Closed yjdy closed 9 months ago
I'm trying to retrain the autoregression head with your train code on vicuna 7b I have 8 v100 with 32G. But even I set bs=1, it still out of memory.
My environment is as follow: CUDA 11.7 python 3.10 pytorch 2.1.2 transformers 4.37.2 accelerate 0.27.2
By the way, I used code of bsne1 branch
Best regards
how do you solve it ?
I'm trying to retrain the autoregression head with your train code on vicuna 7b I have 8 v100 with 32G. But even I set bs=1, it still out of memory.
My environment is as follow: CUDA 11.7 python 3.10 pytorch 2.1.2 transformers 4.37.2 accelerate 0.27.2
By the way, I used code of bsne1 branch
Best regards