LangChain 48 终极解决 实战Langchain访问OpenAI ChatGPT API Account deactivated的另类方法，访问跳板机API_chatgpt的api调用langchain

OpenAI ChatGPT API Account deactivated的另类方法，访问跳板机API 系列

• LangChain 40 实战Langchain访问OpenAI ChatGPT API Account deactivated的另类方法，访问跳板机API
• Langchain访问OpenAI ChatGPT API Account deactivated的另类方法，访问跳板机API

1. 核心其实在于key和url的设置

方法有：

• 使用环境变量来设置
• 使用变量来传入
• 使用手动设置环境变量

笔者用的是macOS, 先添加配置到 .zshrc 文件

export WANDOU_OPENAI_API_KEY='sk-xxx'
export WANDOU_BASE_URL="https://apejhvxcd.cloud.sealos.io/v1”
1
2

重新加载配置

source .zshrc
1

call 一个API示例

import os
import requests
import time
import json
import time
from langchain.llms import OpenAI

API_SECRET_KEY = "你在跳板机的key";
BASE_URL = "https://apejhvxcd.cloud.sealos.io/v1";

os.environ["OPENAI_API_KEY"] = API_SECRET_KEY
os.environ["OPENAI_API_BASE"] = BASE_URL

def text():
    llm = OpenAI(temperature=0.9)
    text = "What would be a good company name for a company that makes colorful socks?"
    print(llm(text))

if __name__ == '__main__':
    text();

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21

踩坑执行代码

print(os.environ["OPENAI_API_KEY"])
1

发现是OpenAI的token，不是跳板机的。

.env 配置

OPENAI_API_KEY = os.getenv("WANDOU_OPENAI_API_KEY")
OPENAI_API_BASE = os.getenv("WANDOU_BASE_URL")
1
2

再次输出就是跳板机的token了，笔者用的跳板机链接，有送美金哦

2. 并行call 6次 OpenAI API试试

2.1 准备工作，打印执行时间

import time

def measure_execution_time(func, *args, **kwargs):
    """测量并返回给定函数的执行时间。

    Args:
        func: 要测量的函数对象。
        *args: 传递给函数的位置参数。
        **kwargs: 传递给函数的关键字参数。

    Returns:
        tuple: (执行结果, 耗时秒数)
    """
    start_time = time.perf_counter()
    result = func(*args, **kwargs)
    end_time = time.perf_counter()
    elapsed_time = end_time - start_time
    return result, elapsed_time

# 示例：假设 chain1 已经定义并且具有 invoke 方法
# chain1 = ...

# 使用 measure_execution_time 来测量 chain1.invoke 的执行时间
# result, exec_time = measure_execution_time(chain1.invoke, {"topic": "bears"})
# print(f"执行结果: {result}")
# print(f"函数执行耗时: {exec_time} 秒")

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

2.2 直接RunnableParallel并行call 6次OpenAI 试试

from langchain_core.runnables import RunnablePassthrough
from langchain.prompts import ChatPromptTemplate
from langchain.chat_models import ChatOpenAI
from langchain_core.output_parsers import StrOutputParser
import time
from calc_time import measure_execution_time
from dotenv import load_dotenv
import os
load_dotenv()

from langchain_core.runnables import RunnableParallel

# print(os.environ["OPENAI_API_KEY"])

# model = ChatOpenAI()
model = ChatOpenAI(model="gpt-3.5-turbo")

chain1 = ChatPromptTemplate.from_template("tell me a joke about {topic}") | model
chain2 = (
    ChatPromptTemplate.from_template("write a short (2 line) poem about {topic}")
    | model
)
combined = RunnableParallel(joke=chain1, poem=chain2)

# result, exec_time = measure_execution_time(chain1.invoke, {"topic": "bears"})
# print(f"执行结果: {result}")
# print(f"函数执行耗时: {exec_time} 秒")


# result, exec_time = measure_execution_time(chain2.invoke, {"topic": "bears"})
# print(f"执行结果: {result}")
# print(f"函数执行耗时: {exec_time} 秒")


# result, exec_time = measure_execution_time(combined.invoke, {"topic": "bears"})
# print(f"执行结果: {result}")
# print(f"函数执行耗时: {exec_time} 秒")

# result, exec_time = measure_execution_time(chain1.batch, [{"topic": "bears"}, {"topic": "cats"}])
# print(f"执行结果: {result}")
# print(f"函数执行耗时: {exec_time} 秒")

# result, exec_time = measure_execution_time(chain2.batch, [{"topic": "bears"}, {"topic": "cats"}])
# print(f"执行结果: {result}")
# print(f"函数执行耗时: {exec_time} 秒")

result, exec_time = measure_execution_time(combined.batch, [{"topic": "bears"}, {"topic": "cats"}, {"topic": "dogs"}])
print(f"执行结果: {result}")
print(f"函数执行耗时: {exec_time} 秒")

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

执行结果

zgpeace at zgpeaces-MacBook-Pro in ~/Workspace/LLM/langchain-llm-app on develop!
(.venv)  ± python LCEL/parallelism.py
执行结果: [{'joke': AIMessage(content="Why don't bears wear shoes?\n\nBecause they have bear feet!"), 'poem': AIMessage(content="Majestic bears roam,\nNature's guardians, fierce yet kind.")}, {'joke': AIMessage(content="Sure, here's a cat joke for you:\n\nWhy was the cat sitting on the computer?\n\nBecause it wanted to keep an eye on the mouse!"), 'poem': AIMessage(content="Whiskers glide, tails entwine,\nGraceful feline, nature's design.")}, {'joke': AIMessage(content="Sure, here's a dog-related joke for you:\n\nWhy don't dogs make good dancers?\n\nBecause they have two left feet!"), 'poem': AIMessage(content='Loyal companions, hearts pure and true,\nDogs bring joy, love, and laughter to you.')}]
函数执行耗时: 3.3068008899572305 秒
1
2
3
4

跳板机

https://docs.qq.com/doc/DRXJxV1Nad3hsb3ZL

代码

https://github.com/zgpeace/pets-name-langchain/tree/develop

参考

• https://blog.csdn.net/zgpeace/article/details/135246321