使用 Hugging Face 的 Transformers 库加载预训练模型遇到的问题

题意：

Size mismatch for embed_out.weight: copying a param with shape torch.Size([0]) from checkpoint - Huggingface PyTorch

这个错误信息 "Size mismatch for embed_out.weight: copying a param with shape torch.Size([0]) from checkpoint - Huggingface PyTorch" 通常出现在使用 Hugging Face 的 Transformers 库加载预训练模型时，模型的某些参数与预训练模型检查点（checkpoint）中的参数形状不匹配。

问题背景：

I want to finetune an LLM. I am able to successfully finetune LLM. But when reload the model after save, gets error. Below is the code

import argparse
import numpy as np
import torch
from datasets import load_dataset
from transformers import AutoTokenizer, AutoModelForCausalLMfrom trl import DPOTrainer, DPOConfig
def preprocess_data(item):return {'prompt': 'Instruct: ' + item['prompt'] + '\n','chosen': 'Output: ' + item['chosen'],'rejected': 'Output: ' + item['rejected']}        def main():parser = argparse.ArgumentParser()parser.add_argument("--epochs", type=int, default=1)parser.add_argument("--beta", type=float, default=0.1)parser.add_argument("--batch_size", type=int, default=4)parser.add_argument("--lr", type=float, default=1e-6)parser.add_argument("--seed", type=int, default=2003)parser.add_argument("--model_name", type=str, default="EleutherAI/pythia-14m")parser.add_argument("--dataset_name", type=str, default="jondurbin/truthy-dpo-v0.1")parser.add_argument("--local_rank", type=int, default=0)args = parser.parse_args()# Determine device based on local_rankdevice = torch.device("cuda", args.local_rank) if torch.cuda.is_available() else torch.device("cpu")tokenizer = AutoTokenizer.from_pretrained(args.model_name)tokenizer.pad_token = tokenizer.eos_tokenmodel = AutoModelForCausalLM.from_pretrained(args.model_name).to(device)ref_model = AutoModelForCausalLM.from_pretrained(args.model_name).to(device)dataset = load_dataset(args.dataset_name, split="train")dataset = dataset.map(preprocess_data)# Split the dataset into training and validation setsdataset = dataset.train_test_split(test_size=0.1, seed=args.seed)train_dataset = dataset['train']val_dataset = dataset['test']training_args = DPOConfig(learning_rate=args.lr,num_train_epochs=args.epochs,per_device_train_batch_size=args.batch_size,logging_steps=10,remove_unused_columns=False,max_length=1024,max_prompt_length=512,fp16=True        )# Verify and print embedding dimensions before finetuningprint("Base model embedding dimension:", model.config.hidden_size)model.train()ref_model.eval()dpo_trainer = DPOTrainer(model,ref_model,beta=args.beta,train_dataset=train_dataset,eval_dataset=val_dataset,tokenizer=tokenizer,args=training_args,)dpo_trainer.train()# Evaluateevaluation_results = dpo_trainer.evaluate()print("Evaluation Results:", evaluation_results)save_model_name = 'finetuned_model'model.save_pretrained(save_model_name)if __name__ == "__main__":main()

Error I was getting as below

return model_class.from_pretrained(File "/.local/lib/python3.10/site-packages/transformers/modeling_utils.py", line 3838, in from_pretrained) = cls._load_pretrained_model(File "/.local/lib/python3.10/site-packages/transformers/modeling_utils.py", line 4349, in _load_pretrained_modelraise RuntimeError(f"Error(s) in loading state_dict for {model.__class__.__name__}:\n\t{error_msg}")RuntimeError: Error(s) in loading state_dict for GPTNeoXForCausalLM:size mismatch for gpt_neox.embed_in.weight: copying a param with shape torch.Size([0]) from checkpoint, the shape in current model is torch.Size([50304, 128]).size mismatch for embed_out.weight: copying a param with shape torch.Size([0]) from checkpoint, the shape in current model is torch.Size([50304, 128]).You may consider adding `ignore_mismatched_sizes=True` in the model `from_pretrained` method.

After finetuning, model works perfectly. But after reloading the saved trained model its not working. Any idea why gets this error when reloading the model ?