Skip to content

Actions: lightseekorg/triton

Actions

Integration Tests

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
ci.yml
This workflow was disabled manually.
19 workflow runs
19 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
[GSAN] Add CUDA feature flag for PTX 7.0 support in GSan build (#10282) Integration Tests #19: Commit fd68aeb pushed by lightseek-bot
1m 13s main
main
1m 13s
Account for duplicated elementwise ops in backward remat cost (#10129) Integration Tests #18: Commit dc57615 pushed by antiagainst
1m 9s main
main
1m 9s
[AMD] Extend chain-dot detection across loop iteration boundaries (#1… Integration Tests #17: Commit 521c2e3 pushed by lightseek-bot
1m 6s main
main
1m 6s
[NFC] Clean up top-level CMakeLists.txt (#10268) Integration Tests #16: Commit 70b3e39 pushed by lightseek-bot
1m 5s main
main
1m 5s
[AMD][GFX1250] Reorder bf16 attn prologue TDM to match loop body (#10… Integration Tests #15: Commit 4cd6bcb pushed by lightseek-bot
1m 10s main
main
1m 10s
[AMD] Fixing mi350 BlockPingpong update waits (#10194) Integration Tests #14: Commit b45acd4 pushed by lightseek-bot
1m 15s main
main
1m 15s
[Frontend] Unwrap constexpr before semantic (#10227) Integration Tests #13: Commit 321275f pushed by antiagainst
1m 32s main
main
1m 32s
[Gluon] Unwrap constexpr bool in dot_scaled and associative_scan (#10… Integration Tests #12: Commit 2bcd105 pushed by lightseek-bot
1m 3s main
main
1m 3s
[AMD][gfx1250] Drop ttg.async_commit_group from TDM async chain (#10215) Integration Tests #11: Commit 12fe342 pushed by antiagainst
1m 12s main
main
1m 12s
[AMD][gfx1250] Explicitly propagate NaN for MXFP attn example (#10208) Integration Tests #10: Commit 886365c pushed by lightseek-bot
1m 2s main
main
1m 2s
[mxfp4] Fix Hopper scale padding mask (#10190) Integration Tests #9: Commit 885569f pushed by lightseek-bot
53s main
main
53s
[Proton] Normalize roofline metric dtypes (#10199) Integration Tests #8: Commit b738eb6 pushed by lightseek-bot
1m 3s main
main
1m 3s
[AMD] Enable IN_THREAD_TRANSPOSE to GFX120* by default (#10185) Integration Tests #7: Commit 5d69e1c pushed by lightseek-bot
1m 4s main
main
1m 4s
[AMD][gfx1250] Add v_pk_*_bf16 conversion and fix sqrt denorm (#10193) Integration Tests #6: Commit e42e0a0 pushed by lightseek-bot
1m 12s main
main
1m 12s
[AMD][gfx9] Restore token-aware wait count derivation on asyncmark ta… Integration Tests #5: Commit 18415f7 pushed by lightseek-bot
1m 13s main
main
1m 13s
[AMD][CI] Switch gfx942 to use normal pytorch docker (#10146) Integration Tests #4: Commit 72eb2f4 pushed by lightseek-bot
1m 8s main
main
1m 8s
[Gluon][Examples] Update MoE BMM1 selector configs (#10124) Integration Tests #3: Commit 9d9a2d7 pushed by lightseek-bot
1m 10s main
main
1m 10s
[FPSAN] Add support for batched matmul, make fpsan errors fatal (#10118) Integration Tests #2: Commit 4da2e26 pushed by lightseek-bot
1m 58s main
main
1m 58s
[GSan] Model TMAReduceOp as atomic (#10053) Integration Tests #1: Commit 07ef288 pushed by lightseek-bot
2m 9s main
main
2m 9s