scikit-build-core
llama-cpp-python
llama-cpp-agent>=0.2.10
spaces
transformers>=4.37.0  # Core library for Hugging Face models [[7]][[10]]
torch>=2.1.0          # Required for GPU acceleration [[6]][[10]]
gradio>=4.0.0         # For the web interface [[10]]
huggingface_hub       # Model downloading and authentication [[9]]