ai\ChatGLM3\composite_demo\main.py", line 72, in
demo_chat.main(
File "C:\Users\user\Desktop\ai./ChatGLM3/composite_demo\demo_chat.py", line 62, in main
for response in client.generate_stream(
File "C:\Users\user\Desktop\ai./ChatGLM3/composite_demo\client.py", line 182, in generate_stream
for new_text, _ in stream_chat(
File "C:\Users\user\Desktop\ai./ChatGLM3/composite_demo\client.py", line 110, in stream_chat
for outputs in self.stream_generate(**inputs, past_key_values=past_key_values,
File "C:\Users\user\anaconda3\envs\kirin\lib\site-packages\torch\utils_contextlib.py", line 35, in generator_context
response = gen.send(None)
File "C:\Users\user.cache\huggingface\modules\transformers_modules\THUDM\chatglm3-6b\b098244a71fbe69ce149682d9072a7629f7e908c\modeling_chatglm.py", line 1159, in stream_generate
outputs = self(
File "C:\Users\user\anaconda3\envs\kirin\lib\site-packages\torch\nn\modules\module.py", line 1518, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "C:\Users\user\anaconda3\envs\kirin\lib\site-packages\torch\nn\modules\module.py", line 1527, in _call_impl
return forward_call(*args, **kwargs)
File "C:\Users\user\anaconda3\envs\kirin\lib\site-packages\accelerate\hooks.py", line 165, in new_forward
output = module._old_forward(*args, **kwargs)
File "C:\Users\user.cache\huggingface\modules\transformers_modules\THUDM\chatglm3-6b\b098244a71fbe69ce149682d9072a7629f7e908c\modeling_chatglm.py", line 937, in forward
transformer_outputs = self.transformer(
File "C:\Users\user\anaconda3\envs\kirin\lib\site-packages\torch\nn\modules\module.py", line 1518, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "C:\Users\user\anaconda3\envs\kirin\lib\site-packages\torch\nn\modules\module.py", line 1527, in _call_impl
return forward_call(*args, **kwargs)
File "C:\Users\user.cache\huggingface\modules\transformers_modules\THUDM\chatglm3-6b\b098244a71fbe69ce149682d9072a7629f7e908c\modeling_chatglm.py", line 811, in forward
past_key_values = self.get_prompt(batch_size=batch_size, device=input_ids.device,
File "C:\Users\user.cache\huggingface\modules\transformers_modules\THUDM\chatglm3-6b\b098244a71fbe69ce149682d9072a7629f7e908c\modeling_chatglm.py", line 773, in get_prompt
past_key_values = self.prefix_encoder(prefix_tokens).type(dtype)
File "C:\Users\user\anaconda3\envs\kirin\lib\site-packages\torch\nn\modules\module.py", line 1518, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "C:\Users\user\anaconda3\envs\kirin\lib\site-packages\torch\nn\modules\module.py", line 1527, in _call_impl
return forward_call(*args, **kwargs)
File "C:\Users\user\anaconda3\envs\kirin\lib\site-packages\accelerate\hooks.py", line 165, in new_forward
output = module._old_forward(*args, **kwargs)
File "C:\Users\user.cache\huggingface\modules\transformers_modules\THUDM\chatglm3-6b\b098244a71fbe69ce149682d9072a7629f7e908c\modeling_chatglm.py", line 89, in forward
past_key_values = self.embedding(prefix)
File "C:\Users\user\anaconda3\envs\kirin\lib\site-packages\torch\nn\modules\module.py", line 1518, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "C:\Users\user\anaconda3\envs\kirin\lib\site-packages\torch\nn\modules\module.py", line 1527, in _call_impl
return forward_call(*args, **kwargs)
File "C:\Users\user\anaconda3\envs\kirin\lib\site-packages\accelerate\hooks.py", line 160, in new_forward
args, kwargs = module._hf_hook.pre_forward(module, *args, **kwargs)
File "C:\Users\user\anaconda3\envs\kirin\lib\site-packages\accelerate\hooks.py", line 294, in pre_forward
module, name, self.execution_device, value=self.weights_map[name], fp16_statistics=fp16_statistics
File "C:\Users\user\anaconda3\envs\kirin\lib\site-packages\accelerate\utils\offload.py", line 118, in getitem
return self.dataset[f"{self.prefix}{key}"]
File "C:\Users\user\anaconda3\envs\kirin\lib\site-packages\accelerate\utils\offload.py", line 165, in getitem
weight_info = self.index[key]
KeyError: 'transformer.prefix_encoder.embedding.weight'

想請問各位大佬該怎麼解決這個問題