DarkKnight_21th

torch의 DDP를 이용하여, huggingface의 kogpt2를 파인튜닝하면 GPU들에서 메모리 불균형이 일어납니다. torch의 DistributedDataSampler로 데이터 셋을 감싸줬고, DDP를 이용해서 학습을 시키면 메모리 불균형이 일어납니다. 왜 일까요? GPU는 RTX 4090 (24GB) 2대로 학습을...

because of size of model, i want to run the mdel in deplabv3_resnet50. but there are some error. how can i do that? do i change the structure of code?...

Operating System : Window 10 ( feat. error of Installing the mmcv-full)

1

Hi :) I have some errors for setting environments. I already installed the cuda (toolkit) 11.0.3, torch+cu110 and torch 1.7.0. But there is a same error with the issue (https://github.com/lhoyer/HRDA/issues/35)....

i wanna train on imagenet21k, not imagenet1k. so i downloaded the imagenet21k(winter) on the official site.(imagenet official site) and then i ran the "write_imagenet.sh" with default argument value (500 0.50...

[BUG] There is the error in timm/train.py when i use the Webdataset (timm/imagent-w21-wds in huggingface) with class map

9

**Describe the bug** A clear and concise description of what the bug is. i wanted to train the the parts of IN21k-winter class , so i made the class map...

**Is your feature request related to a problem? Please describe.** A clear and concise description of what the problem is. i think it is good to add the new image...