Skip to content

Add optional zero_start_index_M argument to triton fp8 rowwise quantization #8858

Add optional zero_start_index_M argument to triton fp8 rowwise quantization

Add optional zero_start_index_M argument to triton fp8 rowwise quantization #8858

Triggered via pull request January 28, 2025 23:01
Status Success
Total duration 2h 54m 44s
Artifacts 6

build_wheels_linux_x86.yml

on: pull_request
generate-matrix  /  generate
4s
generate-matrix / generate
Matrix: pytorch/FBGEMM / build
Matrix: build / upload / upload
Fit to window
Zoom out
Zoom in

Artifacts

Produced during runtime
Name Size
pytorch_FBGEMM__3.9_cpu_x86_64
3.75 MB
pytorch_FBGEMM__3.9_cu118_x86_64
342 MB
pytorch_FBGEMM__3.9_cu124_x86_64
452 MB
pytorch_FBGEMM__3.9_cu126_x86_64
449 MB
pytorch_FBGEMM__3.9_rocm6.2.4_x86_64
127 MB
pytorch_FBGEMM__3.9_rocm6.3_x86_64
122 MB