Star16
93
Code
Code
Issues34
Pull Requests42
Actions
Discussion
Wiki
Analysis
Star16
93
Merged
feat(triton): add chunk_gated_delta_rule_fwd_kkt_solve #135
gcw_K5CvmS79Add on Aug 25
feat(triton): add chunk_gated_delta_rule_fwd_kkt_solve #135
Merged
gcw_K5CvmS79Add on Aug 25
Merged fromgcw_K5CvmS79/MindSpeed-Opsfeature/chunk_gated_delta_rule_fwd_kkt_solve_kernelintoAscend/MindSpeed-Opsmaster
2026-09-21
style: one-arg-per-line in the two CPU-golden calls (ruff-format)
Ggcw_K5CvmS79Submitted on Sep 22
04165bb4
test: run the golden reference on CPU
Ggcw_K5CvmS79Submitted on Sep 21
2b72f0f6
docs/ut: drop stale top-level routing references
Ggcw_K5CvmS79Submitted on Sep 21
3d219e48
chore(api): leave chunk_gated_delta_rule.py untouched vs master
Ggcw_K5CvmS79Submitted on Sep 21
f9a9b082
fix(api): keep chunk_gated_delta_rule_fwd on the stock two-op chain
Ggcw_K5CvmS79Submitted on Sep 21
dc857f12
2026-09-20
style: fix pre-commit findings (ruff E741/F401/format, markdown MD040)
Ggcw_K5CvmS79Submitted on Sep 20
45bbb432
docs(triton): mark the formula code block as text (markdownlint MD040)
Ggcw_K5CvmS79Submitted on Sep 20
31a4edc1
docs(triton): sync operator doc with the final kernel version
Ggcw_K5CvmS79Submitted on Sep 20
4273fb20
feature(triton): fused chunk_gated_delta_rule_fwd_kkt_solve operator
Ggcw_K5CvmS79Submitted on Sep 20
715a6015
Total: 9