手把收带你调用大模型+附上清华大学云盘：chatglm2位置

from transformers import AutoTokenizer, AutoModel
import pandas as pd
tokenizer = AutoTokenizer.from_pretrained("model", trust_remote_code=True, revision="")
model = AutoModel.from_pretrained("model", trust_remote_code=True, revision="").half().cuda()
model = model.eval()
filename = 'data.csv'
data = pd.read_csv(filename)
data = data[['id','ques','question','ans','len']]

# 新增一列
new_column = []  # 存储新列的数据
n = 0
i = 0 
# 循环遍历每一行，并添加新数据到new_column列表中
for index, row in data.iterrows():
    try:
        response, history = model.chat(tokenizer,data['question'], history=[])
        #print(response)
        print("处理第几条：",i)
        i += 1 
        new_column.append(response)
    except:
        new_column.append("无法解答")
        n += 1
print("无法解答有多少条：",n)

# 将new_column列表作为新列添加到DataFrame中
data['chatglm20703'] = new_column
data.to_csv("new_data.csv")

修改的位置，model。

model目录下，放入你从下载的模型文件，比如，我这放的是chatglm2模型。

清华大学云盘 https://cloud.tsinghua.edu.cn/d/674208019e314311ab5c/?p=%2Fchatglm2-6b&mode=list