from langchain_openai import ChatOpenAI
from langchain_core.prompts import ChatPromptTemplate
async def build_rag_pipeline(query):
retriever = vectorstore.as_retriever()
prompt = ChatPromptTemplate.from_messages([
("system", "基于资料回答"),
("human", "{question}")
])
chain = (
{"context": retriever}
| prompt | llm
)
return await chain.ainvoke(query)
from langgraph.prebuilt import create_react_agent
agent = create_react_agent(llm, tools)
lora_config = LoraConfig(
r=8, lora_alpha=16,
target_modules=["q_proj", "v_proj"],
lora_dropout=0.1
)
trainer = SFTTrainer(
model=model,
args=training_args,
train_dataset=dataset
)
trainer.train()
python -m vllm.entrypoints.openai.api_server \
--model qwen2.5-7b-lora-merged \
--port 8000 \
--max-model-len 4096
results = collection.query(
query_embeddings=[query_vec],
n_results=3
)