-
### 🐛 Describe the bug
@fmassa
First of all, I want to express my gratitude to the team for creating and maintaining this amazing open-source project. It’s a fantastic tool, and I deeply apprecia…
-
### 🐛 Describe the bug
I am following [the qualcomm tutorial](https://github.com/pytorch/executorch/blob/17ad8d3f69d365a77dda590e1d40497befaa6b69/examples/demo-apps/android/LlamaDemo/docs/delegates…
-
### 🐛 Describe the bug
getting inconsistent results of ```torch.Tensor.logcumsumexp``` between CPU and GPU
```python #
import torch
self = torch.tensor([[2.4375, -0.3340, -0.3711]], dtype=torch.…
-
### 🐛 Describe the bug
I am trying to install cuda enabled torchcodec from the documentation. https://github.com/pytorch/torchcodec?tab=readme-ov-file#installing-cuda-enabled-torchcodec
I could ab…
-
### 🐛 Describe the bug
### Background
The FCOS object detector accepts `**kwargs`, of which is the `_skip_resize` flag, to be passed directly to `GeneralizedRCNNTransform` at FCOS init, [here](h…
-
### 🐛 Describe the bug
``` python
import os
import torch
from torch._subclasses import FakeTensorMode, CrossRefFakeMode
import torch.nn.functional as F
fake_mode = FakeTensorMode()
cross_ref_…
-
### 🐛 Describe the bug
When testing flex attention with the mask from issue #141126, I found that apart from being incorrect it is also significantly slower than torch's generic `scaled_dot_product…
-
### 🐛 Describe the bug
PyTorch 2.5.0 and latest 2.5.1 releases revive #111469 and so breaks when importing.
Temporary fix: getting back to 2.4.1
### Versions
PyTorch version: N/A
Is debug b…
-
### 🐛 Describe the bug
Under specific inputs, `replication_pad1d_backward` triggered a crash.
```python
import torch
grad_output = torch.full((9, 9, 0,), 0, dtype=torch.float)
self = torch.full…
-
### 🐛 Describe the bug
It is found that the vectorized kernels are not performing well. For instance, copy_() is just utilizing only 40-50% of the theoretical bandwidth.
### Versions
Collecting en…