Inference or finetuning with RTX 3000 / A100 on CUDA 11

lucidrains / DALLE-pytorch

Implementation / replication of DALL-E, OpenAI's Text to Image Transformer, in Pytorch

MIT License

5.57k stars 642 forks source link

Inference or finetuning with RTX 3000 / A100 on CUDA 11 #433

Open tianjianh opened 2 years ago

tianjianh commented 2 years ago

Hey I am just wondering is there way to load any pre-trained models with A100 on CUDA 11? It seems that deepspeed==0.4.4 and triton==0.4.2 do not work with CUDA 11 but pre-trained models require those old versions of deepspeed and triton. Thanks in advance!