openrl/examples/nlp at main · OpenRL-Lab/openrl

Name	Name	Last commit message	Last commit date
parent directory ..
README.md	README.md
chat.py	chat.py
chat_6b.py	chat_6b.py
ds_config.json	ds_config.json
eval_ds_config.json	eval_ds_config.json
nlp_ppo.yaml	nlp_ppo.yaml
nlp_ppo_ds.yaml	nlp_ppo_ds.yaml
train_ppo.py	train_ppo.py

Name

Last commit message

Last commit date

How to Use

Users can train the dialog task via:

python train_ppo.py --config nlp_ppo.yaml

Users can train the dialog task with deepspeed via:

deepspeed train_ppo.py --config nlp_ppo_ds.yaml

After the training, users can chat with the agent via:

python chat.py