nv-dlasalle
nv-dlasalle
**Is your feature request related to a problem? Please describe.** In HPS, for GNNs it would be great to have a python lookup in function which could take in indices...
## 🚀 Feature It seems unlikely that a users calling `dgl.remove_self_loop()` (https://docs.dgl.ai/generated/dgl.remove_self_loop.html?highlight=remove_self_loop#dgl.remove_self_loop) would desire the side effect of batch information being removed. It can also make code confusing to read:...
## 🐛 Bug In the dataloader, we do nothing with `pin_prefetcher` if the graph is not a DGLHeteroGraph object, and thus it defaults to `None`: https://github.com/dmlc/dgl/blob/master/python/dgl/dataloading/dataloader.py#L776 Here `pin_prefetcher` gets passed...
## Description Adds support for storing graph features partitioned in GPU memory. This doesn't implement new functionality, but rather better exposes that which was already implemented in the NCCL based...
[bug] dgl.dataloading.DataLoader forwards args to torch.utils.data.DataLoader but many are not valid
## 🐛 Bug Currently, the DGL's DataLoader is a child of PyTorch's DataLoader, and passes `kwargs` to it, however DGL's dataloader does several things under the hood which breaks many...
## 🐛 Bug When a DGLBlock is cloned, the duplicate srctype and dsttype get re-merged in the metagraph, which makes it unusable. This appears to be the result of the...
## 🚀 Feature Starting in PyTorch 1.11, `send` and `recv` are supported by `torch.distributed`, which largely means we no longer need direct access to NCCL. To do this, we should...
## 🐛 Bug See PyTorch issue (thanks to @as51340 for raising this): https://github.com/pytorch/pytorch/issues/83733 And reproducible code here: https://gist.github.com/as51340/69691482e2a204e2a8e12c07954e0553 From what I can tell the issue is caused by the cast...