2024-03-05 03:16:35 +08:00
|
|
|
Usage:
|
|
|
|
|
|
2024-03-07 20:26:31 +08:00
|
|
|
- `pretrain.sh`: do pre-train (optional)
|
2024-03-31 18:29:50 +08:00
|
|
|
- `sft.sh`: do supervised fine-tuning
|
2024-03-07 20:26:31 +08:00
|
|
|
- `reward.sh`: do reward modeling (must after sft.sh)
|
|
|
|
|
- `ppo.sh`: do PPO training (must after sft.sh and reward.sh)
|
|
|
|
|
- `dpo.sh`: do DPO training (must after sft.sh)
|
2024-03-31 18:29:50 +08:00
|
|
|
- `orpo.sh`: do ORPO training
|
2024-03-07 20:26:31 +08:00
|
|
|
- `predict.sh`: do predict (must after sft.sh and dpo.sh)
|
