Skip to content

Actions: pytorch/FBGEMM

Actions

Build FBGEMM GenAI Aarch64 Linux Wheels

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
build_wheels_genai_linux_aarch64.yml
5,581 workflow runs
5,581 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

Migrate GenAI kv cache kernel arguments to use PTA_B Build FBGEMM GenAI Aarch64 Linux Wheels #6005: Pull request #4899 opened by q10
1h 59m 33s q10:export-D82775043
q10:export-D82775043
1h 59m 33s
Migrate GenAI kv cache kernels to FBGEMM_LAUNCH_KERNEL, pt 4 (#4895) Build FBGEMM GenAI Aarch64 Linux Wheels #6004: Commit 944647a pushed by facebook-github-bot
22m 1s main
main
22m 1s
Restrict to fp8, and patch 4.2.0 release changes Build FBGEMM GenAI Aarch64 Linux Wheels #6003: Pull request #4898 opened by henrylhtsang
2h 4m 37s henrylhtsang:export-D82792657
henrylhtsang:export-D82792657
2h 4m 37s
Add softmax_scale to blackwell cutlass fmha Build FBGEMM GenAI Aarch64 Linux Wheels #6002: Pull request #4897 opened by henrylhtsang
2h 6m 35s henrylhtsang:export-D82788784
henrylhtsang:export-D82788784
2h 6m 35s
Small changes to improve blackwell_fmha_test.py Build FBGEMM GenAI Aarch64 Linux Wheels #6001: Pull request #4896 synchronize by henrylhtsang
2h 0m 31s henrylhtsang:export-D81992869
henrylhtsang:export-D81992869
2h 0m 31s
Small changes to improve blackwell_fmha_test.py Build FBGEMM GenAI Aarch64 Linux Wheels #6000: Pull request #4896 synchronize by henrylhtsang
19s henrylhtsang:export-D81992869
henrylhtsang:export-D81992869
19s
Small changes to improve blackwell_fmha_test.py Build FBGEMM GenAI Aarch64 Linux Wheels #5999: Pull request #4896 opened by henrylhtsang
17m 56s henrylhtsang:export-D81992869
henrylhtsang:export-D81992869
17m 56s
Migrate GenAI kv cache kernels to FBGEMM_LAUNCH_KERNEL, pt 4 Build FBGEMM GenAI Aarch64 Linux Wheels #5998: Pull request #4895 opened by q10
1h 59m 2s q10:export-D82773258
q10:export-D82773258
1h 59m 2s
Migrate GenAI gqa attn splitk kernels to FBGEMM_LAUNCH_KERNEL, pt 1 Build FBGEMM GenAI Aarch64 Linux Wheels #5997: Pull request #4894 opened by q10
2h 6m 10s q10:export-D81817544
q10:export-D81817544
2h 6m 10s
Back out "Enable CUTLASS grouped GEMM for pretraining wgrad on GB200 … Build FBGEMM GenAI Aarch64 Linux Wheels #5996: Commit 53f9e51 pushed by facebook-github-bot
2h 22m 47s main
main
2h 22m 47s
upgrade cutlass to 4.2.0 for fbcode Build FBGEMM GenAI Aarch64 Linux Wheels #5995: Pull request #4893 opened by henrylhtsang
2h 20m 49s henrylhtsang:export-D82567751
henrylhtsang:export-D82567751
2h 20m 49s
Back out "Enable CUTLASS grouped GEMM for pretraining wgrad on GB200 and H100" Build FBGEMM GenAI Aarch64 Linux Wheels #5994: Pull request #4892 synchronize by jiawenliu64
2h 2m 10s jiawenliu64:export-D82746070
jiawenliu64:export-D82746070
2h 2m 10s
Back out "Enable CUTLASS grouped GEMM for pretraining wgrad on GB200 and H100" Build FBGEMM GenAI Aarch64 Linux Wheels #5993: Pull request #4892 opened by jiawenliu64
11m 27s jiawenliu64:export-D82746070
jiawenliu64:export-D82746070
11m 27s
add blog link Build FBGEMM GenAI Aarch64 Linux Wheels #5992: Pull request #4888 synchronize by sijiac
2h 15m 43s sijiac:export-D82681043
sijiac:export-D82681043
2h 15m 43s
Add ability to pad the rowwise quantized tensors Build FBGEMM GenAI Aarch64 Linux Wheels #5991: Pull request #4877 synchronize by yyetim
2h 14m 16s yyetim:export-D82486197
yyetim:export-D82486197
2h 14m 16s
dequantize_fp8_cache_kernel: Move D=128 device-side-assertion check to host Build FBGEMM GenAI Aarch64 Linux Wheels #5990: Pull request #4869 synchronize by ColinPeppler
2h 12m 42s ColinPeppler:export-D82320518
ColinPeppler:export-D82320518
2h 12m 42s
dequantize_fp8_cache_kernel: Move D=128 device-side-assertion check to host Build FBGEMM GenAI Aarch64 Linux Wheels #5989: Pull request #4869 synchronize by ColinPeppler
39s ColinPeppler:export-D82320518
ColinPeppler:export-D82320518
39s
symmetric quantization to FBGEMM prefill token-wise FP8 (fixed) Build FBGEMM GenAI Aarch64 Linux Wheels #5988: Pull request #4868 synchronize by ColinPeppler
2h 17m 15s ColinPeppler:export-D82320500
ColinPeppler:export-D82320500
2h 17m 15s
fix the scaled input issue Build FBGEMM GenAI Aarch64 Linux Wheels #5987: Pull request #4884 synchronize by pyjhzwh
2h 12m 49s pyjhzwh:export-D82147819
pyjhzwh:export-D82147819
2h 12m 49s
Add ability to pad the rowwise quantized tensors Build FBGEMM GenAI Aarch64 Linux Wheels #5986: Pull request #4877 synchronize by yyetim
20m 8s yyetim:export-D82486197
yyetim:export-D82486197
20m 8s
Add ability to pad the rowwise quantized tensors Build FBGEMM GenAI Aarch64 Linux Wheels #5985: Pull request #4877 synchronize by yyetim
7m 9s yyetim:export-D82486197
yyetim:export-D82486197
7m 9s
Backward optimizations + rocm and gfx950 build support Build FBGEMM GenAI Aarch64 Linux Wheels #5984: Pull request #4729 synchronize by liligwu
Action required ROCm:shbiswas/uptream_develop_0815
ROCm:shbiswas/uptream_develop_0815
Action required
2025-09-18 nightly release (fd32631d837b41251311099c393af7d7be5cfbf5) Build FBGEMM GenAI Aarch64 Linux Wheels #5983: Commit 5206781 pushed by pytorchbot
2h 19m 49s nightly
nightly
2h 19m 49s
Optimize wgrad CUTLASS grouped gemm Build FBGEMM GenAI Aarch64 Linux Wheels #5982: Pull request #4891 opened by jiawenliu64
2h 17m 21s export-D82700455
export-D82700455
2h 17m 21s
Support multiple total-k and total-m in quantize bench Build FBGEMM GenAI Aarch64 Linux Wheels #5981: Pull request #4890 opened by jiawenliu64
2h 17m 19s jiawenliu64:export-D82700396
jiawenliu64:export-D82700396
2h 17m 19s