HF testing

app/services/vector_store.py

@@ -8,11 +8,15 @@ from app.config import params
 # Paths for vector store and metadata
 VECTOR_STORE_DIR = "/app/data"
 os.makedirs(VECTOR_STORE_DIR, exist_ok=True)
+os.makedirs("/app/cache", exist_ok=True)
 INDEX_PATH = os.path.join(VECTOR_STORE_DIR, params["paths"]["index_file"])
 META_PATH = os.path.join(VECTOR_STORE_DIR, params["paths"]["metadata_file"])
 
 # Embedding model
-EMBED_MODEL = SentenceTransformer(params["embedding"]["model_name"])
+EMBED_MODEL = SentenceTransformer(
+    params["embedding"]["model_name"],
+    cache_folder="/app/cache"
+)
 
 def embed_texts(texts: List[str]):
     """

requirements.txt

@@ -27,7 +27,8 @@ pdf2image --trusted-host=pypi.org --trusted-host=files.pythonhosted.org
 python-dotenv
 requests
 tiktoken
-sentence-transformers
+sentence-transformers==2.6.1
+transformers>=4.36.2
 faiss-cpu
 pytesseract
 nltk