torch
trl
openai
transformers
peft
datasets
pandas
jinja2
tqdm
pyyaml
gradio_client
python-dotenv
wheel
python-dotenv
tensorboard
numpy
alfworld[full]
sentencepiece
wandb
requests
sglang[all]
# pip install flashinfer -i https://flashinfer.ai/whl/cu121/torch2.4/
# from webshop
# gym==0.24.0
# bs4
# selenium==4.2.0
# cleantext==1.1.4
# rank_bm25==0.2.2