# MoodSyncAI runtime dependencies # Tested on Python 3.10–3.13 (Windows / Linux / macOS, CPU) # # Install: # pip install --upgrade pip # pip install -r requirements.txt --only-binary=:all: # # The --only-binary flag forces wheels, which avoids needing Rust/MSVC to # compile tokenizers on Python 3.13. # --- UI --- gradio>=4.44,<6 # --- Deep-learning stack --- torch>=2.2 torchvision>=0.17 transformers>=4.46 tokenizers>=0.20 sentencepiece>=0.2 accelerate>=0.30 safetensors>=0.4 # --- Vision / data --- pillow>=10 numpy>=1.26 opencv-python>=4.9 plotly>=5.20 pandas>=2.0 # --- Audio (Whisper + video audio extraction) --- imageio-ffmpeg>=0.5 # bundles ffmpeg binary; no system install required soundfile>=0.12 librosa>=0.10