-
### Your current environment
```text
The output of `python collect_env.py`
```
Collecting environment information...
PyTorch version: 2.1.2+cu121
Is debug build: False
CUDA used to build PyTorc…
-
### Bug description
The CPU meter shows the values since boot, and the RAM meter ignores the caches...
This makes the available resource information rather misleading, for the casual user.
I …
-
I'm trying to play around with the synth project from the 2023 sycl conference and I'm getting a build error when trying to use the latest microzig version. I've made some changes to the build script …
tanj updated
2 months ago
-
### 🐛 Describe the bug
When using dcp.save or dcp.async_save as torch.save alternative and then loading state_dicts back, non-tensor values is not updated as expected, for example(edited from [async_…
-
### 🐛 Describe the bug
Runtime Error noticed while running llama2 on aarch64 Grace super server, onednn built on the native system from https://github.com/oneapi-src/oneDNN/blob/main/README.md and A…
-
## 🚀 Feature
Modeling of accessing shape attribute of Tensor/TensorProxy has raising up to discussion in separate PRs & discussions.
We are trying to debate on whether we would want to lift shap…
-
### Your current environment
```
PyTorch version: 2.3.1+cu121
Is debug build: False
CUDA used to build PyTorch: 12.1
ROCM used to build PyTorch: N/A
OS: Amazon Linux 2 (x86_64)
GCC version:…
-
### Component
server, agent
### Describe the bug
Hello. I recently migrated some CI jobs from Drone and noticed their running time increasing from 4 seconds to about two minutes.
The reason se…
-
### Is there an existing issue for this?
- [X] I have searched the existing issues and none matched.
### Operating System
Windows 10
### CPU
Intel (R) Core (TM) i5-1035G4 CPU @ 1.10GHz
### GPU
…
-
### 🐛 Describe the bug
`torch.export.export` in strict mode seems to produce an incorrect exported program when the exported module contains a python generator.
Here is a minimal reproducible ex…