kdyang
kdyang
After some experiments, it is found out that this basline here should use bottleneck as the block type, so that the model can reach the result. Right?
@JiexiPinkman 您好！请问您下载数据的时候，从注册账号到申请通过大约经过了多长时间？谢谢！
Yeah, agree with @KaleabTessera. BTW, this makes the net run a little slower.
I think it would be better if the author updates the requirements in README.
> cuda 10.2 python 3.7 certifi 2022.9.24 flit_core 3.6.0 mkl-fft 1.3.1 mkl-random 1.2.2 mkl-service 2.4.0 numpy 1.21.5 Pillow 9.2.0 pip 22.2.2 setuptools 65.5.0 six 1.12.0 torch 1.8.1 torchvision 0.9.1 typing_extensions...
@TingheZhang Thanks.