Updating README
#2
by
mlinmg
- opened
I would like to know how to use the weights you provided. Also could I follow the alpaca-lora guide on merging the weights(I would like to finetune starting from your checkpoints)? What should be use, the hf one or the pytorch one? Thx
I have already-merged weights you can use here: https://huggingface.co/TheBloke/gpt4-alpaca-lora-30b-HF
I merged them using this script: https://github.com/tloen/alpaca-lora/blob/main/export_hf_checkpoint.py