File size: 5,009 Bytes
1359487 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 | """
FastAPI application entry-point.
Local:
uvicorn api.main:app --reload --port 8000
Render:
uvicorn api.main:app --host 0.0.0.0 --port $PORT
(set REDIS_URL in the Render dashboard under Environment)
"""
from __future__ import annotations
import logging
import os
from contextlib import asynccontextmanager
from pathlib import Path
from dotenv import load_dotenv
# Load .env before any os.getenv() calls — works locally and is a no-op on
# Render (where env vars are injected directly by the platform).
load_dotenv(Path(__file__).resolve().parent.parent / ".env")
from fastapi import FastAPI
from fastapi.middleware.cors import CORSMiddleware
from api.event_logger import EventLogger
from api.routes import router, set_engine, set_event_logger
from serving.feature_store import FeatureStore
from serving.inference import RecommendationEngine
logging.basicConfig(
level=logging.INFO,
format="%(asctime)s | %(levelname)s | %(name)s | %(message)s",
)
logger = logging.getLogger(__name__)
_RECOMMENDER_ROOT = Path(__file__).resolve().parent.parent # recommender/
ARTIFACT_DIR = Path(os.getenv("ARTIFACT_DIR", str(_RECOMMENDER_ROOT / "artifacts")))
_REQUIRED_ARTIFACTS = [
"preprocessor.pkl",
"movie_meta.csv",
"two_tower.pt",
"deepfm_best.pt",
"faiss.index",
"item_embeddings.npy",
"item_features.npy",
"user_features.npy",
]
# ------------------------------------------------------------------
# Step 1 — Check
# ------------------------------------------------------------------
def _check_artifacts(artifact_dir: Path) -> None:
"""Raise a descriptive RuntimeError if any required file is missing."""
missing = [f for f in _REQUIRED_ARTIFACTS if not (artifact_dir / f).exists()]
if missing:
raise RuntimeError(
f"\n\n{'='*60}\n"
f" Artifacts missing in: {artifact_dir}\n"
f" Missing: {', '.join(missing)}\n\n"
f" Run training to generate them:\n"
f" ./run_training.sh\n"
f"{'='*60}\n"
)
# ------------------------------------------------------------------
# Step 2 — Load
# ------------------------------------------------------------------
def load_models(artifact_dir: Path) -> tuple[RecommendationEngine, EventLogger]:
"""
Instantiate the feature store, recommendation engine, and event logger.
Kept as a plain function so it can be called from tests or CLI without
going through the full FastAPI lifespan.
"""
feature_store = FeatureStore(
redis_host=os.getenv("REDIS_HOST", "localhost"),
redis_port=int(os.getenv("REDIS_PORT", 6379)),
sqlite_path=artifact_dir / "feature_store.db",
)
engine = RecommendationEngine.load(
artifact_dir,
device_str=os.getenv("DEVICE", "cpu"),
feature_store=feature_store,
)
event_logger = EventLogger(
kafka_bootstrap=os.getenv("KAFKA_BOOTSTRAP", "localhost:9092"),
sqlite_path=artifact_dir / "events.db",
)
return engine, event_logger
# ------------------------------------------------------------------
# Lifespan — wires the three steps together
# ------------------------------------------------------------------
@asynccontextmanager
async def lifespan(app: FastAPI):
logger.info(f"Artifact dir: {ARTIFACT_DIR}")
# Artifacts come from GitHub directly — no download needed
_check_artifacts(ARTIFACT_DIR)
engine, event_logger = load_models(ARTIFACT_DIR)
set_engine(engine)
set_event_logger(event_logger)
logger.info("Startup complete.")
yield
event_logger.close()
# ------------------------------------------------------------------
# App
# ------------------------------------------------------------------
app = FastAPI(
title="CineMatch Recommendation API",
description=(
"Production recommendation system: Two-Tower retrieval, "
"DeepFM ranking, MMR diversity re-ranking."
),
version="1.0.0",
lifespan=lifespan,
)
# CORS: accept local dev ports + any *.onrender.com subdomain.
# CORS_ORIGINS env var lets Render / CI override this without a code change.
_extra_origins = [o.strip() for o in os.getenv("CORS_ORIGINS", "").split(",") if o.strip()]
app.add_middleware(
CORSMiddleware,
allow_origins=[
"http://localhost:5173",
"http://localhost:5174",
"http://localhost:5175",
"http://localhost:5176",
"http://localhost:5177",
"http://localhost:3000",
"http://127.0.0.1:5173",
"http://127.0.0.1:5177",
*_extra_origins,
],
allow_origin_regex=r"https://.*\.onrender\.com", # matches any Render deploy URL
allow_credentials=True,
allow_methods=["*"],
allow_headers=["*"],
)
app.include_router(router, prefix="/api")
@app.get("/", tags=["system"])
def root():
return {
"service": "CineMatch Recommendation API",
"version": "1.0.0",
"docs": "/docs",
}
|