Skip to content

安装必要的软件

ollama

安装教程

安装教程

启动命令

# 运行 llama3 8b 版本，内存占用大概 4.7g
ollama run llama3:8b

配置文件 config.json

{
    "model":"llama3:8b"
}

python代码

import json
import gradio as gr

from langchain_community.chat_models import ChatOllama
from langchain_core.prompts import PromptTemplate
from langchain_community.document_loaders import PyPDFLoader
from langchain_community.vectorstores import Chroma
from langchain_community.embeddings import FastEmbedEmbeddings
from langchain.vectorstores.utils import filter_complex_metadata
from langchain_core.runnables import RunnablePassthrough
from langchain_core.output_parsers import StrOutputParser

#Create the prompte from the template.
promptTemplate = """Answer the question as precise as possible using the provided context. If the answer is
    not contained in the context, say "answer not available in context" \n\n
    Context: {context}
    Question: {question}
    Answer:

     """
modelSel = ""
#Load the PDF file to ChromaDB
def loadDataFromPDFFile(filePath):
    loader = PyPDFLoader(filePath)
    pages = loader.load_and_split()
    chunks = filter_complex_metadata(pages)
    vector_store = Chroma.from_documents(documents=chunks, embedding=FastEmbedEmbeddings())
    return vector_store

def modelResponse(message , history):
    print(message)
    llm = ChatOllama(model = conf["model"])

    prompt = PromptTemplate(template=promptTemplate , input_variables=["context","question"])

    # 要加载的 pdf 路径
    pdfPath = ""
    #Initiate the retriever
    dbLoaded = loadDataFromPDFFile(pdfPath)
    retriever = dbLoaded.as_retriever(search_type="similarity_score_threshold" , search_kwargs = {
        "k":  5,
        "score_threshold": 0.2
    })

    hpChain = (
            {"context": retriever , "question": RunnablePassthrough()}
            | prompt
            | llm
            | StrOutputParser()
    )
    return hpChain.invoke(message)

if __name__ == "__main__":

    #read configuration file
    conf = {}
    with open("config.json" , "r") as confFile:
        conf = json.load(confFile)
        print(conf["model"])

    chatUI = gr.ChatInterface(fn=modelResponse , title="施瓦辛格健身大全")
    chatUI.launch()

页面示例

示例