huggingface_hub==0.25.2
numpy
tqdm
torch
peft
transformers>=4.39.1
datasets
evaluate
scikit-learn