Pull requests: antirez/ds4

New pull request New

128 Open 127 Closed

Pull requests list

rocm: fix distributed inference on unified-memory APUs (strix halo / gfx1151)

#407 opened Jun 13, 2026 by kyuz0

[3/N] add prefetch support for CUDA backend : running ds4 for any GPU with cache (2.75 x faster!)

#402 opened Jun 12, 2026 by yiakwy-xpu-ml-framework-team

Local gen with dist prefill

#401 opened Jun 12, 2026 by lobanov • Draft

Extract dashboard into standalone ds4_dashboard module

#400 opened Jun 12, 2026 by EdwQ

Add multi-column attn-out low projection kernel for small batches

#399 opened Jun 11, 2026 by rwl4

CUDA backend (DGX-Spark) — refactored into modular .cuh files mirroring ROCm structure

#398 opened Jun 11, 2026 by gundemirbas

ROCm runtime: configurable weight cache limit and arena chunk size via environment variables

#397 opened Jun 11, 2026 by gundemirbas

Add env-gated prompt-lookup speculative decoding for greedy generation

#396 opened Jun 11, 2026 by rwl4

Add DwarfStar logo SVG

#395 opened Jun 11, 2026 by darchidev

fix(kv-cache): refresh cold anchor after partial prefix hits

#394 opened Jun 11, 2026 by TerryChengTW

3 tasks done

Add teaching mode to ds4-agent, with teach-bench benchmark

#391 opened Jun 11, 2026 by rowantrollope

Skip MTP probes while drafts go unused

#385 opened Jun 10, 2026 by pandysp Contributor

ds4_agent: fix rocm support

#383 opened Jun 10, 2026 by mcmalayalam

Harden CDP page target creation

#382 opened Jun 10, 2026 by sk8erboi17

Clamp MTP draft depth to the prefill capacity

#381 opened Jun 10, 2026 by pandysp Contributor

feat: add native Agent Skills support to ds4-agent

#380 opened Jun 10, 2026 by fry69 Contributor

feat: add runtime file discovery chain

#379 opened Jun 10, 2026 by fry69 Contributor

Keep live KV reusable when clients strip transient metadata blocks

#378 opened Jun 10, 2026 by adv0r

[2/N] add cuda imatrix support for custom RL model

#377 opened Jun 10, 2026 by yiakwy-xpu-ml-framework-team

server: abort prefill when the client disconnects

#376 opened Jun 9, 2026 by adv0r

ds4_server: Add /health endpoint that returns HTTP 200 once model is fully loaded

#374 opened Jun 9, 2026 by mcmalayalam

Fix agent edit: accept [upto] markers indented or padded with blanks (+ golden cases)

#373 opened Jun 9, 2026 by rinaldofesta

Add continuous depth-1 MTP speculation (DS4_MTP_CONTINUOUS)

#371 opened Jun 9, 2026 by pandysp Contributor

[1/N] add fp8 fp32 scale support for custom RL model

#368 opened Jun 9, 2026 by yiakwy-xpu-ml-framework-team

make: consistent ROCm targets (rocm-strix-halo / rocm-generic) + portable lib paths (#357, #179)

#365 opened Jun 8, 2026 by jamesburton

ProTip! Type g i on any issue or pull request to go back to the issue listing page.

Navigation Menu

Search code, repositories, users, issues, pull requests...

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Pull requests: antirez/ds4

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Pull requests list