# Vetrina — Python dependencies # Pinned to the exact versions the project was developed and evaluated with # (Python 3.12, macOS Apple Silicon). Adjust torch for your platform/CUDA if needed. # --- LangGraph orchestration --- langgraph==1.0.9 langchain==1.2.10 langchain-core==1.2.15 langchain-community==0.4.1 langchain-openai==1.1.10 langchain-huggingface==1.2.0 # --- LLM (GPT-4o-mini: translation + greeting node) --- openai==2.23.0 # --- Embeddings + vector search --- sentence-transformers==5.2.3 # paraphrase-multilingual-MiniLM-L12-v2 faiss-cpu==1.13.2 # --- Visual recommending (Trendyol DinoV2, 256d) --- torch==2.10.0 transformers==4.57.6 pillow==12.1.1 # --- Dataset loading (HuggingFace ecommerce-intent-routing, for the Router) --- datasets==4.5.0 # --- Web interface --- flask==3.1.3 # --- Utilities --- python-dotenv==1.2.1 numpy==2.4.2 # --- Recommender: LightFM (hybrid collaborative + content-based) --- # LightFM has no prebuilt wheels for Python 3.12 and must be built from source # with a small Cython compatibility patch. See the README section # "Installing LightFM on Python 3.12" for step-by-step instructions. # lightfm==1.17