Add HF hub fallback and dependency

app.py

@@ -1,12 +1,21 @@
 import streamlit as st
 import os
 from pinecone import Pinecone
-from sentence_transformers import SentenceTransformer
 from typing import List, Dict
+import logging
+
+# Compatibility shim for older huggingface_hub versions
+try:
+    from huggingface_hub import cached_download
+except Exception:
+    from huggingface_hub import hf_hub_download as cached_download
+    import huggingface_hub
+    huggingface_hub.cached_download = hf_hub_download
+
+from sentence_transformers import SentenceTransformer
 import re # For parsing timestamp and extracting video ID
 import streamlit.components.v1 as components # For embedding HTML
 from openai import OpenAI # Import OpenAI library
-import logging
 
 # Setup logging
 logging.basicConfig(level=logging.INFO)
@@ -84,7 +93,8 @@ def generate_youtube_embed_html(youtube_url: str, timestamp: str) -> str | None:
 PINECONE_API_KEY = os.getenv("PINECONE_API_KEY","pcsk_PZHLK_TRAvMCyNmJM4FKGCX7rbbY22a58fhnWYasx1mf3WL6sRasoASZXfsbnJYvCQ13w") # Load from environment variable
 PINECONE_ENV = os.getenv("PINECONE_ENV", "us-east-1")
 INDEX_NAME = "video-embeddings"
-EMBEDDING_MODEL = SentenceTransformer('sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2')
+# Use a lightweight multilingual model that works well with Korean
+EMBEDDING_MODEL_NAME = "intfloat/multilingual-e5-small"
 
 # OpenAI 설정
 OPENAI_API_KEY = "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"
@@ -109,8 +119,8 @@ def init_pinecone():
 def load_embedding_model():
     """Sentence Transformer 모델을 로드합니다."""
     try:
-        model = SentenceTransformer(EMBEDDING_MODEL)
-        logger.info(f"Successfully loaded embedding model: {EMBEDDING_MODEL}")
+        model = SentenceTransformer(EMBEDDING_MODEL_NAME)
+        logger.info(f"Successfully loaded embedding model: {EMBEDDING_MODEL_NAME}")
         return model
     except Exception as e:
         st.error(f"임베딩 모델 로딩 중 오류 발생: {e}")

requirements.txt

@@ -2,6 +2,7 @@ streamlit
 pinecone
 numpy<2.0
 pandas
-openai 
+openai
+huggingface-hub
 sentence-transformers==2.2.2
 torch==2.2.0