Skip to content

[pull] master from mudler:master - #1453

Merged
pull[bot] merged 3 commits into
soitun:masterfrom
mudler:master
Aug 16, 2026
Merged

[pull] master from mudler:master#1453
pull[bot] merged 3 commits into
soitun:masterfrom
mudler:master

Conversation

@pull

@pull pull Bot commented Aug 16, 2026

Copy link
Copy Markdown

See Commits and Changes for more details.


Created by pull[bot] (v2.0.0-alpha.4)

Can you help keep this open source service alive? 💖 Please sponsor : )

mudler's LocalAI [bot] and others added 3 commits August 16, 2026 00:31
…1409cd1b027fbc9` (#11542)

⬆️ Update ikawrakow/ik_llama.cpp

Signed-off-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Co-authored-by: mudler <2420543+mudler@users.noreply.github.com>
…11541)

⬆️ Update vllm-project/vllm-metal (darwin)

Signed-off-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Co-authored-by: mudler <2420543+mudler@users.noreply.github.com>
Adds eight curated vllm-cpp entries to the model gallery. Until now the
backend had gallery coverage only for MiniMax-H3 video, so serving text on
it meant hand-writing engine_args.

The flagship tier is what vllm.cpp gates its correctness and speed claims
on: Qwen3.6-27B and Qwen3.6-35B-A3B in NVFP4, each with a speculative
sibling (MTP on both, DFlash on the 27B). Qwen3-Coder-30B-A3B covers
agentic tool use, and Qwen3-4B / Qwen3-0.6B in bf16 are the entries that
run where NVFP4 cannot, CPU included.

Three details are load-bearing rather than incidental:

- The 27B entries pin revision 890bdef7. That repository was later
  re-quantized in place from NVFP4 to FP8 W8A8 under the same name, so an
  unpinned entry resolves to different weights and reports nothing.
- Qwen3-Coder names tool_parser: qwen3_coder explicitly. Its dialect is
  byte-identical on the wire to step3p5's, so chat-template sniffing
  cannot separate them and auto-detection picks wrong.
- enable_prefix_caching is deliberately left unset everywhere. It defaults
  on for dense models and off for the GDN hybrids, and that per-model
  default is the right answer.

num_blocks is sized per model from its real KV footprint rather than
copied between entries, which ranges from 20 KiB/token on the 35B to
144 KiB/token on the 4B.

Docs: adds features/vllm-cpp.md covering installation, the model table,
the pinning rationale and how to choose between the speculative variants,
and cross-links it from the existing engine_args reference. It also
records that the CUDA images are built for Blackwell only, which is
narrower than vllm.cpp's own ten-architecture release and makes an
otherwise cryptic "no kernel image is available" failure legible.

Verified: gallery suite green; all eight decode and validate as a
ModelConfig. qwen3-0.6b-vllm-cpp confirmed end to end on a real cluster,
chat plus engine-parsed tool_calls. The NVFP4 entries are not yet
runtime-verified: no available node has kernels for them.


Assisted-by: Claude Code:claude-opus-5[1m] [Read] [Bash] [Edit]

Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
Co-authored-by: Ettore Di Giacinto <mudler@localai.io>
@pull pull Bot locked and limited conversation to collaborators Aug 16, 2026
@pull pull Bot added the ⤵️ pull label Aug 16, 2026
@pull
pull Bot merged commit 6fb9ab3 into soitun:master Aug 16, 2026
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

0 participants