Skip to content

Pull requests: cactus-compute/cactus

New pull request New
13 Open 364 Closed
13 Open 364 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

Cactus torch api clean
#529 opened Mar 21, 2026 by cattermelon1234 Loading…
8
Update docs site_url to docs.cactuscompute.com
#527 opened Mar 20, 2026 by rshemet Loading…
Per-layer KV heads, attention logit capping, MoE per-expert scales, NPU multi-input
#526 opened Mar 19, 2026 by ncylich Loading…
4 tasks done
20
Accelerate MatMul FP16 for Apple GPUs
#523 opened Mar 17, 2026 by aarav18 Loading…
reverting attn exp calculations to before 3n
#511 opened Mar 9, 2026 by ncylich Loading…
1
Fix gemma multi tool call and logit biasing
#510 opened Mar 8, 2026 by lennartvoelz Loading…
1
new approximation for exponent on (0,1)
#500 opened Mar 6, 2026 by kar-m Loading…
9
add Android benchmarking harness with battery and thermal metrics
#494 opened Mar 5, 2026 by Vinesh2929 Loading…
Optimized Attention
#480 opened Mar 2, 2026 by ncylich Loading…
10
Axis reductions fixes
#473 opened Feb 28, 2026 by cattermelon1234 Loading…
3
refactor(converter): optimize MoE expert weight conversion and naming (#468)
#469 opened Feb 27, 2026 by UsmanCh1226 Loading…
1
Benchmarking against other quantized kernels
#458 opened Feb 26, 2026 by ncylich Loading…
2
Whisper/ASR cloud handoff
#448 opened Feb 25, 2026 by ammesatyajit Loading…
1
ProTip! Find all pull requests that aren't related to any open issues with -linked:issue.