rxqy

Adding new datasets from CVPR2018

gpustat reports only the first program on nv driver 535

5

## TL;DR NVIDIA Driver >= 535.43, < 535.98 are broken. Avoid buggy driver and upgrade your NVIDIA driver software to higher (or lower) versions. ## Original bug report Hi, we...

`FullSync` gives distributed not initialized error for old `DataLoader`

5

### 🐛 Describe the bug Hi, here we are training in webdataset format with torchdata. Everything else works fine without fullsync, just the training hangs at the last iteration. So...

Hi, do we need to manual launch the vllm server for the grpo demo? I used the training command from the jupyter notebook, but soon get the following outputs, and...