Migrate GenAI gqa attn splitk kernels to `FBGEMM_LAUNCH_KERNEL`, pt 1 · pytorch/FBGEMM@58118b3

Triggered via pull request September 18, 2025 20:24

q10

opened #4894

q10:export-D81817544

Status Success

Total duration 21m 30s

Artifacts 4

fbgemm_gpu_benchmark_cpu.yml

on: pull_request

Matrix: build_artifact

Matrix: benchmark_artifact

Produced during runtime

Name	Size	Digest
fbgemm_gpu_nightly_cpu_arm_gcc_py3.13.whl	4.2 MB	`sha256:464644c3e1c4e64d6b013e4a075a60a40254b7b8d6a3737dfa54808e634bcd96`
fbgemm_gpu_nightly_cpu_x86_gcc_py3.13.whl	5.35 MB	`sha256:dd8396a1807196a1d3651c2f5537c3ca66096fa513a24eb286d7d8d3f3bb2536`
fbgemm_gpu_traces_arm_gcc_py3.13_cpu.zip	238 KB	`sha256:da67547172c95432283c797268171464a2e4344ca13dbe14f841db8f050a6912`
fbgemm_gpu_traces_x86_gcc_py3.13_cpu.zip	238 KB	`sha256:b887057c655781a6df89e617870bd8b4c450a43cdda36d24b1b20bd2857bd801`