Upgraded Gradio. More resilient to cases where LLM calls do not return valid markdown tables (will reattempt with different temperature). Minor fixes
b9301bd
| pandas==2.2.3 | |
| gradio==5.20.1 | |
| spaces==0.31.0 | |
| boto3==1.35.71 | |
| pyarrow==18.1.0 | |
| openpyxl==3.1.3 | |
| markdown==3.7 | |
| tabulate==0.9.0 | |
| lxml==5.3.0 | |
| google-generativeai==0.8.3 | |
| html5lib==1.1 | |
| beautifulsoup4==4.12.3 | |
| rapidfuzz==3.10.1 | |
| torch==2.4.1 --extra-index-url https://download.pytorch.org/whl/cu121 | |
| #llama-cpp-python==0.2.90 --extra-index-url https://abetlen.github.io/llama-cpp-python/whl/cu121 | |
| # Specify exact llama_cpp wheel for huggingface compatibility | |
| https://github.com/abetlen/llama-cpp-python/releases/download/v0.2.90-cu121/llama_cpp_python-0.2.90-cp310-cp310-linux_x86_64.whl | |
| transformers==4.49.0 | |
| numpy==1.26.4 | |
| typing_extensions==4.12.2 |