Skip to content

Issues: pytorch/pytorch

[v2.3.1] Release Tracker
#125425 opened May 2, 2024 by atalman
Open 4
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

12,957 Open 28,768 Closed
12,957 Open 28,768 Closed
Author
Filter by author
Label
Filter by label
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Milestones
Filter by milestone
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

[BE][DTensor] get rid of @with_comms decorator for tests, and use run_test to init pg better-engineering Relatively self-contained tasks for better engineering contributors oncall: distributed Add this issue/PR to distributed oncall triage queue
#125800 opened May 8, 2024 by wanchaol
1
_will_engine_execute_node expects an grad_fn, but got CppFunction
#125797 opened May 8, 2024 by soulitzer
Method for loading a distributed checkpoint into a single state_dict is being deprecrated without alternative, request to make it possible to keep that feature
#125777 opened May 8, 2024 by RuABraun
DISABLED test_binary_op_list_slow_path__foreach_div_cuda_uint8 (__main__.TestForeachCUDA) skipped Denotes a (flaky) test currently skipped in CI.
#125775 opened May 8, 2024 by huydhn
2
DISABLED test_parity__foreach_abs_fastpath_inplace_cuda_complex128 (__main__.TestForeachCUDA) skipped Denotes a (flaky) test currently skipped in CI.
#125774 opened May 8, 2024 by huydhn
1
RFC: Integrate Arm Compute Library (ACL) into PyTorch as a submodule and add build support in setup.py
#125767 opened May 8, 2024 by snadampal
torch C10 half type not compatible with nvidia __half atomicAdd
#125766 opened May 8, 2024 by SamKG
dynamo test (test_model_output.py) failing on cpu devices because of cuda hardcoding for the device
#125760 opened May 8, 2024 by snadampal
1
1
Excution difference between Mac and Linux module: macos Mac OS related issues needs reproduction Someone else needs to try reproducing the issue given the instructions. No action needed from user
#125754 opened May 8, 2024 by NicolayP
2
ONNX export success, but Load model from onnx failed oncall: pt2
#125753 opened May 8, 2024 by xiaocode
Support third-party devices emit an range for each autograd operator module: PrivateUse1 private use oncall: profiler profiler-related issues (cpu, gpu, kineto) OSS contribution wanted PR from open source contributors welcome to solve this issue.
#125752 opened May 8, 2024 by 1274085042
'MultiHeadAttention.attention' is being compiled since it was called from 'MultiHeadAttention.forward' oncall: jit Add this issue/PR to JIT oncall triage queue
#125749 opened May 8, 2024 by SRbone
DefaultCPUAllocator: not enough memory in matrix multiplication broadcasting
#125746 opened May 8, 2024 by LeiZhang-116-4
torch.compile error: Attempting to broadcast a dimension of length 2 at -1 oncall: pt2
#125745 opened May 8, 2024 by syheliel
Improve oneDNN memory alloction performance for pytorch Windows intel This tag is for PR from Intel module: cpu CPU specific problem (e.g., perf, algorithm) module: performance Issues related to performance, either of kernel code or framework glue module: windows Windows support for PyTorch triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#125744 opened May 8, 2024 by xuhancn
@xuhancn
1
[DTensor][Tensor Parallel] transformer test numerical issue when dtype=torch.float32 module: dtensor distributed tensor tag 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
#125741 opened May 8, 2024 by tianyu-l
[Distributed Checkpoint] When loading FSDP sharded checkpointing each rank needs all the checkpointing files oncall: distributed Add this issue/PR to distributed oncall triage queue
#125740 opened May 8, 2024 by bigning
1
[FSDP2] _sharded_param_data is sitll on meta while sharded_param moved to cuda after calling initialize_parameters() module: fsdp triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#125738 opened May 8, 2024 by weifengpy
2
[ARM] Vectorized<half>::loadu(x, 8) yields slow code if -fno-unsafe-math-optimizations are used module: arm Related to ARM architectures builds of PyTorch. Includes Apple M1 module: performance Issues related to performance, either of kernel code or framework glue oncall: cpu inductor CPU Inductor issues for Intel team to triage oncall: pt2
#125735 opened May 8, 2024 by malfet
@malfet
3
☂️ torch.compile generates slower code for LLMs than eager on ARM platform (M1/AARCH64) module: arm Related to ARM architectures builds of PyTorch. Includes Apple M1 module: performance Issues related to performance, either of kernel code or framework glue oncall: cpu inductor CPU Inductor issues for Intel team to triage oncall: pt2
#125734 opened May 8, 2024 by malfet
torch._inductor.config.max_autotune_gemm_backends = "TRITON" crashes with Convolution layer module: inductor oncall: pt2 triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#125728 opened May 8, 2024 by amodab01
Migrate multiple/custom runner labels before deprecation module: ci Related to continuous integration triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#125721 opened May 7, 2024 by ZainRizvi
Support dynamo tracing weakref obj module: dynamo oncall: pt2
#125720 opened May 7, 2024 by ydwu4
torch.compile and complex numbers oncall: pt2
#125718 opened May 7, 2024 by cw-tan
2
KINETO_USE_DAEMON causing issues oncall: profiler profiler-related issues (cpu, gpu, kineto)
#125705 opened May 7, 2024 by gjit-juniper
@briancoutinho
ProTip! Updated in the last three days: updated:>2024-05-05.