linyu0219
linyu0219
> I had to change a lot of .cpp files replacing AT_CHECK with TORCH_CHECK. But it worked for me! Thankyou I use pytorch 1.7 and success made it.
wrong result example 2 
> No, not for this reason. I use 0.1 also has this problem. When I use cpu a pip install version, results is correct.
#3545 Maybe this error is same like mine. xnnpack is cpu , cpu is ok gpu is wrong. something error in gpu.
my old apex ok, it's behind 23 commit . git reset --hard 0c2c6eea
> Same error, pip3 install -v --no-cache-dir --global-option="--cpp_ext" --global-option="--cuda_ext" ./ > > Ubuntu 20.04, Cuda 11.1 > > P.S. Old Apex june 30 version 0.1 - OK. Useful!
thanks @JanuszL , I understand.
It's a big problem, has impact to other libs @richzhang https://github.com/Lightning-AI/lightning/issues/18712
> [你可以看看这个项目，不光支持了GPT-4,](https://github.com/n3d1117/chatgpt-telegram-bot) DALL·E和Whisper模型也支持 这个是api接口的gpt4，楼主需要的是plus的gpt4。 我也想要
https://github.com/Zulko/moviepy/issues/1625#issuecomment-973056577 fixed decorator problem