Skip to content

Pull requests: AMD-AGI/TraceLens

New pull request New
30 Open 393 Closed
30 Open 393 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Add deterministic marker identification evals
#656 opened May 21, 2026 by Ahmedhasssan-aig Contributor Loading…
3 tasks done
fix(jaxopkeys): route TE-adjacent buffer-init kernels to TE
#655 opened May 20, 2026 by gphuang Contributor Loading…
3 tasks done
2
@gphuang
9
Changes to TraceLens files for comparative analysis
#654 opened May 19, 2026 by unagiboi Collaborator Loading…
feat(perfmodel): add varlen aiter FA + aten::_flash_attention_forward perf models (Wan 2.2 coverage)
#651 opened May 19, 2026 by gphuang Contributor Loading…
4 tasks done
1
3
Propagate trace metadata through TraceLens analyzers
#643 opened May 14, 2026 by ajassani Collaborator • Draft
fix: ops_summary shows GPU kernel names for orphan HIP launches (#641)
#642 opened May 14, 2026 by gabeweisz Collaborator Loading…
1
feat(perfmodel): add primus_turbo and aiter MXFP4 GEMM and quantize perf models
#638 opened May 11, 2026 by gphuang Contributor Loading…
6 tasks done
1
@gphuang
9
Include atleast one prefill step in the split if available (Bug fix)
#636 opened May 8, 2026 by devalshahamd Contributor Loading…
2
PerfModel: simplify torch op categorization registry
#630 opened May 6, 2026 by ajassani Collaborator Loading…
3
Support fp8 quantization for KV cache in Inference perf models
#625 opened May 6, 2026 by devalshahamd Contributor Loading…
Feat/add gpu op uids to report enhancement New feature or request
#624 opened May 5, 2026 by unagiboi Collaborator Loading…
Add op category breakdown to perf report comparison (#331)
#622 opened May 5, 2026 by ajassani Collaborator Loading…
2 of 4 tasks
1
3
Perfmodel for torch.compile (triton) enhancement New feature or request
#621 opened May 5, 2026 by janmatai Contributor Loading…
2
Add IdleTimeAnalyser: classify GPU idle gaps by root cause
#620 opened May 4, 2026 by ajassani Collaborator • Draft
3 tasks
Runtime improvement for pseudo op extension
#617 opened Apr 30, 2026 by devalshahamd Contributor Loading…
3
fix(treeperf): preserve kernel time when perf model fails
#611 opened Apr 29, 2026 by unagiboi Collaborator Loading…
1
EventReplay: extend beyond aten ops with auto-import and custom initializers
#607 opened Apr 28, 2026 by ajassani Collaborator Loading…
2 tasks done
Adding Claude.md file for Tracelens documentation Improvements or additions to documentation
#606 opened Apr 28, 2026 by janmatai Contributor Loading…
fix(trace2tree): attach "bleeding" host events to a containing ancestor
#603 opened Apr 23, 2026 by unagiboi Collaborator Loading…
1
[Feat] Idle Time Analysis for the GPU Kernel Stream
#601 opened Apr 22, 2026 by spandoesai Collaborator • Draft
Syncing up with the main branch
#594 opened Apr 21, 2026 by devalshahamd Contributor Loading…
Fix/trace2tree: fix cross-rank GPU attribution and merged-trace hang
#577 opened Apr 1, 2026 by brieflynn Contributor Loading…
2
456 add support for traces generated by jax 08 suppress warning
#572 opened Mar 28, 2026 by devalshahamd Contributor Loading…
1
Add perf model coverage for DeepEP EP communication ops perf_model Add performance model for calculating TFLOPS/s and TB/s
#520 opened Mar 10, 2026 by gphuang Contributor Loading…
1
@gphuang
28
get_df_xla_perf raises KeyError for FP8 and s8 dtypes in XLA kernel operands
#447 opened Dec 12, 2025 by brieflynn Contributor Loading…
ProTip! Filter pull requests by the default branch with base:main.