Update retriever.py

retriever.py

@@ -1,5 +1,7 @@
+import os
 import logging
 
+# Optional FAISS (keeps your original behavior)
 try:
     import faiss
     _HAS_FAISS = True
@@ -9,8 +11,41 @@ except ImportError:
 
 from sentence_transformers import SentenceTransformer
 
-# load embedding model (still works even if FAISS missing)
-_model = SentenceTransformer("all-MiniLM-L6-v2")
+# ---- Writable cache + stable repo id for Spaces ----
+_ST_CACHE = os.getenv("SENTENCE_TRANSFORMERS_HOME", "/data/.cache/sentence-transformers")
+_ST_MODEL_ID = "sentence-transformers/all-MiniLM-L6-v2"  # canonical repo id
+
+def _load_st_model():
+    """
+    Load SentenceTransformer using an explicit cache folder to avoid
+    Hugging Face 'xet' transport / permission issues on Spaces.
+    """
+    # Ensure cache dir exists
+    try:
+        os.makedirs(_ST_CACHE, exist_ok=True)
+    except Exception as e:
+        logging.warning(f"Could not create cache directory {_ST_CACHE}: {e}")
+
+    # Primary attempt
+    try:
+        return SentenceTransformer(_ST_MODEL_ID, cache_folder=_ST_CACHE)
+    except Exception as e1:
+        logging.warning(f"Primary load failed for '{_ST_MODEL_ID}' with cache '{_ST_CACHE}': {e1}")
+
+        # Secondary attempt (allow trust_remote_code just in case)
+        try:
+            return SentenceTransformer(_ST_MODEL_ID, cache_folder=_ST_CACHE, trust_remote_code=True)
+        except Exception as e2:
+            logging.exception("Failed loading SentenceTransformer model on both attempts.")
+            raise RuntimeError(
+                f"Failed loading SentenceTransformer '{_ST_MODEL_ID}'.\n"
+                f"First error: {e1}\nSecond error: {e2}\n"
+                f"Check cache dir permissions at: {_ST_CACHE}\n"
+                f"Tip: ensure app.py sets HF_HUB_ENABLE_XET=0 and uses writable caches under /data."
+            )
+
+# Load embedding model (works even if FAISS missing)
+_model = _load_st_model()
 
 _index = None
 _docs = []
@@ -27,7 +62,7 @@ def init_retriever(docs=None):
 
     if docs:
         _docs = docs
-        embeddings = _model.encode(docs, convert_to_numpy=True)
+        embeddings = _model.encode(docs, convert_to_numpy=True, normalize_embeddings=False)
         d = embeddings.shape[1]
         _index = faiss.IndexFlatL2(d)
         _index.add(embeddings)
@@ -35,11 +70,11 @@ def init_retriever(docs=None):
 def retrieve_context(query: str, k: int = 5):
     """
     Retrieve top-k docs matching query.
-    Falls back to empty list if FAISS unavailable.
+    Falls back to empty list if FAISS unavailable or not initialized.
     """
     if not _HAS_FAISS or _index is None or not _docs:
         return []
 
-    q_emb = _model.encode([query], convert_to_numpy=True)
+    q_emb = _model.encode([query], convert_to_numpy=True, normalize_embeddings=False)
     D, I = _index.search(q_emb, k)
-    return [_docs[i] for i in I[0] if i < len(_docs)]
+    return [_docs[i] for i in I[0] if 0 <= i < len(_docs)]