Make FloatToFloat16 conversion 75x faster using SVE2 instructions #8853
Triggered via pull request
January 29, 2025 02:26
embg
Status
Success
Total duration
1h 53m 22s
Artifacts
2
build_wheels_linux_aarch64.yml
on: pull_request
generate-matrix
/
generate
3s
Matrix: build
Artifacts
Produced during runtime
| Name | Size | |
|---|---|---|
|
pytorch_FBGEMM__3.9_cpu_aarch64
|
2.69 MB |
|
|
pytorch_FBGEMM__3.9_cu126_aarch64
|
446 MB |
|