Open Mr-IT007 opened 1 year ago
For training:
Training with RRHF You can train your own model with generated or released datasets using the script train.sh, please note that the training process requires 8*A100 80GB GPUs, bf16 and FSDP. In the future, we will try efficient training methods such as LoRA or Prefix-tuning or Adapter to lower the computational resource requirements.
bash ./train.sh
For using Wombat: Use recover_wombat_7b.sh and single_sentence_inference.py
How to use it. Is there some code examples?