LLMChain的使用

hehui0921

已于 2024-03-09 11:37:42 修改

阅读量355

点赞数 3

分类专栏： LangChain 文章标签： python 前端数据库

于 2024-02-07 11:00:45 首次发布

本文链接：https://blog.csdn.net/oHeHui1/article/details/136066732

版权

LangChain 专栏收录该内容

59 篇文章 4 订阅

订阅专栏

文章介绍了如何使用ChatGLM和LLMChain创建对话式链，通过模板生成描述制造产品（如鞋子）的公司在中国（如深圳、北京）的理想名称。展示了如何在Gradio中集成这些工具并执行预测和生成操作。

摘要由CSDN通过智能技术生成

import ChatGLM
from langchain.chains import LLMChain

from langchain_core.prompts import ChatPromptTemplate

from langchain.chains import SimpleSequentialChain

prompt = ChatPromptTemplate.from_template("What is the best name to describe a company that makes {product},please respond in chinese, must contian {a2}")
llm = ChatGLM.ChatGLM_LLM()
chain_one = LLMChain(
    llm=llm,
    prompt=prompt,
    verbose=True
)
print(chain_one.invoke({"product":"shoes","a2":"深圳"}))

from langchain.prompts import PromptTemplate
from langchain_core.prompts import ChatPromptTemplate
import gradio as gr
from langchain.chains import LLMChain
from LLMs import myllm
hf = myllm()
template = """{question}"""
prompt = PromptTemplate.from_template(template)
chain = prompt | hf


prompt = ChatPromptTemplate.from_template("What is the best name to describe a company that makes {product},please respond in chinese, must contian {words}")
prompt2 = ChatPromptTemplate.from_template("{input}")

llm = hf
chain_one = LLMChain(
    llm=llm,
    prompt=prompt,
    verbose=False
)

chain_one2 = LLMChain(
    llm=llm,
    prompt=prompt2,
    verbose=False
)

chains =  {"input": chain_one } |  chain_one2

print(chains.invoke({"product":"shoes","words":"深圳"}))

@chain的使用

from langchain.prompts import PromptTemplate
from langchain_core.prompts import ChatPromptTemplate
import gradio as gr
from langchain.chains import LLMChain
from langchain_core.runnables import chain
from langchain_core.output_parsers import StrOutputParser
from LLMs import myllm
hf = myllm()
prompt = ChatPromptTemplate.from_template("What is the best name to describe a company that makes {product},please respond in chinese, must contian {words}")
prompt2 = ChatPromptTemplate.from_template("{input}")

llm = hf
chain_one = LLMChain(
    llm=llm,
    prompt=prompt,
    verbose=False
)

chain_one2 = LLMChain(
    llm=llm,
    prompt=prompt2,
    verbose=False
)

@chain
def  myadd(name:str):
    print(name)
    # return "myadd"
    return name

# chains = {"input": chain_one } | prompt2|llm | StrOutputParser()

chains2 =  {"input": chain_one } | chain_one2 | myadd
# print(chains.invoke({"product":"shoes","words":"深圳"}))
print(chains2.invoke({"product":"shoes","words":"北京"}))

# print(chains.predict(product="shoes",words="深圳"))

# print(chains.generate(
#     [
#         {"product":"shoes","words":"深圳"},
#        {"product":"shoes","words":"北京"},
#      ]
#     ))