Skip to content

Enable CUTLASS grouped GEMM for pretraining wgrad on GB200 and H100 (resubmit) #6077

Enable CUTLASS grouped GEMM for pretraining wgrad on GB200 and H100 (resubmit)

Enable CUTLASS grouped GEMM for pretraining wgrad on GB200 and H100 (resubmit) #6077

Triggered via pull request September 23, 2025 16:18
Status Failure
Total duration 2h 27m 28s
Artifacts 1
generate-matrix  /  generate
7s
generate-matrix / generate
filter-matrix
6s
filter-matrix
Matrix: pytorch/FBGEMM / build
Matrix: pytorch/FBGEMM / upload / upload
Fit to window
Zoom out
Zoom in

Annotations

8 errors and 1 warning
pytorch/FBGEMM / build-manywheel-py3_10-cuda13_0
The self-hosted runner lost communication with the server. Verify the machine is running and has a healthy network connection. Anything in your workflow that terminates the runner process, starves it for CPU/Memory, or blocks its network access can cause this error.
pytorch/FBGEMM / build-manywheel-py3_10-cuda12_6
The operation was canceled.
pytorch/FBGEMM / build-manywheel-py3_10-rocm6_4
The operation was canceled.
pytorch/FBGEMM / build-manywheel-py3_10-rocm6_3
The self-hosted runner lost communication with the server. Verify the machine is running and has a healthy network connection. Anything in your workflow that terminates the runner process, starves it for CPU/Memory, or blocks its network access can cause this error.
pytorch/FBGEMM / upload / upload-manywheel-py3_10-rocm6_4
Unable to download artifact(s): Artifact not found for name: pytorch_FBGEMM__3.10_rocm6.4_x86_64 Please ensure that your artifact is not expired and the artifact was uploaded using a compatible version of toolkit/upload-artifact. For more information, visit the GitHub Artifacts FAQ: https://github.com/actions/toolkit/blob/main/packages/artifact/docs/faq.md
pytorch/FBGEMM / upload / upload-manywheel-py3_10-rocm6_3
Unable to download artifact(s): Artifact not found for name: pytorch_FBGEMM__3.10_rocm6.3_x86_64 Please ensure that your artifact is not expired and the artifact was uploaded using a compatible version of toolkit/upload-artifact. For more information, visit the GitHub Artifacts FAQ: https://github.com/actions/toolkit/blob/main/packages/artifact/docs/faq.md
pytorch/FBGEMM / upload / upload-manywheel-py3_10-cuda13_0
Unable to download artifact(s): Artifact not found for name: pytorch_FBGEMM__3.10_cu130_x86_64 Please ensure that your artifact is not expired and the artifact was uploaded using a compatible version of toolkit/upload-artifact. For more information, visit the GitHub Artifacts FAQ: https://github.com/actions/toolkit/blob/main/packages/artifact/docs/faq.md
pytorch/FBGEMM / upload / upload-manywheel-py3_10-cuda12_6
Unable to download artifact(s): Artifact not found for name: pytorch_FBGEMM__3.10_cu126_x86_64 Please ensure that your artifact is not expired and the artifact was uploaded using a compatible version of toolkit/upload-artifact. For more information, visit the GitHub Artifacts FAQ: https://github.com/actions/toolkit/blob/main/packages/artifact/docs/faq.md
filter-matrix
The `python-version` input is not set. The version of Python currently in `PATH` will be used.

Artifacts

Produced during runtime
Name Size Digest
pytorch_FBGEMM__3.10_cu128_x86_64
48 MB
sha256:fca66fcc541227408087b3a8b68762c1c8cfc1c9f90ba66ab138cec98050495a