Spaces:
Sleeping
Sleeping
File size: 9,162 Bytes
6d6a4c1 8662bec 6d6a4c1 8662bec 6d6a4c1 bd907d1 6d6a4c1 8662bec 6d6a4c1 bd907d1 6d6a4c1 8662bec 6d6a4c1 bd907d1 6d6a4c1 bd907d1 6d6a4c1 bd907d1 6d6a4c1 bd907d1 6d6a4c1 bd907d1 6d6a4c1 8662bec 6d6a4c1 8662bec 6d6a4c1 bd907d1 6d6a4c1 bd907d1 6d6a4c1 bd907d1 6d6a4c1 bd907d1 6d6a4c1 8662bec 6d6a4c1 bd907d1 6d6a4c1 bd907d1 6d6a4c1 bd907d1 6d6a4c1 bd907d1 6d6a4c1 8662bec 6d6a4c1 8662bec 6d6a4c1 8662bec 6d6a4c1 8662bec 6d6a4c1 8662bec 6d6a4c1 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 | import os
import json
import time
from typing import List, Dict, Any, Optional
import numpy as np
import faiss
from fastapi import FastAPI, HTTPException
from fastapi.middleware.cors import CORSMiddleware
from pydantic import BaseModel
from sentence_transformers import SentenceTransformer
# -----------------------------
# Paths
# -----------------------------
BASE_DIR = os.path.dirname(os.path.abspath(__file__))
JSON_PATH = os.path.join(BASE_DIR, "hadith_corpus25k.json")
ART_DIR = os.path.join(BASE_DIR, "artifacts_hadith_faiss")
INDEX_PATH = os.path.join(ART_DIR, "faiss.index")
# IMPORTANT: np.save adds ".npy" if not present; keep path WITHOUT extension
EMB_PATH = os.path.join(ART_DIR, "embeddings") # will produce embeddings.npy
ID_BY_POS_PATH = os.path.join(ART_DIR, "id_by_pos.json")
POS_BY_ID_PATH = os.path.join(ART_DIR, "pos_by_id.json")
# Settings
MODEL_NAME = os.getenv("MODEL_NAME", "intfloat/multilingual-e5-base")
BATCH_SIZE = int(os.getenv("BATCH_SIZE", "64"))
TOPK_MAX = int(os.getenv("TOPK_MAX", "50"))
# -----------------------------
# App
# -----------------------------
app = FastAPI(title="Hadith FAISS API", version="1.0")
app.add_middleware(
CORSMiddleware,
allow_origins=["*"], # ูู ุชุจู ุชููููุง ุนูู ุฏูู
ูู ู
ููุนู ููุท ูู ูู
allow_credentials=True,
allow_methods=["*"],
allow_headers=["*"],
)
# -----------------------------
# Globals (loaded at startup)
# -----------------------------
_items: List[Dict[str, Any]] = []
_item_by_id: Dict[int, Dict[str, Any]] = {}
_model: Optional[SentenceTransformer] = None
_index: Optional[faiss.Index] = None
_emb: Optional[np.ndarray] = None
_id_by_pos: List[int] = []
_pos_by_id: Dict[int, int] = {}
_DIM: int = 0
_READY: bool = False
# -----------------------------
# Helpers
# -----------------------------
def build_text(x: Dict[str, Any]) -> str:
ar = (x.get("arabic_clean") or x.get("arabic") or "").strip()
en = (x.get("english") or "").strip()
if ar and en:
return ar + " [SEP] " + en
return ar or en
def ensure_dirs():
os.makedirs(ART_DIR, exist_ok=True)
def artifacts_exist() -> bool:
return (
os.path.exists(INDEX_PATH)
and os.path.exists(EMB_PATH + ".npy")
and os.path.exists(ID_BY_POS_PATH)
and os.path.exists(POS_BY_ID_PATH)
)
def load_items():
global _items, _item_by_id
if not os.path.exists(JSON_PATH):
raise RuntimeError(f"Missing dataset file: {JSON_PATH}")
with open(JSON_PATH, "r", encoding="utf-8") as f:
_items = json.load(f)
_item_by_id = {}
for it in _items:
cid = it.get("corpusID")
if cid is not None:
_item_by_id[int(cid)] = it
def get_model() -> SentenceTransformer:
global _model
if _model is None:
_model = SentenceTransformer(MODEL_NAME)
return _model
def save_artifacts(
index: faiss.Index,
emb: np.ndarray,
id_by_pos: List[int],
pos_by_id: Dict[int, int],
):
ensure_dirs()
faiss.write_index(index, INDEX_PATH)
np.save(EMB_PATH, emb) # will create embeddings.npy
with open(ID_BY_POS_PATH, "w", encoding="utf-8") as f:
json.dump([int(x) for x in id_by_pos], f, ensure_ascii=False)
pos_by_id_str = {str(k): int(v) for k, v in pos_by_id.items()}
with open(POS_BY_ID_PATH, "w", encoding="utf-8") as f:
json.dump(pos_by_id_str, f, ensure_ascii=False)
def load_artifacts():
global _index, _emb, _id_by_pos, _pos_by_id, _DIM
_index = faiss.read_index(INDEX_PATH)
_emb = np.load(EMB_PATH + ".npy").astype("float32")
with open(ID_BY_POS_PATH, "r", encoding="utf-8") as f:
_id_by_pos = [int(x) for x in json.load(f)]
with open(POS_BY_ID_PATH, "r", encoding="utf-8") as f:
raw = json.load(f)
_pos_by_id = {int(k): int(v) for k, v in raw.items()}
_DIM = int(_emb.shape[1])
def build_all():
"""
Build embeddings + FAISS then save.
This should run only if artifacts are missing.
"""
global _index, _emb, _id_by_pos, _pos_by_id, _DIM
t0 = time.time()
model = get_model()
texts = [build_text(x) for x in _items]
passages = ["passage: " + t for t in texts] # E5 passage prefix
emb = model.encode(
passages,
normalize_embeddings=True,
batch_size=BATCH_SIZE,
show_progress_bar=True,
)
emb = np.asarray(emb, dtype="float32")
dim = int(emb.shape[1])
index = faiss.IndexFlatIP(dim) # cosine via IP since normalized
index.add(emb)
id_by_pos = [int(x["corpusID"]) for x in _items]
pos_by_id = {cid: i for i, cid in enumerate(id_by_pos)}
save_artifacts(index, emb, id_by_pos, pos_by_id)
_index = index
_emb = emb
_id_by_pos = id_by_pos
_pos_by_id = pos_by_id
_DIM = dim
dt = time.time() - t0
print(f"[build_all] Built + saved artifacts in {dt:.2f}s. dim={_DIM}, n={len(_id_by_pos)}")
def require_ready():
if not _READY or _index is None or _emb is None:
raise HTTPException(status_code=503, detail="API is not ready yet. Try again in a moment.")
def pack_item(it: Dict[str, Any]) -> Dict[str, Any]:
return {
"corpusID": it.get("corpusID"),
"book": it.get("book"),
"chapter": it.get("chapter"),
"arabic": it.get("arabic_clean") or it.get("arabic"),
"english": it.get("english"),
"grade": it.get("grade"),
"meta": it.get("meta"),
}
def embed_query(q: str) -> np.ndarray:
model = get_model()
vec = model.encode(["query: " + q], normalize_embeddings=True) # E5 query prefix
return np.asarray(vec, dtype="float32")
# -----------------------------
# Request Models
# -----------------------------
class SearchRequest(BaseModel):
query: str
topk: int = 10
# -----------------------------
# Startup
# -----------------------------
@app.on_event("startup")
def on_startup():
global _READY
try:
print("[startup] Loading items...")
load_items()
print(f"[startup] Loaded items: {len(_items)}")
if artifacts_exist():
print("[startup] Artifacts found. Loading...")
load_artifacts()
print(f"[startup] Loaded artifacts: dim={_DIM}, n={len(_id_by_pos)}")
else:
print("[startup] Artifacts NOT found. Building now (first run)...")
build_all()
_READY = True
print("[startup] READY โ
")
except Exception as e:
_READY = False
print("[startup] FAILED โ", str(e))
# keep app up but not ready
# -----------------------------
# Routes
# -----------------------------
@app.get("/")
def root():
return {"name": "Hadith FAISS API", "ready": _READY}
@app.get("/health")
def health():
return {
"ready": _READY,
"items": len(_items),
"dim": _DIM,
"has_artifacts": artifacts_exist(),
"model": MODEL_NAME,
}
@app.get("/stats")
def stats():
require_ready()
return {
"items": len(_items),
"dim": _DIM,
"index_type": type(_index).__name__,
"topk_max": TOPK_MAX,
}
@app.get("/item/{corpus_id}")
def get_item(corpus_id: int):
require_ready()
it = _item_by_id.get(int(corpus_id))
if not it:
raise HTTPException(status_code=404, detail="corpusID not found")
return pack_item(it)
@app.get("/similar/{corpus_id}")
def similar(corpus_id: int, topk: int = 10):
require_ready()
cid = int(corpus_id)
if cid not in _pos_by_id:
raise HTTPException(status_code=404, detail="corpusID not found in index")
topk = max(1, min(int(topk), TOPK_MAX))
pos = _pos_by_id[cid]
q = _emb[pos:pos + 1] # already normalized
scores, idxs = _index.search(q, topk + 1) # +1 to skip itself
scores = scores[0].tolist()
idxs = idxs[0].tolist()
results = []
for sc, p in zip(scores, idxs):
if p < 0:
continue
hit_id = _id_by_pos[p]
if hit_id == cid:
continue
it = _item_by_id.get(int(hit_id))
if not it:
continue
results.append({
"corpusID": int(hit_id),
"score": float(sc),
"item": pack_item(it),
})
if len(results) >= topk:
break
return {"query_id": cid, "topk": topk, "results": results}
@app.post("/search")
def search(req: SearchRequest):
require_ready()
q = (req.query or "").strip()
if not q:
raise HTTPException(status_code=400, detail="query is empty")
topk = max(1, min(int(req.topk), TOPK_MAX))
qv = embed_query(q)
scores, idxs = _index.search(qv, topk)
scores = scores[0].tolist()
idxs = idxs[0].tolist()
results = []
for sc, p in zip(scores, idxs):
if p < 0:
continue
hit_id = _id_by_pos[p]
it = _item_by_id.get(int(hit_id))
if not it:
continue
results.append({
"corpusID": int(hit_id),
"score": float(sc),
"item": pack_item(it),
})
return {"query": q, "topk": topk, "results": results}
|