-
### 🚀 The feature, motivation and pitch
Currently Pytorch DDP supports a few default compression hooks, namely fp16, bf16, and PowerSGD. However, there are a myraid of other compression types that …
-
## Problem
Currently, many meta tests are skipped for foreach ops. See https://github.com/pytorch/pytorch/blob/b18d8d4595aa6e0768eedd5fc7d4a4402c567181/torch/testing/_internal/common_methods_invocati…
-
Hi, In **neural-style**, inappropriate dependency versioning constraints can cause risks.
Below are the dependencies and version constraints that the project is using
```
numpy
Pillow
scipy
…
-
## Issue description
Currently, resizing sparse compressed tensors works correctly only in the CSR layout and non-hybrid case. Other cases are silently broken and the resize operation produces inva…
pearu updated
2 years ago
-
# Compromise design for extensions
* [x] Proposed
* [ ] Prototype: Not Started
* [ ] Implementation: Not Started
* [ ] Specification: Not Started
## Summary
[summary]: #summary
This proposal mixes …
-
# The issue
I am trying to understand how gradients are computed for Burgers, implemented by `FDM_Burgers()` in `train_utils/losses.py`, as pasted below:
```python
def FDM_Burgers(u, v, D=1):
…
-
### Request description
Looking for support for primarily int4 (also int2 and int3) weight quantization as described in GPTQ source https://arxiv.org/abs/2210.17323, with code implementation for ll…
-
## 🐛 Bug
scripting test failed in https://github.com/pytorch/pytorch/pull/52316 with error
```
Falsifying example: test_fn(, batch_size=1, dilation=1, groups=1, in_channels_per_group=2, kernel=1,…
-
This contains some notes from @renxida on onnx.If
@PhaneeshB is currently working on onnx.Loop
Find the if in [retinanet_resnet50_fpn_vaiq_int8 model support](https://github.com/nod-ai/SHARK-Tes…
-
## 🐛 Bug
PyTorch deadlocks when using queues and events with `mp.spawn`.
## To Reproduce
Minimal example:
```
import torch.multiprocessing as mp
def main_worker(gpu, queue, event):
pr…