| 12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667686970717273747576777879808182838485868788899091929394959697 |
- """FastAPI 应用入口:挂载路由、配置 CORS、启动 uvicorn
- 启动方式:
- python run_backend.py
- 或
- uvicorn backend.api.main:app --host 0.0.0.0 --port 8000 --reload
- """
- from __future__ import annotations
- import logging
- from fastapi import FastAPI
- from fastapi.middleware.cors import CORSMiddleware
- from backend.api.routes import router
- # 配置日志
- logging.basicConfig(
- level=logging.INFO,
- format="%(asctime)s [%(levelname)s] %(name)s: %(message)s",
- )
- logger = logging.getLogger(__name__)
- app = FastAPI(
- title="Paper Innovation Agent API",
- description="论文创新 Agent 异步接口:提交分析、SSE 进度推送、报告查询",
- version="1.0.0",
- )
- # 配置 CORS:允许前端 localhost:5173 访问
- app.add_middleware(
- CORSMiddleware,
- allow_origins=[
- "http://localhost:5173",
- "http://127.0.0.1:5173",
- ],
- allow_credentials=True,
- allow_methods=["*"],
- allow_headers=["*"],
- )
- # 挂载路由
- app.include_router(router)
- @app.on_event("startup")
- async def on_startup() -> None:
- """应用启动日志 + 预加载重型模块
- 预加载 torch/transformers/chromadb 等重型依赖,避免首次处理论文时
- 才触发 import(实测 import torch 单独就要 100+ 秒)。
- """
- import time as _t
- _t0 = _t.perf_counter()
- logger.info("Paper Innovation Agent API 已启动,开始预加载重型模块...")
- # 顺序预加载:torch → transformers → chromadb → embedding(含模型加载)
- try:
- _t1 = _t.perf_counter()
- import torch # noqa: F401
- logger.info("[预加载] import torch: %.2fs", _t.perf_counter() - _t1)
- _t2 = _t.perf_counter()
- from transformers import AutoModel, AutoTokenizer # noqa: F401
- logger.info("[预加载] import transformers: %.2fs", _t.perf_counter() - _t2)
- _t3 = _t.perf_counter()
- import chromadb # noqa: F401
- logger.info("[预加载] import chromadb: %.2fs", _t.perf_counter() - _t3)
- # 预加载 Embedder 单例(首次会触发模型加载 + CUDA 上下文初始化)
- _t4 = _t.perf_counter()
- from backend.rag.embedding import get_default_embedder
- _embedder = get_default_embedder() # 注册为模块级单例
- # 触发模型加载 + 首次 forward(完成 CUDA 上下文初始化)
- _embedder.embed_texts(["预加载测试文本"])
- logger.info(
- "[预加载] Embedder 模型加载 + CUDA warmup: %.2fs (device=%s)",
- _t.perf_counter() - _t4,
- getattr(_embedder, "_device", "unknown"),
- )
- except Exception as exc:
- logger.warning("[预加载] 失败(不影响启动): %s", exc)
- logger.info("[预加载] 全部完成: %.2fs", _t.perf_counter() - _t0)
- if __name__ == "__main__":
- import uvicorn
- uvicorn.run(
- "backend.api.main:app",
- host="0.0.0.0",
- port=8000,
- reload=False,
- )
|