forked from datawhalechina/self-llm
-
Notifications
You must be signed in to change notification settings - Fork 2
/
04-Qwen-7B-Chat Lora 微调.py
77 lines (69 loc) · 3.38 KB
/
04-Qwen-7B-Chat Lora 微调.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
from datasets import Dataset
import pandas as pd
from transformers import AutoTokenizer, AutoModelForCausalLM, DataCollatorForSeq2Seq, TrainingArguments, Trainer
import torch
from peft import LoraConfig, TaskType, get_peft_model
# 用于处理数据集的函数
def process_func(example):
MAX_LENGTH = 384 # Llama分词器会将一个中文字切分为多个token,因此需要放开一些最大长度,保证数据的完整性
input_ids, attention_mask, labels = [], [], []
instruction = tokenizer("\n".join(["<|im_start|>system", "现在你要扮演皇帝身边的女人--甄嬛.<|im_end|>" + "\n<|im_start|>user\n" + example["instruction"] + example["input"] + "<|im_end|>\n"]).strip(), add_special_tokens=False) # add_special_tokens 不在开头加 special_tokens
response = tokenizer("<|im_start|>assistant\n" + example["output"] + "<|im_end|>\n", add_special_tokens=False)
input_ids = instruction["input_ids"] + response["input_ids"] + [tokenizer.pad_token_id]
attention_mask = instruction["attention_mask"] + response["attention_mask"] + [1] # 因为eos token咱们也是要关注的所以 补充为1
labels = [-100] * len(instruction["input_ids"]) + response["input_ids"] + [tokenizer.pad_token_id] # Qwen的特殊构造就是这样的
if len(input_ids) > MAX_LENGTH: # 做一个截断
input_ids = input_ids[:MAX_LENGTH]
attention_mask = attention_mask[:MAX_LENGTH]
labels = labels[:MAX_LENGTH]
return {
"input_ids": input_ids,
"attention_mask": attention_mask,
"labels": labels
}
# loraConfig
config = LoraConfig(
task_type=TaskType.CAUSAL_LM,
target_modules=["c_attn", "c_proj", "w1", "w2"], # 这个不同的模型需要设置不同的参数,需要看模型中的attention层
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
)
# 配置训练参数
args = TrainingArguments(
output_dir="./output/Qwen",
per_device_train_batch_size=8,
gradient_accumulation_steps=2,
logging_steps=10,
num_train_epochs=3,
gradient_checkpointing=True,
save_steps=100,
learning_rate=1e-4,
save_on_each_node=True
)
if "__main__" == __name__:
# 处理数据集
# 将JSON文件转换为CSV文件
df = pd.read_json('./data/huanhuan.json')
ds = Dataset.from_pandas(df)
# 加载tokenizer
tokenizer = AutoTokenizer.from_pretrained('./qwen/Qwen-7B-Chat/', use_fast=False, trust_remote_code=True)
tokenizer.pad_token_id = tokenizer.eod_id
# 将数据集变化为token形式
tokenized_id = ds.map(process_func, remove_columns=ds.column_names)
# 创建模型并以半精度形式加载
model = AutoModelForCausalLM.from_pretrained('./qwen/Qwen-7B-Chat/', trust_remote_code=True, torch_dtype=torch.half, device_map="auto")
model.enable_input_require_grads() # 开启梯度检查点时,要执行该方法
# 加载lora参数
model = get_peft_model(model, config)
# 使用trainer训练
trainer = Trainer(
model=model,
args=args,
train_dataset=tokenized_id,
data_collator=DataCollatorForSeq2Seq(tokenizer=tokenizer, padding=True),
)
trainer.train() # 开始训练
response, history = model.chat(tokenizer, "你是谁", history=[], system="现在你要扮演皇帝身边的女人--甄嬛.")
print(response)