Skip to content

Actions: ROCm/iris

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
2,500+ workflow runs
2,500+ workflow runs

Filter by Workflow

⌥ + click/return to exclude

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
Iris Nightly Triton Test Iris Nightly Triton Test #51: Scheduled
Queued main
main
Queued
Nightly Docker Build Nightly Docker Build #38: Scheduled
In progress main
main
In progress
Add graph-capturable one-shot all-reduce Lint & Auto-Fix #4122: Pull request #531 synchronize by mawad-amd
23s muhaawad/one-shot-vllm
muhaawad/one-shot-vllm
23s
Add graph-capturable one-shot all-reduce Iris External Validation Test #1217: Pull request #531 synchronize by mawad-amd
2m 17s muhaawad/one-shot-vllm
muhaawad/one-shot-vllm
2m 17s
Add graph-capturable one-shot all-reduce Iris Performance Regression Test #1097: Pull request #531 synchronize by mawad-amd
1h 32m 37s muhaawad/one-shot-vllm
muhaawad/one-shot-vllm
1h 32m 37s
Add graph-capturable one-shot all-reduce Iris Tests #819: Pull request #531 synchronize by mawad-amd
In progress muhaawad/one-shot-vllm
muhaawad/one-shot-vllm
In progress
Add graph-capturable one-shot all-reduce Build and Deploy Documentation #970: Pull request #531 synchronize by mawad-amd
1m 28s muhaawad/one-shot-vllm
muhaawad/one-shot-vllm
1m 28s
Default all-reduce to gluon one-shot Lint & Auto-Fix #4121: Commit ee3ec96 pushed by mawad-amd
27s muhaawad/one-shot-vllm
muhaawad/one-shot-vllm
27s
Add graph-capturable one-shot all-reduce Lint & Auto-Fix #4120: Pull request #531 synchronize by mawad-amd
36s muhaawad/one-shot-vllm
muhaawad/one-shot-vllm
36s
Add graph-capturable one-shot all-reduce Iris External Validation Test #1216: Pull request #531 synchronize by mawad-amd
2m 11s muhaawad/one-shot-vllm
muhaawad/one-shot-vllm
2m 11s
Add graph-capturable one-shot all-reduce Iris Performance Regression Test #1096: Pull request #531 synchronize by mawad-amd
6m 31s muhaawad/one-shot-vllm
muhaawad/one-shot-vllm
6m 31s
Add graph-capturable one-shot all-reduce Build and Deploy Documentation #969: Pull request #531 synchronize by mawad-amd
1m 39s muhaawad/one-shot-vllm
muhaawad/one-shot-vllm
1m 39s
Add graph-capturable one-shot all-reduce Iris Tests #818: Pull request #531 synchronize by mawad-amd
6m 30s muhaawad/one-shot-vllm
muhaawad/one-shot-vllm
6m 30s
Add gluon one-shot all-reduce with graph-capture safety Lint & Auto-Fix #4119: Commit 3382fe8 pushed by mawad-amd
38s muhaawad/one-shot-vllm
muhaawad/one-shot-vllm
38s
Always write to scratch first, barrier, then copy for in-place Lint & Auto-Fix #4118: Commit 129c02a pushed by mawad-amd
26s muhaawad/gluon-one-shot-ar
muhaawad/gluon-one-shot-ar
26s
Handle in-place all-reduce inside the gluon kernel Lint & Auto-Fix #4117: Commit d75f9c9 pushed by mawad-amd
29s muhaawad/gluon-one-shot-ar
muhaawad/gluon-one-shot-ar
29s
Revert to direct write for graph-captured in-place all-reduce Lint & Auto-Fix #4116: Commit 14b3dfa pushed by mawad-amd
18s muhaawad/gluon-one-shot-ar
muhaawad/gluon-one-shot-ar
18s
Revert scratch to torch.empty — heap allocation causes worse failures Lint & Auto-Fix #4115: Commit f037a45 pushed by mawad-amd
19s muhaawad/gluon-one-shot-ar
muhaawad/gluon-one-shot-ar
19s
Allocate in-place scratch from symmetric heap Lint & Auto-Fix #4114: Commit 337f2b6 pushed by mawad-amd
30s muhaawad/gluon-one-shot-ar
muhaawad/gluon-one-shot-ar
30s
Fix in-place all-reduce for graph capture with persistent scratch buffer Lint & Auto-Fix #4113: Commit 5c46d5e pushed by mawad-amd
27s muhaawad/gluon-one-shot-ar
muhaawad/gluon-one-shot-ar
27s
Make all-reduce graph-capture-safe for vLLM integration Lint & Auto-Fix #4112: Commit 0829d9a pushed by mawad-amd
29s muhaawad/gluon-one-shot-ar
muhaawad/gluon-one-shot-ar
29s
Add barrier before in-place copy-back to prevent cross-rank race Lint & Auto-Fix #4111: Commit 85292a7 pushed by mawad-amd
35s muhaawad/gluon-one-shot-ar
muhaawad/gluon-one-shot-ar
35s
Simplify in-place handling: always allocate fresh scratch buffer Lint & Auto-Fix #4110: Commit bd18f7a pushed by mawad-amd
34s muhaawad/gluon-one-shot-ar
muhaawad/gluon-one-shot-ar
34s
Add in-place all-reduce test for default gluon config Lint & Auto-Fix #4109: Commit f88e628 pushed by mawad-amd
16s muhaawad/gluon-one-shot-ar
muhaawad/gluon-one-shot-ar
16s
Default to gluon one-shot all-reduce, handle in-place transparently Lint & Auto-Fix #4108: Commit f29dd9b pushed by mawad-amd
37s muhaawad/gluon-one-shot-ar
muhaawad/gluon-one-shot-ar
37s