Skip to content

Improve performance of prefill mode FP8 Grouped Gemm #2931

Improve performance of prefill mode FP8 Grouped Gemm

Improve performance of prefill mode FP8 Grouped Gemm #2931

Triggered via pull request December 31, 2024 00:36
Status Failure
Total duration 27m 22s
Artifacts 20

fbgemm_gpu_ci_genai.yml

on: pull_request
Matrix: build_artifact
Matrix: test_and_publish_artifact
Fit to window
Zoom out
Zoom in

Annotations

20 errors and 1 warning
build_artifact (x86, linux.24xlarge, 3.10, 12.6.3, gcc)
Process completed with exit code 1.
build_artifact (x86, linux.24xlarge, 3.10, 12.6.3, clang)
Process completed with exit code 1.
build_artifact (x86, linux.24xlarge, 3.11, 12.6.3, gcc)
Process completed with exit code 1.
build_artifact (x86, linux.24xlarge, 3.12, 12.6.3, gcc)
Process completed with exit code 1.
build_artifact (x86, linux.24xlarge, 3.13, 12.6.3, gcc)
Process completed with exit code 1.
build_artifact (x86, linux.24xlarge, 3.9, 12.6.3, gcc)
Process completed with exit code 1.
build_artifact (x86, linux.24xlarge, 3.11, 12.6.3, clang)
Process completed with exit code 1.
build_artifact (x86, linux.24xlarge, 3.12, 12.6.3, clang)
Process completed with exit code 1.
build_artifact (x86, linux.24xlarge, 3.13, 12.6.3, clang)
Process completed with exit code 1.
build_artifact (x86, linux.24xlarge, 3.9, 12.6.3, clang)
Process completed with exit code 1.
test_and_publish_artifact (x86, linux.g5.4xlarge.nvidia.gpu, 3.12, 12.6.3, 12.4.1, gcc)
Unable to find an artifact with the name: fbgemm_gpu_nightly_genai_x86_gcc_py3.12_cu12.6.3.whl
test_and_publish_artifact (x86, linux.g5.4xlarge.nvidia.gpu, 3.10, 12.6.3, 12.4.1, gcc)
Unable to find an artifact with the name: fbgemm_gpu_nightly_genai_x86_gcc_py3.10_cu12.6.3.whl
test_and_publish_artifact (x86, linux.g5.4xlarge.nvidia.gpu, 3.12, 12.6.3, 12.4.1, clang)
Unable to find an artifact with the name: fbgemm_gpu_nightly_genai_x86_clang_py3.12_cu12.6.3.whl
test_and_publish_artifact (x86, linux.g5.4xlarge.nvidia.gpu, 3.9, 12.6.3, 12.4.1, clang)
Unable to find an artifact with the name: fbgemm_gpu_nightly_genai_x86_clang_py3.9_cu12.6.3.whl
test_and_publish_artifact (x86, linux.g5.4xlarge.nvidia.gpu, 3.9, 12.6.3, 12.4.1, gcc)
Unable to find an artifact with the name: fbgemm_gpu_nightly_genai_x86_gcc_py3.9_cu12.6.3.whl
test_and_publish_artifact (x86, linux.g5.4xlarge.nvidia.gpu, 3.13, 12.6.3, 12.4.1, gcc)
Unable to find an artifact with the name: fbgemm_gpu_nightly_genai_x86_gcc_py3.13_cu12.6.3.whl
test_and_publish_artifact (x86, linux.g5.4xlarge.nvidia.gpu, 3.11, 12.6.3, 12.4.1, clang)
Unable to find an artifact with the name: fbgemm_gpu_nightly_genai_x86_clang_py3.11_cu12.6.3.whl
test_and_publish_artifact (x86, linux.g5.4xlarge.nvidia.gpu, 3.10, 12.6.3, 12.4.1, clang)
Unable to find an artifact with the name: fbgemm_gpu_nightly_genai_x86_clang_py3.10_cu12.6.3.whl
test_and_publish_artifact (x86, linux.g5.4xlarge.nvidia.gpu, 3.11, 12.6.3, 12.4.1, gcc)
Unable to find an artifact with the name: fbgemm_gpu_nightly_genai_x86_gcc_py3.11_cu12.6.3.whl
test_and_publish_artifact (x86, linux.g5.4xlarge.nvidia.gpu, 3.13, 12.6.3, 12.4.1, clang)
Unable to find an artifact with the name: fbgemm_gpu_nightly_genai_x86_clang_py3.13_cu12.6.3.whl
Deprecation notice: v1, v2, and v3 of the artifact actions
The following artifacts were uploaded using a version of actions/upload-artifact that is scheduled for deprecation: "fbgemm_gpu_nightly_genai_x86_clang_py3.10_cu11.8.0.whl", "fbgemm_gpu_nightly_genai_x86_clang_py3.10_cu12.4.1.whl", "fbgemm_gpu_nightly_genai_x86_clang_py3.11_cu11.8.0.whl", "fbgemm_gpu_nightly_genai_x86_clang_py3.11_cu12.4.1.whl", "fbgemm_gpu_nightly_genai_x86_clang_py3.12_cu11.8.0.whl", "fbgemm_gpu_nightly_genai_x86_clang_py3.12_cu12.4.1.whl", "fbgemm_gpu_nightly_genai_x86_clang_py3.13_cu11.8.0.whl", "fbgemm_gpu_nightly_genai_x86_clang_py3.13_cu12.4.1.whl", "fbgemm_gpu_nightly_genai_x86_clang_py3.9_cu11.8.0.whl", "fbgemm_gpu_nightly_genai_x86_clang_py3.9_cu12.4.1.whl", "fbgemm_gpu_nightly_genai_x86_gcc_py3.10_cu11.8.0.whl", "fbgemm_gpu_nightly_genai_x86_gcc_py3.10_cu12.4.1.whl", "fbgemm_gpu_nightly_genai_x86_gcc_py3.11_cu11.8.0.whl", "fbgemm_gpu_nightly_genai_x86_gcc_py3.11_cu12.4.1.whl", "fbgemm_gpu_nightly_genai_x86_gcc_py3.12_cu11.8.0.whl", "fbgemm_gpu_nightly_genai_x86_gcc_py3.12_cu12.4.1.whl", "fbgemm_gpu_nightly_genai_x86_gcc_py3.13_cu11.8.0.whl", "fbgemm_gpu_nightly_genai_x86_gcc_py3.13_cu12.4.1.whl", "fbgemm_gpu_nightly_genai_x86_gcc_py3.9_cu11.8.0.whl", "fbgemm_gpu_nightly_genai_x86_gcc_py3.9_cu12.4.1.whl". Please update your workflow to use v4 of the artifact actions. Learn more: https://github.blog/changelog/2024-04-16-deprecation-notice-v3-of-the-artifact-actions/

Artifacts

Produced during runtime
Name Size
fbgemm_gpu_nightly_genai_x86_clang_py3.10_cu11.8.0.whl
2.62 MB
fbgemm_gpu_nightly_genai_x86_clang_py3.10_cu12.4.1.whl
5.66 MB
fbgemm_gpu_nightly_genai_x86_clang_py3.11_cu11.8.0.whl
2.62 MB
fbgemm_gpu_nightly_genai_x86_clang_py3.11_cu12.4.1.whl
5.66 MB
fbgemm_gpu_nightly_genai_x86_clang_py3.12_cu11.8.0.whl
2.62 MB
fbgemm_gpu_nightly_genai_x86_clang_py3.12_cu12.4.1.whl
5.66 MB
fbgemm_gpu_nightly_genai_x86_clang_py3.13_cu11.8.0.whl
2.62 MB
fbgemm_gpu_nightly_genai_x86_clang_py3.13_cu12.4.1.whl
5.66 MB
fbgemm_gpu_nightly_genai_x86_clang_py3.9_cu11.8.0.whl
2.62 MB
fbgemm_gpu_nightly_genai_x86_clang_py3.9_cu12.4.1.whl
5.66 MB
fbgemm_gpu_nightly_genai_x86_gcc_py3.10_cu11.8.0.whl
2.5 MB
fbgemm_gpu_nightly_genai_x86_gcc_py3.10_cu12.4.1.whl
5.53 MB
fbgemm_gpu_nightly_genai_x86_gcc_py3.11_cu11.8.0.whl
2.5 MB
fbgemm_gpu_nightly_genai_x86_gcc_py3.11_cu12.4.1.whl
5.53 MB
fbgemm_gpu_nightly_genai_x86_gcc_py3.12_cu11.8.0.whl
2.5 MB
fbgemm_gpu_nightly_genai_x86_gcc_py3.12_cu12.4.1.whl
5.53 MB
fbgemm_gpu_nightly_genai_x86_gcc_py3.13_cu11.8.0.whl
2.5 MB
fbgemm_gpu_nightly_genai_x86_gcc_py3.13_cu12.4.1.whl
5.53 MB
fbgemm_gpu_nightly_genai_x86_gcc_py3.9_cu11.8.0.whl
2.5 MB
fbgemm_gpu_nightly_genai_x86_gcc_py3.9_cu12.4.1.whl
5.53 MB