LCA-LLM/LCA_RAG/batchGLM.ipynb

In [5]:

# 创建批处理xlsx文件

import pandas as pd

customid = 1
method = "POST"
url = "/v4/chat/completions"
model = "glm-4"
role = "system"
instruction = "你是生命周期领域富有经验和知识的专家。根据你所掌握的知识回答问题；不要列出几点来回答，不需要换行，只需要用1句话回答问题。"

temperature = 0.95
top_p = 0.7
max_tokens = 4096

df = pd.DataFrame(columns=["custom_id","method","url","model","role","content","role1","content1","temperature","top_p","max_tokens"])
  

In [4]:

question  = []
with open("/home/zhangxj/WorkFile/LCA-GPT/QA/filters/question.txt","r",encoding="utf-8") as file:
    for line in file.readlines():
        question.append(line.strip())

In [12]:

data = []
for ques in question:
    row = {
        "custom_id": f"request-{customid}",
        "method":method,
        "url":url,
        "model":model,
        "role":role,
        "content":instruction,
        "role1":"user",
        "content1":ques,
        "temperature":temperature,
        "top_p":top_p,
        "max_tokens":max_tokens
    }
    data.append(row)
    customid+=1

In [13]:

len(data)

Out[13]:

3933

In [14]:

df = pd.DataFrame(data)

In [15]:

df.head()

# "custom_id","method","url","model","role","content","role1","content1","temperature","top_p","max_tokens"])

Out[15]:

	custom_id	method	url	model	role	content	role1	content1	temperature	top_p	max_tokens
0	request-1	POST	/v4/chat/completions	glm-4	system	你是生命周期领域富有经验和知识的专家。根据你所掌握的知识回答问题；不要列出几点来回答，不需要...	user	什么是生命周期分析（LCA）的主要目标？	0.95	0.7	4096
1	request-2	POST	/v4/chat/completions	glm-4	system	你是生命周期领域富有经验和知识的专家。根据你所掌握的知识回答问题；不要列出几点来回答，不需要...	user	在LCA中，如何确定研究的范围？	0.95	0.7	4096
2	request-3	POST	/v4/chat/completions	glm-4	system	你是生命周期领域富有经验和知识的专家。根据你所掌握的知识回答问题；不要列出几点来回答，不需要...	user	医疗废物如何处理？	0.95	0.7	4096
3	request-4	POST	/v4/chat/completions	glm-4	system	你是生命周期领域富有经验和知识的专家。根据你所掌握的知识回答问题；不要列出几点来回答，不需要...	user	LCA数据清单收集阶段需要哪些信息？	0.95	0.7	4096
4	request-5	POST	/v4/chat/completions	glm-4	system	你是生命周期领域富有经验和知识的专家。根据你所掌握的知识回答问题；不要列出几点来回答，不需要...	user	生命周期影响评价阶段的目标是什么？	0.95	0.7	4096

In [16]:

df.to_excel("/home/zhangxj/WorkFile/LCA-GPT/QA/questionForBatch.xlsx",index=False)

In [20]:

from zhipuai import ZhipuAI
 
client = ZhipuAI(api_key="434790cf952335f18b6347e7b6de9777.V50p55zfk8Ye4ojV")  # 填写您自己的APIKey

create = client.batches.create(
    input_file_id="1723556210_f79e4160ab3840b4b02f44c821d27752",
    endpoint="/v4/chat/completions", 
    completion_window="24h", #完成时间只支持 24 小时
    metadata={
        "description": "回答问题"
    }
)
print(create)

Batch(id='batch_1823353255129645056', completion_window='24h', created_at=1723556266945, endpoint='/v4/chat/completions', input_file_id='1723556210_f79e4160ab3840b4b02f44c821d27752', object='batch', status='validating', cancelled_at=None, cancelling_at=None, completed_at=None, error_file_id=None, errors=None, expired_at=None, expires_at=None, failed_at=None, finalizing_at=None, in_progress_at=None, metadata={'description': '回答问题'}, output_file_id=None, request_counts=BatchRequestCounts(completed=None, failed=None, total=3933))

In [21]:

batch_job = client.batches.retrieve("batch_id")
print(batch_job)

Batch(id=None, completion_window=None, created_at=None, endpoint=None, input_file_id=None, object=None, status=None, cancelled_at=None, cancelling_at=None, completed_at=None, error_file_id=None, errors=None, expired_at=None, expires_at=None, failed_at=None, finalizing_at=None, in_progress_at=None, metadata=None, output_file_id=None, request_counts=None)

In [ ]: