deepseek-llm-7b-chat微调报错

[lzh123415]

transformers version: 4.41.2
Python version: 3.12
System: windows

    peft_config = LoraConfig(
        task_type=TaskType.CAUSAL_LM,
        target_modules=["q_proj", "k_proj", "v_proj", "o_proj", "gate_proj", "up_proj", "down_proj"],
        inference_mode=False,  # 训练模式
        r=8,  # Lora 秩
        lora_alpha=32,  # Lora alaph，具体作用参见 Lora 原理
        lora_dropout=0.1  # Dropout 比例
    )

class ModifiedTrainer(Trainer):
def compute_loss(self, model, inputs, return_outputs=False):
# 7B
# print(model)
# print(inputs)
return model(
input_ids=inputs["input_ids"],
labels=inputs["labels"]
).loss

def save_model(self, output_dir=None, _internal_call=False):
    from transformers.trainer import TRAINING_ARGS_NAME

    os.makedirs(output_dir, exist_ok=True)
    torch.save(self.args, os.path.join(output_dir, TRAINING_ARGS_NAME))
    saved_params = {
        k: v.to("cuda:0") for k, v in self.model.named_parameters() if v.requires_grad
    }
    torch.save(saved_params, os.path.join(output_dir, "adapter_model.bin"))
    
    
    
    完整的报错信息如下：
    
    2024-06-18 11:58:33-train-INFO: 从dataset/huanhuan.json加载数据集成功

log_name: log.log
2024-06-18 11:58:33-train-INFO: 开始 LoRA 训练
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
Loading checkpoint shards: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 2/2 [00:08<00:00, 4.03s/it]
2024-06-18 11:58:41-train-INFO: 从deepseek-ai/deepseek-llm-7b-chat加载模型成功
2024-06-18 11:59:20-train-INFO: 加载 LoRA 参数成功
Found cached dataset generator (C:/Users/admin/.cache/huggingface/datasets/generator/default-d2f54e55ff33160c/0.0.0)
2024-06-18 11:59:20-train-INFO: 从dataset/huanhuan.json加载数据集成功
2024-06-18 11:59:20-train-INFO: 成功加载 Trainer
0%| | 0/1401 [00:00<?, ?it/s]Traceback (most recent call last):
File "D:\Pythonprojects\chat_test\deepseek_test2.py", line 376, in
main()
File "D:\Pythonprojects\chat_test\deepseek_test2.py", line 368, in main
trainer.train()
File "D:\Pythonprojects\chat_test\venv\Lib\site-packages\transformers\trainer.py", line 1885, in train
return inner_training_loop(
^^^^^^^^^^^^^^^^^^^^
File "D:\Pythonprojects\chat_test\venv\Lib\site-packages\transformers\trainer.py", line 2216, in _inner_training_loop
tr_loss_step = self.training_step(model, inputs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "D:\Pythonprojects\chat_test\venv\Lib\site-packages\transformers\trainer.py", line 3238, in training_step
loss = self.compute_loss(model, inputs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "D:\Pythonprojects\chat_test\deepseek_test2.py", line 156, in compute_loss
return model(
^^^^^^
File "D:\Pythonprojects\chat_test\venv\Lib\site-packages\torch\nn\modules\module.py", line 1532, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "D:\Pythonprojects\chat_test\venv\Lib\site-packages\torch\nn\modules\module.py", line 1541, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "D:\Pythonprojects\chat_test\venv\Lib\site-packages\peft\peft_model.py", line 922, in forward
return self.base_model(
^^^^^^^^^^^^^^^^
return self.base_model(
return self.base_model(
^^^^^^^^^^^^^^^^
File "D:\Pythonprojects\chat_test\venv\Lib\site-packages\torch\nn\modules\module.py", line 1532, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "D:\Pythonprojects\chat_test\venv\Lib\site-packages\torch\nn\modules\module.py", line 1541, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
TypeError: LlamaModel.forward() got an unexpected keyword argument 'labels'

[KMnO4-zx]

这个代码可以在linux或者autodl平台跑通嘛？因为看到您的环境是windows，所以可能会存在一些不稳定原因，无法准备确定问题。
可以尝试使用和教程一样的环境~