botbw

Hi, is it possible to do back-propagation on the model ensembled using WBF? Or would we have a pytorch implementation for WBF?

A lot of errors when running `hatch run train`

6

Hi, I cloned the latest master but somehow ran into a series of problems when running the demo command. Could you please have a try on `hatch run train` and...

Further explanation for `batch_isend_irecv`

1

### 📚 The doc issue [here](https://github.com/pytorch/pytorch/blob/ce503c1b40207dab770c28cbd4568cd9e105277b/torch/distributed/distributed_c10d.py#L2067) the doc string says the function will return `dist.Work` by calling corresponding ops, however the returned reqs might not be "corresponding" (e.g. `batch_isend_irect([op1, op2])`...

### 🚀 The feature, motivation and pitch Hey there! Currently `torch.distributed._tensor.DeviceMesh` only supports 1-D sub-meshes, will it be possible to manipulate it as an NDArray and generate N-dim sub-meshes? For...

## 📌 Checklist before creating the PR - [ ] I have created an issue for this PR for traceability - [x] The title follows the standard format: `[doc/gemini/tensor/...]: A...

[gemini] async grad chunk reduce (all-reduce&reduce-scatter)

1

## 📌 Checklist before creating the PR - [ ] I have created an issue for this PR for traceability - [x] The title follows the standard format: `[doc/gemini/tensor/...]: A...

[Gemini] Prefetch next chunk before each op

1

## 📌 Checklist before creating the PR - [ ] I have created an issue for this PR for traceability - [ ] The title follows the standard format: `[doc/gemini/tensor/...]:...

Better to fix the version of `face_alignment` and `numpy` in `pose_requirements.txt`