Skip to content

Actions: ROCm/iris

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
7,110 workflow runs
7,110 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

Iris Nightly Triton Test Iris Nightly Triton Test #10: Scheduled
5h 59m 22s main
main
5h 59m 22s
Merge branch 'dev/dasidler/sdma-benchmark' of https://github.com/ROCm… Lint & Auto-Fix #3594: Commit 39ac9d7 pushed by dsidler
11s dev/dasidler/sdma-benchmark
dev/dasidler/sdma-benchmark
11s
Merge branch 'dev/dasidler/sdma-benchmark' of https://github.com/ROCm… Lint & Auto-Fix #3593: Commit 9d4c7f9 pushed by dsidler
36s dev/dasidler/sdma-benchmark
dev/dasidler/sdma-benchmark
36s
Iris Nightly Triton Test Iris Nightly Triton Test #9: Scheduled
5h 58m 45s main
main
5h 58m 45s
Iris Nightly Triton Test Iris Nightly Triton Test #8: Scheduled
12h 39m 31s main
main
12h 39m 31s
Remove pre-kernel barrier, match all_gather convention Lint & Auto-Fix #3592: Commit 5f0002f pushed by mawad-amd
20s muhaawad/broadcast-reduce
muhaawad/broadcast-reduce
20s
Use device_barrier instead of host barrier for lower overhead Lint & Auto-Fix #3591: Commit d39a630 pushed by mawad-amd
26s muhaawad/broadcast-reduce
muhaawad/broadcast-reduce
26s
Add broadcast and reduce collectives to iris CCL Lint & Auto-Fix #3590: Commit 54b4496 pushed by mawad-amd
44s muhaawad/broadcast-reduce
muhaawad/broadcast-reduce
44s
Revert to separated wait/data/signal kernels to fix large-size regres… Lint & Auto-Fix #3589: Commit 3376f00 pushed by mawad-amd
21s muhaawad/p2p-send-recv
muhaawad/p2p-send-recv
21s
Fuse wait+store and wait+copy kernels, tune num_warps=8 Lint & Auto-Fix #3588: Commit eae1cb1 pushed by mawad-amd
17s muhaawad/p2p-send-recv
muhaawad/p2p-send-recv
17s
Add send-side flow control to prevent buffer overwrite Lint & Auto-Fix #3587: Commit 9b63732 pushed by mawad-amd
34s muhaawad/p2p-send-recv
muhaawad/p2p-send-recv
34s
Optimize fused AG+GEMM: tracing, alignment hints, parameter sweep Lint & Auto-Fix #3586: Commit b5ce35f pushed by mawad-amd
46s muhaawad/fused-allgather-gemm
muhaawad/fused-allgather-gemm
46s
Scale test tolerances with problem dimensions Lint & Auto-Fix #3585: Commit c0604cb pushed by mawad-amd
39s muhaawad/fused-gemm-reducescatter
muhaawad/fused-gemm-reducescatter
39s
Fix test reference: use all_reduce+column slice instead of reduce_sca… Lint & Auto-Fix #3584: Commit 437c0aa pushed by mawad-amd
23s muhaawad/fused-gemm-reducescatter
muhaawad/fused-gemm-reducescatter
23s
Debug: disable swizzle, add deterministic test Lint & Auto-Fix #3583: Commit bdf5df6 pushed by mawad-amd
25s muhaawad/fused-gemm-reducescatter
muhaawad/fused-gemm-reducescatter
25s
Fix data race in sequential P2P sends with double-buffering Lint & Auto-Fix #3582: Commit 8b5fdb5 pushed by mawad-amd
46s muhaawad/p2p-send-recv
muhaawad/p2p-send-recv
46s
Relax fp32 tolerance for GEMM+RS tests Lint & Auto-Fix #3581: Commit 51ce2aa pushed by mawad-amd
32s muhaawad/fused-gemm-reducescatter
muhaawad/fused-gemm-reducescatter
32s
Fix BLOCK_N > shard_size causing wrong scatter destinations Lint & Auto-Fix #3580: Commit 7ad013d pushed by mawad-amd
28s muhaawad/fused-gemm-reducescatter
muhaawad/fused-gemm-reducescatter
28s
Fix ruff lint issues in benchmark and p2p module Lint & Auto-Fix #3579: Commit 90caddd pushed by mawad-amd
38s muhaawad/p2p-send-recv
muhaawad/p2p-send-recv
38s
Widen fp32 atol for large-K TF32 GEMM accumulation error Lint & Auto-Fix #3578: Commit 354d90e pushed by mawad-amd
26s muhaawad/fused-allgather-gemm
muhaawad/fused-allgather-gemm
26s
Relax fp32 tolerance for TF32 GEMM accumulation Lint & Auto-Fix #3577: Commit e770fb8 pushed by mawad-amd
32s muhaawad/fused-allgather-gemm
muhaawad/fused-allgather-gemm
32s
Skip post-barrier for all_pairs_chunked variant Lint & Auto-Fix #3576: Commit b3127d9 pushed by mawad-amd
44s muhaawad/allreduce-allpairs-chunked
muhaawad/allreduce-allpairs-chunked
44s
Add fused all-gather + GEMM (prologue fusion) for column-parallel TP Lint & Auto-Fix #3575: Commit f954f3d pushed by mawad-amd
16s muhaawad/fused-allgather-gemm
muhaawad/fused-allgather-gemm
16s
Add fused GEMM + reduce-scatter collective Lint & Auto-Fix #3574: Commit 3828536 pushed by mawad-amd
30s muhaawad/fused-gemm-reducescatter
muhaawad/fused-gemm-reducescatter
30s
Add all-pairs chunked all-reduce variant Lint & Auto-Fix #3573: Commit aa8f54f pushed by mawad-amd
18s muhaawad/allreduce-allpairs-chunked
muhaawad/allreduce-allpairs-chunked
18s