Skip to content

Pull requests: antirez/ds4

New pull request New
224 Open 159 Closed
224 Open 159 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

cuda: copy model into VRAM on single-GPU via --gpu-resident
#622 opened Jul 28, 2026 by pvaccarello Loading…
6 of 7 tasks
Support AProjQ4 GGUFs: Q4_K dense attention projections
#621 opened Jul 28, 2026 by GiorgioOppo Loading…
2
allow overriding FORCE_HF_DOWNLOAD via environment variable
#619 opened Jul 28, 2026 by alexlipa91 Loading…
Rocm prefill performance improvement on 16.3%
#616 opened Jul 26, 2026 by Farenheith • Draft
Fix Laguna KV ring commits for oversized prefills
#614 opened Jul 26, 2026 by christophsturm Loading…
Update Laguna to corrected GGUF revision
#613 opened Jul 26, 2026 by christophsturm Loading…
Server: chat/completions live tool continuation via tool_call_id
#611 opened Jul 26, 2026 by starforge-labs Loading…
1
CUDA: resident expert cache and faster selected-expert uploads for --ssd-streaming decode
#605 opened Jul 25, 2026 by iCreil Loading…
Server: fused single-GPU decode for --batched-session (co-decode 2-4 sessions in one graph pass)
#604 opened Jul 25, 2026 by iCreil Loading…
rocm: AMD Instinct (CDNA/gfx90a) support with multi-GPU device placement
#602 opened Jul 24, 2026 by ewindisch Loading…
Fix batched prefill decode starvation
#601 opened Jul 24, 2026 by krunkosaurus • Draft
1
rocm: fix build on gfx1201 (RDNA4, R9700) — gate the WMMA v1 kernel
#599 opened Jul 24, 2026 by cm999club Loading…
1
Add CUDA backend for Laguna S 2.1
#594 opened Jul 23, 2026 by devteapot • Draft
fix(cuda): avoid null output-head dereference in distributed coordinators
#592 opened Jul 22, 2026 by hy-sde Loading…
1
server: disable tool calls during compaction requests
#591 opened Jul 21, 2026 by ober Loading…
DSpark: replay-free partial accepts + encoder-batched captures (+40% greedy, byte-lossless)
#590 opened Jul 21, 2026 by gilbert-barajas Loading…
3
Support SSD streaming and MTP drafter when using two GPUs and caching of experts in VRAM
#589 opened Jul 21, 2026 by iSevenDays Loading…
speed-bench: add M5 Max (Q2+Q4) results
#587 opened Jul 21, 2026 by thomasjohnflaherty Loading…
1
speed-bench: add Apple M5 Max results
#584 opened Jul 21, 2026 by altdotboy Loading…
README: add dark-mode logo variant
#583 opened Jul 20, 2026 by moretti Loading…
server: emit GLM tool-call args with their real JSON types (fixes #569)
#582 opened Jul 20, 2026 by gilbert-barajas Loading…
1
1
fix: Hermes agent cache token-mismatch
#581 opened Jul 20, 2026 by vestigecounty Loading…
rax.c: add OOM guard in raxRemove for incomplete ancestor stack
#580 opened Jul 20, 2026 by ft-ioxcs Loading…
Add styles for light and dark logo modes
#579 opened Jul 20, 2026 by S1M0N38 Loading…
2
metal: restore streamed-expert mlock pinning (fixes #532)
#578 opened Jul 20, 2026 by gilbert-barajas Loading…
1
ProTip! What’s not been updated in a month: updated:<2026-06-28.