Skip to content

Actions: Navi-AI-Lab/nvllm

Actions

pre-commit

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
pre-commit.yml
This workflow was disabled manually.
31 workflow runs
31 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
Feat/own the stack phase b pre-commit #31: Pull request #2 opened by Natfii
7s feat/own-the-stack-phase-b
feat/own-the-stack-phase-b
7s
fix(cute-paged): thread torch CUDA stream into CuTe DSL launches pre-commit #30: Commit a78ab51 pushed by Natfii
4m 18s main
main
4m 18s
feat: PIECEWISE CUDA graphs working on CuTe paged attention pre-commit #29: Commit 643a4fe pushed by Natfii
5m 12s main
main
5m 12s
fix: 5 CUDA graph infrastructure bugs, disable fusion pending kernel … pre-commit #28: Commit 9e0550c pushed by Natfii
4m 27s main
main
4m 27s
debug: disable fusion_active for baseline validation pre-commit #27: Commit 00d0a14 pushed by Natfii
4m 25s main
main
4m 25s
fix: save max_num_seqs at __init__ time, not during forward pre-commit #26: Commit 4960225 pushed by Natfii
4m 22s main
main
4m 22s
fix: Qwen3_5DecoderLayer skips parent __init__, needs own _fusion_bound pre-commit #25: Commit 4672044 pushed by Natfii
4m 3s main
main
4m 3s
feat(cuda-graphs): wire CuTe fusion into Qwen3NextAttention/DecoderLayer pre-commit #24: Commit 5ce5583 pushed by Natfii
4m 17s main
main
4m 17s
docs: CUDA graph pipeline state implementation plan pre-commit #23: Commit c757d98 pushed by Natfii
7m 48s main
main
7m 48s
docs:Update status of serve-nemotron.sh to 'Not Ready' pre-commit #22: Commit b273032 pushed by Natfii
8m 19s main
main
8m 19s
docs:Update AGENTS.md to emphasize Readme relevance pre-commit #21: Commit c8383ee pushed by Natfii
8m 46s main
main
8m 46s
docs(readme): recommend building from source over prebuilt images pre-commit #20: Commit 22783f5 pushed by Natfii
9m 12s main
main
9m 12s
chore(scripts): consolidate serve scripts, remove unused clutter pre-commit #19: Commit 0434cc1 pushed by Natfii
4m 56s main
main
4m 56s
pre-commit pre-commit #18: Commit 8c81280 pushed by Natfii
4m 34s main
main
4m 34s
fix(bench): correct misleading Phase B+C fusion GSM8K results pre-commit #17: Commit bdbb3ec pushed by Natfii
4m 47s main
main
4m 47s
feat(kernel): Phase C RMSNorm fusion — residual add + RMSNorm in CuTe… pre-commit #16: Commit ba5a7fd pushed by Natfii
4m 36s main
main
4m 36s
fix(kernel): Phase B W_O GEMV — DSL compilation fixes, GSM8K gate passes pre-commit #15: Commit 8056437 pushed by Natfii
4m 30s main
main
4m 30s
fix: hidden_dim 3584→5120 — got too excited claude pre-commit #14: Commit f5fce0d pushed by Natfii
4m 30s main
main
4m 30s
test: standalone test for fused attention + W_O GEMV vs reference pre-commit #13: Commit e712732 pushed by Natfii
4m 39s main
main
4m 39s
bench: re-baseline CuTe vs FlashInfer after .contiguous() elimination pre-commit #12: Commit e326143 pushed by Natfii
4m 30s main
main
4m 30s
feat(bench): CuTe vs FlashInfer kernel profile + Navi AI Labs attribu… pre-commit #11: Commit 6e78e8e pushed by Natfii
4m 33s main
main
4m 33s
feat(kernel): CuTe DSL paged attention decode kernel — first working … pre-commit #10: Commit 26bda34 pushed by Natfii
4m 41s main
main
4m 41s
feat(kernel): bridge datacenter Blackwell core instructions for SM121… pre-commit #9: Commit 5e8b9c6 pushed by Natfii
4m 45s main
main
4m 45s
docs: CuTe DSL kernel replacement spec v2 (7-expert audit) pre-commit #8: Commit 51187ce pushed by Natfii
4m 9s main
main
4m 9s
feat: CuTe paged attention v1 — prototype validated end-to-end pre-commit #7: Commit 5383121 pushed by Natfii
4m 29s main
main
4m 29s