问题验证

• 我已在文档和Discord中寻找答案。

问题

from weaviate.classes.config import Configure, Property, DataType
import time
import weaviate
import chromadb
from llama_index.core import VectorStoreIndex, SimpleDirectoryReader
from llama_index.vector_stores.chroma import ChromaVectorStore
from llama_index.vector_stores.weaviate import WeaviateVectorStore
from llama_index.core.node_parser import SentenceSplitter
from llama_index.core.ingestion import IngestionPipeline
from llama_index.readers.file import PyMuPDFReader
from llama_index.core import Settings
from llama_index.core import StorageContext
from llama_index.embeddings.huggingface import HuggingFaceEmbedding
from llama_index.core.postprocessor import SentenceTransformerRerank
from llama_index.core.retrievers import QueryFusionRetriever
from llama_index.retrievers.bm25 import BM25Retriever
from llama_index.core.query_engine import RetrieverQueryEngine
from llama_index.core.chat_engine import CondenseQuestionChatEngine
from llama_index.llms.openai import OpenAI
from llama_index.embeddings.openai import OpenAIEmbedding
from dotenv import load_dotenv, find_dotenv

# 加载环境变量
_ = load_dotenv(find_dotenv())
import nest_asyncio
nest_asyncio.apply() # 只在Jupyter笔记环境中需要此操作，否则会报错

# 连接到local,需要启动docker
weaviate_client = weaviate.connect_to_local(host="localhost", port=8080, grpc_port=50051, skip_init_checks=True)
# weaviate_client = weaviate.connect_to_local(host="localhost", port=8080, grpc_port=50051)
# 指定全局llm与embedding模型
Settings.llm = OpenAI(temperature=0, model="gpt-4o")
Settings.embed_model = OpenAIEmbedding(model="text-embedding-3-small", dimensions=512)
# Settings.embed_model = HuggingFaceEmbedding(model_name=r"C:\Users\79033\.cache\huggingface\hub\models--BAAI--bge-m3\snapshots\5617a9f61b028005a4858fdac845db406aefb181", )

splitter = SentenceSplitter(chunk_size=512, chunk_overlap=100)

# 加载 pdf 文档
documents = SimpleDirectoryReader("./data1", file_extractor={".pdf": PyMuPDFReader()}).load_data()

# 切分nodes
nodes = splitter.get_nodes_from_documents(documents)

# 创建 Vector Store
# vector_store = ChromaVectorStore(chroma_collection=chroma_collection)
vector_store = WeaviateVectorStore(weaviate_client=weaviate_client, index_name="LlamaIndex1")
# 指定 Vector Store 用于 index
storage_context = StorageContext.from_defaults(vector_store=vector_store)
index = VectorStoreIndex(
    nodes, storage_context=storage_context
)
print(index.index_struct)

query_engine = index.as_query_engine()
while True:
    question = input("User:")
    if question.strip() == "":
        break
    start_time = time.time()
    response = query_engine.query(question)
    for node in response.source_nodes:
        print(f"Text: {node.node.text}")
        print(f"Score: {node.score}")
        print(f"Metadata: {node.node.metadata}")
        print("---")
    end_time = time.time()
    print(end_time-start_time)
    print(f"AI: {response}")