frankzhao112
frankzhao112
I get the same question. Does the traing code here only support llama or opt model? Can we finetune the bloom using its official training framework with stanford_alpaca data?
No, I have the same issue. Do u know BELLE, they use bloom as the base model instead of llama.
Are u Chinese? 说汉语呗
u can check bloom training code in bloom github. Bloom already opens its trainning code, I think u can find tranning code there.