Skip to content

Pull requests: JustVugg/colibri

New pull request New
37 Open 516 Closed
37 Open 516 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

The VRAM expert tier could only be fed by disk reads
#877 opened Aug 7, 2026 by ZacharyZcR Contributor Loading…
Avoid reserving DSpark cache for unsupported DeepSeek-V4 checkpoints
#875 opened Aug 7, 2026 by waizuichougou Loading…
A CUDA fallback is a property of the call, not of the tensor bug Difetto verificato nel codice cuda Backend CUDA/NVIDIA
#849 opened Aug 6, 2026 by ZacharyZcR Contributor Loading…
3
docs: M1 Ultra fmt=2 Metal perf report + benchmarks row + fmt4→fmt2 converter benchmark Datapoint di misurazione hardware docs Documentazione metal Backend Metal/Apple
#836 opened Aug 5, 2026 by boblaublaw Loading…
inkling: ring-buffer KV cache for sliding-window layers (~10x less KV memory at long context) enhancement New feature or request performance Velocità / tok-s / ottimizzazioni
#830 opened Aug 4, 2026 by dpanelli Loading…
2
fix(metal): int4-g64 (fmt=4) models never used the GPU — enable MoE experts + fused attention bug Difetto verificato nel codice metal Backend Metal/Apple needs-rebase Confligge, serve rebase dell'autore
#829 opened Aug 4, 2026 by aaristov Loading…
1
engine: notice when a tensor's format silently disables the fused Metal decode path enhancement New feature or request metal Backend Metal/Apple
#827 opened Aug 4, 2026 by monotophic Contributor Loading…
1
Feat/dstorage transport enhancement New feature or request performance Velocità / tok-s / ottimizzazioni
#825 opened Aug 4, 2026 by khalilswdp Contributor Loading…
4 of 5 tasks
managed runner, deterministic benchmark, and causal evidence (PR3 draft of #377 split) discussion Proposta / discussione aperta, non un task enhancement New feature or request performance Velocità / tok-s / ottimizzazioni
#822 opened Aug 4, 2026 by BColsey • Draft
ramdisk: headless planning, staging, mounts, recovery + tokenized CLI (PR2 of #377 split) discussion Proposta / discussione aperta, non un task enhancement New feature or request performance Velocità / tok-s / ottimizzazioni
#821 opened Aug 4, 2026 by BColsey • Draft
engine: RAMMAP, NUMA, and telemetry (PR1 of #377 split) discussion Proposta / discussione aperta, non un task enhancement New feature or request performance Velocità / tok-s / ottimizzazioni
#820 opened Aug 4, 2026 by BColsey • Draft
Add a registry for the 214 environment variables, and check the environment against it enhancement New feature or request
#800 opened Aug 3, 2026 by ZacharyZcR Contributor Loading…
3
Dead code, de-duplication, repo-wide clang-format, and lint gates in CI enhancement New feature or request needs-rebase Confligge, serve rebase dell'autore
#798 opened Aug 3, 2026 by ZacharyZcR Contributor Loading…
2
Metal (Apple GPU) backend for Kimi K3 feature Nuova funzionalità metal Backend Metal/Apple model-support Supporto a nuovi modelli
#790 opened Aug 2, 2026 by RDouglasSharp Contributor Loading…
2
[WIP] Vulkan stats enhancement New feature or request vulkan Backend Vulkan/AMD
#789 opened Aug 2, 2026 by Neppord Loading…
5 tasks
1
feature: Hy3 support discussion Proposta / discussione aperta, non un task model-support Supporto a nuovi modelli
#775 opened Aug 2, 2026 by ErikTromp Loading…
5 tasks done
3
DeepSeek V4: the CUDA kernels, as a tier rather than an engine discussion Proposta / discussione aperta, non un task model-support Supporto a nuovi modelli
#772 opened Aug 2, 2026 by ZacharyZcR Contributor Loading…
9
Vulkan MoE GEMV backend for integrated/AMD GPUs (draft, complements #418) enhancement New feature or request feature Nuova funzionalità vulkan Backend Vulkan/AMD
#729 opened Jul 31, 2026 by minne100 • Draft
1
feat(qwen36): CUDA VRAM expert tier — heat-based placement across GPUs via the shared CUDA backend cuda Backend CUDA/NVIDIA model-support Supporto a nuovi modelli
#713 opened Jul 30, 2026 by kreuzzelg Contributor • Draft
2
feat(qwen36): Qwen3.6-35B-A3B engine (CPU): hybrid Gated Attention + Gated DeltaNet + streaming MoE model-support Supporto a nuovi modelli
#712 opened Jul 30, 2026 by kreuzzelg Contributor Loading…
16
feat(win): fix silent CPU fallback, launcher suite, DirectStorage expert loads enhancement New feature or request needs-rebase Confligge, serve rebase dell'autore
#670 opened Jul 28, 2026 by khalilswdp Contributor Loading…
8 tasks done
3
feat(core): model-architecture seam, chat templates, text antiprompt enhancement New feature or request needs-rebase Confligge, serve rebase dell'autore
#667 opened Jul 28, 2026 by khalilswdp Contributor Loading…
5 tasks done
8
QLoRA training path: fine-tune GLM-5.2 (744B) in 64 GB RAM enhancement New feature or request feature Nuova funzionalità
#626 opened Jul 26, 2026 by pavolbauer Loading…
5 tasks done
8
WIP: MiniMax-M3 support — GQA + MSA block-sparse attention, o200k tokenizer, converter (follow-up to #418) model-support Supporto a nuovi modelli
#601 opened Jul 24, 2026 by steve-m Contributor Loading…
5 tasks
3
Metal fmt=4 grouped-int4 decode: attention + routed experts (#585) metal Backend Metal/Apple needs-rebase Confligge, serve rebase dell'autore
#587 opened Jul 24, 2026 by RDouglasSharp Contributor Loading…
31
ProTip! Adding no:label will show everything without a label.