68 lines
1.8 KiB
Python
68 lines
1.8 KiB
Python
import torch
|
|
from transformers import AutoModelForCausalLM, AutoTokenizer
|
|
|
|
# 使用你本地的检查点路径
|
|
model_path = "/root/Qwen2.5-7B-Instruct-R1-forfinance/"
|
|
|
|
# 加载模型和分词器
|
|
print("正在加载模型...")
|
|
model = AutoModelForCausalLM.from_pretrained(
|
|
model_path,
|
|
torch_dtype=torch.bfloat16, # 根据config.json中的torch_dtype
|
|
device_map="auto",
|
|
trust_remote_code=True # 如果需要的话
|
|
)
|
|
|
|
tokenizer = AutoTokenizer.from_pretrained(
|
|
model_path,
|
|
trust_remote_code=True
|
|
)
|
|
|
|
print("模型加载完成！")
|
|
|
|
# 准备输入
|
|
prompt = "假设你是一位金融行业专家，请回答下列问题。\n在宏观分析中，描述在既定利率水平下产品市场达到均衡状态的曲线是什么？\n请一步步思考。"
|
|
|
|
messages = [
|
|
{"role": "system", "content": "You are Qwen, created by Alibaba Cloud. You are a helpful assistant."},
|
|
{"role": "user", "content": prompt}
|
|
]
|
|
|
|
# 应用聊天模板
|
|
text = tokenizer.apply_chat_template(
|
|
messages,
|
|
tokenize=False,
|
|
add_generation_prompt=True
|
|
)
|
|
|
|
print("输入文本:")
|
|
print(text)
|
|
print("\n" + "="*50 + "\n")
|
|
|
|
# 编码输入
|
|
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
|
|
|
|
# 生成回答
|
|
print("正在生成回答...")
|
|
with torch.no_grad(): # 节省显存
|
|
generated_ids = model.generate(
|
|
**model_inputs,
|
|
max_new_tokens=2048, # 适当减少避免太长
|
|
do_sample=True,
|
|
temperature=0.7,
|
|
top_p=0.8,
|
|
repetition_penalty=1.05,
|
|
pad_token_id=tokenizer.eos_token_id
|
|
)
|
|
|
|
# 解码生成的tokens
|
|
generated_ids = [
|
|
output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
|
|
]
|
|
|
|
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
|
|
|
|
# 输出结果
|
|
print("模型回答:")
|
|
print(response)
|