Skip to content

Issues: pytorch/pytorch

[v.2.5.0] Release Tracker
#135522 opened Sep 9, 2024 by kit1980
Open 17
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Clear current search query, filters, and sorts
442 Open 536 Closed
442 Open 536 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

Python's REPL trying to do invoke .T / .mH / .mT on 0-dim tensor when doing tab-completion for some reason (default install of Python on new Ubuntu 24.04) needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user release notes: python_frontend python frontend release notes category triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#135693 opened Sep 11, 2024 by vadimkantorov
8
Bug with "make latexpdf" module: docs Related to our documentation, both in docs/ and docblocks needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#135420 opened Sep 7, 2024 by raisin
1
Problem with conversion of torch model (htdemucs) to onnx module: onnx Related to torch.onnx needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#135343 opened Sep 6, 2024 by gianlourbano
@justinchuby
2
signbit on bf16 tensor causes compile error module: inductor needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user oncall: pt2 triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#135341 opened Sep 6, 2024 by jthakurH
1
torch.onnx.export Doesn't allow dynamic shapes when updating a tensor bug module: onnx Related to torch.onnx needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#135233 opened Sep 5, 2024 by Akshat-Tripathi
@justinchuby
5
ModuleNotFoundError: No module named 'torch.fx.experimental.shape_inference' module: fx needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#135196 opened Sep 5, 2024 by PHLens
1
@yf225
8
torch.unique behaves strange for large input arrays on Windows module: correctness (silent) issue that returns an incorrect result silently module: windows Windows support for PyTorch needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#135019 opened Sep 3, 2024 by SkafteNicki
2
[bug] import torch failed after install the cpu version of pytorch module: binaries Anything related to official binaries that we release to users needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#134993 opened Sep 3, 2024 by nctu6
2
On Kaggle : libcusparse.so.12: undefined symbol: __nvJitLinkComplete_12_4, version libnvJitLink.so.12 module: binaries Anything related to official binaries that we release to users module: cuda Related to torch.cuda, and CUDA support in general module: regression It used to work, and now it doesn't module: sparse Related to torch.sparse needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#134929 opened Sep 2, 2024 by FurkanGozukara
10
CMake Error: When installing PyTorch from source, CUDA not being detected. module: build Build system issues needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#134331 opened Aug 23, 2024 by efsotr
2
[Intel CPU] Same CPU operations on Torch 2.3+ many times slower those on Torch 2.1 module: cpu CPU specific problem (e.g., perf, algorithm) module: intel Specific to x86 architecture module: performance Issues related to performance, either of kernel code or framework glue module: regression It used to work, and now it doesn't needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#134198 opened Aug 22, 2024 by Kite0011
7
torch::tensor function does not work module: cpp Related to C++ API needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#133860 opened Aug 19, 2024 by walkacross
2
pytorch/aten/src/ATen/cuda/CUDABlas.cpp:1422:28: error module: build Build system issues module: cuda Related to torch.cuda, and CUDA support in general needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#133808 opened Aug 18, 2024 by spamidig
1
Errors with torch.compile after upgrading to 2.4.0 high priority module: aotdispatch umbrella label for AOTAutograd issues module: dynamic shapes module: inductor module: pt2-dispatcher PT2 dispatcher-related issues (e.g., aotdispatch, functionalization, faketensor, custom-op, needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user oncall: pt2 oncall: transformer/mha Issues related to Transformers and MultiheadAttention triage review
#133571 opened Aug 15, 2024 by GLivshits
@eellison @bdhirsh
20
Floating point exception on H20 module: cuda Related to torch.cuda, and CUDA support in general needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#133548 opened Aug 15, 2024 by songh11
2
[flexAttention] compiled flexAttention with sliding window runs in 32bit float precision but crashes on 16bit float & mixed precision module: crash Problem manifests as a hard crash, as opposed to a RuntimeError module: flex attention module: higher order operators torch.cond and similar module: multi-headed-attention module: pt2-dispatcher PT2 dispatcher-related issues (e.g., aotdispatch, functionalization, faketensor, custom-op, needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user oncall: pt2 triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#133431 opened Aug 14, 2024 by cathalobrien
3
Issue Exporting PyTorch Model with pytorch3d.ops.ball_query to ONNX: Bug or Compatibility Problem? module: onnx Related to torch.onnx needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user oncall: export
#133227 opened Aug 12, 2024 by florentjousse
1
Inconsistent results of Conv2d in pytorch between CPU and CUDA module: convolution Problems related to convolutions (THNN, THCUNN, CuDNN) module: cuda Related to torch.cuda, and CUDA support in general needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#133093 opened Aug 9, 2024 by VadimShabashov
1
JITerator: Torch.prod produces RuntimeError: CUDA driver error: invalid argument module: jiterator needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user oncall: pt2 triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#132756 opened Aug 6, 2024 by ezyang
关于yolov5在mac设备上使用pytorch的mps加速出现的各种问题 module: mps Related to Apple Metal Performance Shaders framework needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#132226 opened Jul 31, 2024 by xxxkkw
6
[DEBUG] Strange behavior observed with PyTorch 2.4.0 + Windows + CPU inference high priority module: cpu CPU specific problem (e.g., perf, algorithm) module: regression It used to work, and now it doesn't module: windows Windows support for PyTorch needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#131958 opened Jul 26, 2024 by Burhan-Q 2.4.1
@xuhancn
40
RuntimeError: Expected grad_output.numel() <= std::numeric_limits<int32_t>::max() to be true, but got false. module: autograd Related to torch.autograd, and the autograd engine in general module: error checking Bugs related to incorrect/lacking error checking module: nn Related to torch.nn module: 64-bit Problems related to incorrectly using 32-bit integers when 64-bit is needed (e.g., 8G tensors) needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#131185 opened Jul 19, 2024 by emmaking-smith
3
torch.cuda.empty_cache() doesn't work correctly for the first instance of nn.Module set to cuda (and a potential method to deal with that). module: cuda Related to torch.cuda, and CUDA support in general module: memory usage PyTorch is using more memory than it should, or it is leaking memory needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#130728 opened Jul 15, 2024 by kangchengX
7
Converting a numpy array of size larger than 32,768 to a tensor causes a segmentation fault module: crash Problem manifests as a hard crash, as opposed to a RuntimeError module: macos Mac OS related issues module: numpy Related to numpy support, and also numpy compatibility of our operators needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#130708 opened Jul 15, 2024 by chadeos
7
Fail to offload FSDP model weights and optimizer states without using CPUOffload(offload_params=True) needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user oncall: distributed Add this issue/PR to distributed oncall triage queue triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#130530 opened Jul 11, 2024 by PeterSH6
1
ProTip! no:milestone will show everything without a milestone.