Skip to content

Actions: pytorch/FBGEMM

Actions

Build FBGEMM_GPU x86 Linux Wheels

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
build_wheels_linux_x86.yml
2,500+ workflow runs
2,500+ workflow runs

Filter by Event

⌥ + click/return to exclude

Filter by Status

⌥ + click/return to exclude

Filter by Branch

⌥ + click/return to exclude

Filter by Actor

⌥ + click/return to exclude
Add FP8 rowwise padding to quantized AllToAll pooled embeddings (#5673) Build FBGEMM_GPU x86 Linux Wheels #20214: Pull request #5673 synchronize by RohanVardhan
In progress RohanVardhan:export-D100069631
RohanVardhan:export-D100069631
In progress
support warpSize 32 and 64 in the same build (#5739) Build FBGEMM_GPU x86 Linux Wheels #20213: Commit d8f6d05 pushed by meta-codesync Bot
In progress main
main
In progress
Make float16/bfloat16 distinct types Build FBGEMM_GPU x86 Linux Wheels #20212: Pull request #5736 synchronize by cyyever
In progress cyyever:bf16-distinct-types
cyyever:bf16-distinct-types
In progress
Make float16/bfloat16 distinct types Build FBGEMM_GPU x86 Linux Wheels #20211: Pull request #5736 synchronize by cyyever
3h 34m 58s cyyever:bf16-distinct-types
cyyever:bf16-distinct-types
3h 34m 58s
TLX IKBO FA benchmarking with latest commit hash + bug fix (#5734) Build FBGEMM_GPU x86 Linux Wheels #20210: Commit 7c0a1a7 pushed by meta-codesync Bot
5h 6m 16s main
main
5h 6m 16s
support warpSize 32 and 64 in the same build (#5739) Build FBGEMM_GPU x86 Linux Wheels #20209: Pull request #5739 synchronize by q10
5h 7m 51s q10:export-D102479962
q10:export-D102479962
5h 7m 51s
support warpSize 32 and 64 in the same build (#5739) Build FBGEMM_GPU x86 Linux Wheels #20208: Pull request #5739 opened by q10
1h 28m 16s q10:export-D102479962
q10:export-D102479962
1h 28m 16s
Optimize jagged_unique_indices_cuda (binary-search length + custom cub pipeline) (#5718) Build FBGEMM_GPU x86 Linux Wheels #20207: Pull request #5718 synchronize by AlbertDachiChen
5h 3m 19s AlbertDachiChen:export-D103207899
AlbertDachiChen:export-D103207899
5h 3m 19s
TLX IKBO FA benchmarking with latest commit hash + bug fix (#5734) Build FBGEMM_GPU x86 Linux Wheels #20206: Pull request #5734 synchronize by liptds
3h 38m 18s liptds:export-D103791854
liptds:export-D103791854
3h 38m 18s
fbcode/deeplearning/fbgemm/fbgemm_gpu/test/tbe/cache [B] Build FBGEMM_GPU x86 Linux Wheels #20205: Pull request #5738 opened by meta-codesync Bot
5h 3m 30s export-D103335905
export-D103335905
5h 3m 30s
fbcode/deeplearning/fbgemm/fbgemm_gpu/test/tbe/utils/generate_vbe_metadata_test.py Build FBGEMM_GPU x86 Linux Wheels #20204: Pull request #5737 opened by meta-codesync Bot
5h 5m 13s export-D103819080
export-D103819080
5h 5m 13s
Add KVZCH inference read-time hit rate metrics via fb303 ODS counters… Build FBGEMM_GPU x86 Linux Wheels #20203: Commit 15b2a43 pushed by meta-codesync Bot
5h 6m 39s main
main
5h 6m 39s
Add KVZCH inference read-time hit rate metrics via fb303 ODS counters (#5730) Build FBGEMM_GPU x86 Linux Wheels #20202: Pull request #5730 synchronize by hy-NJU
5h 5m 31s hy-NJU:export-D101879296
hy-NJU:export-D101879296
5h 5m 31s
remove uneccesarry field for FixedBlockPool in inference (#5729) Build FBGEMM_GPU x86 Linux Wheels #20201: Pull request #5729 synchronize by hy-NJU
5h 2m 40s hy-NJU:export-D91654624
hy-NJU:export-D91654624
5h 2m 40s
Make float16/bfloat16 distinct types Build FBGEMM_GPU x86 Linux Wheels #20200: Pull request #5736 synchronize by cyyever
3h 37m 20s cyyever:bf16-distinct-types
cyyever:bf16-distinct-types
3h 37m 20s
2026-05-05 nightly release (0872fda2f2a43e7cf75a5e3e126c568820089c0d) Build FBGEMM_GPU x86 Linux Wheels #20199: Commit dfa9de0 pushed by pytorchbot
5h 7m 7s nightly
nightly
5h 7m 7s
Make float16/bfloat16 distinct types Build FBGEMM_GPU x86 Linux Wheels #20198: Pull request #5736 synchronize by cyyever
Action required cyyever:bf16-distinct-types
cyyever:bf16-distinct-types
Action required
Make float16/bfloat16 distinct types Build FBGEMM_GPU x86 Linux Wheels #20197: Pull request #5736 opened by cyyever
Action required cyyever:bf16-distinct-types
cyyever:bf16-distinct-types
Action required
Switch bf16 quantize ops to at::kBFloat16 Build FBGEMM_GPU x86 Linux Wheels #20196: Pull request #5735 opened by cyyever
5h 4m 54s cyyever:bf16-quantize-kbfloat16
cyyever:bf16-quantize-kbfloat16
5h 4m 54s
Add bf16 avx2 8bit dequant and Nbit dequant Build FBGEMM_GPU x86 Linux Wheels #20195: Pull request #5634 synchronize by cyyever
5h 2m 17s cyyever:impl-bf16-avx2-8bit-dequant
cyyever:impl-bf16-avx2-8bit-dequant
5h 2m 17s
Fix DramKV race: hold rlock during inplace update writes Build FBGEMM_GPU x86 Linux Wheels #20194: Pull request #5536 synchronize by cyyever
Action required cyyever:fix3
cyyever:fix3
Action required
Fix bf16 rounding to IEEE 754 ties-to-even Build FBGEMM_GPU x86 Linux Wheels #20193: Pull request #5648 synchronize by cyyever
Action required cyyever:fix/bf16-ties-to-even
cyyever:fix/bf16-ties-to-even
Action required
Enable fp32 tests for Windows Build FBGEMM_GPU x86 Linux Wheels #20192: Pull request #4722 synchronize by cyyever
Action required cyyever:enable_fp32_win
cyyever:enable_fp32_win
Action required
Bump ASMJIT to 1.18 Build FBGEMM_GPU x86 Linux Wheels #20191: Pull request #4850 synchronize by cyyever
Action required cyyever:asmjit_1.18
cyyever:asmjit_1.18
Action required
Enable device-side assertions on ROCm Build FBGEMM_GPU x86 Linux Wheels #20190: Pull request #5723 synchronize by cyyever
5h 3m 20s cyyever:enable-rocm-dsa
cyyever:enable-rocm-dsa
5h 3m 20s