Issues: pytorch/pytorch
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.
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
Author
Label
Projects
Milestones
Assignee
Sort
Issues list
dataloader crashes after several epochs if the trained model contains triton-based operators
#126620
opened May 18, 2024 by
flishwang
compile inductor falling operator ‘at::vec::CPU_CAPABILITY::VectorizedN<long int, 2>’ and ‘int’
oncall: pt2
#126619
opened May 18, 2024 by
johnnv1
inductor compile failling for batched tensor on fuse
oncall: pt2
#126617
opened May 18, 2024 by
johnnv1
[ONNX] export() with dynamic shapes fails when only part of input dimensions are dynamic
#126607
opened May 18, 2024 by
borisfom
UserWarning: Plan failed with a cudnnException: CUDNN_BACKEND_EXECUTION_PLAN_DESCRIPTOR
#126605
opened May 18, 2024 by
tjasmin111
[Pipelining] Support for uneven microbatch sizes
module: pipelining
Pipeline Parallelism
oncall: distributed
Add this issue/PR to distributed oncall triage queue
#126600
opened May 18, 2024 by
wconstab
DCP sees 1/2 of the expected size of each tensor in 3D parallel
#126595
opened May 18, 2024 by
wconstab
RuntimeError when using Adam(fused=True) with torch.compile
#126585
opened May 17, 2024 by
mitchellgoffpc
codegen error on .item() as a Triton kernel arg
module: aotinductor
aot inductor
triaged
This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#126574
opened May 17, 2024 by
desertfire
torch.compiler.allow_in_graph
does not create a call_module
op in fx.Graph in torch 2.3.0
#126566
opened May 17, 2024 by
kilianyp
[RFC] Deprecation support for Amazon Linux 2 support for PyTorch Release 2.5
#126551
opened May 17, 2024 by
atalman
Error: Exporting the operator 'aten::searchsorted' to ONNX opset version 17
#126549
opened May 17, 2024 by
CaioDaumann
2D TP+FSDP with device mesh
oncall: distributed
Add this issue/PR to distributed oncall triage queue
release notes: distributed (fsdp)
release notes category
triaged
This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#126548
opened May 17, 2024 by
ad8e
[AOTI][UX] One has no way of knowing whether they need to load DSO as CPU or CUDA runner
module: aotinductor
aot inductor
oncall: pt2
triaged
This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#126547
opened May 17, 2024 by
malfet
ncclCommWatchdog always terminates the process and prevents error handling if CUDA context is corrupted
#126544
opened May 17, 2024 by
szmigacz
Inductor: Codegen for sympy Trunc is incorrect
module: dynamic shapes
module: inductor
oncall: pt2
triaged
This issue has been looked at a team member, and triaged and prioritized into an appropriate module
#126537
opened May 17, 2024 by
ezyang
[Inductor] Masked
tl.load
operations should explicitly include other
if the masked out values are expected to be used
#126535
opened May 17, 2024 by
alexbaden
torch.set_default_device
does not change torch.Tensor().device
#126533
opened May 17, 2024 by
rafaol
Previous Next
ProTip!
Adding no:label will show everything without a label.