本脚本是对千问7B模型的微调和测试,使得Qwen 7B能够有角色扮演的能力
This script fine-tunes and tests the Qwen 7B model to give Qwen 7B the capability of role playing.
项目链接 https://github.com/LC1332/Chat-Haruhi-Suzumiya
118K训练数据由李鲁鲁收集,
模型是由豆角训练的
Qwen inference代码由米唯实编写,并接入Chatharuhi,目前进行本模型维护和bug解决
李鲁鲁编写了ChatHaruhi内部的prompt组织函数
A Harry Potter test see in https://github.com/LC1332/Chat-Haruhi-Suzumiya/blob/main/notebook/Harry_Potter_test_on_Qwen7B.ipynb
使用方法
载入函数
from transformers import AutoTokenizer, AutoModel, AutoModelForCausalLM
tokenizer = AutoTokenizer.from_pretrained("silk-road/ChatHaruhi_RolePlaying_qwen_7b", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("silk-road/ChatHaruhi_RolePlaying_qwen_7b", device_map="auto", trust_remote_code=True)
model = model.eval()
具体看https://github.com/LC1332/Chat-Haruhi-Suzumiya/blob/main/notebook/ChatHaruhi_x_Qwen7B.ipynb 这个notebook
from ChatHaruhi import ChatHaruhi
chatbot = ChatHaruhi( role_name = 'haruhi', max_len_story = 1000 )
prompt = chatbot.generate_prompt(role='阿虚', text = '我看新一年的棒球比赛要开始了!我们要去参加吗?')
response, history = model.chat(tokenizer, prompt, history=[])
print(response)
chatbot.append_response(response)
目前支持 role_name
role_from_hf
role_from_jsonl
多种角色格式载入。
- Downloads last month
- 12
Inference API (serverless) does not yet support model repos that contain custom code.