COMMITS
June 3, 2026
R
Add google/gemma-4-12B-it recipe (#507)
Roy Wang committed
L
[Gemma 4] Add Gemma 4 Unified 12B (encoder-free) to recipe (#506)
Luciano Martins committed
M
Update the recipe for Mistral-Medium-3.5-128B on MI300X (#490)
matti-palomaki-amd committed
J
[new model] add GLM-GA recipes (#489)
Jared Wen committed
M
Include documented TPU recipes (#477)
Mateusz Sokół committed
S
Add plamo recipe (#475)
Shinichi Hemmi committed
June 2, 2026
R
Add OpenBMB MiniCPM-V 4.6 + MiniCPM5-1B recipes (#505)
Roy Wang committed
R
Add NVIDIA Cosmos3 omnimodal world model recipes (#504)
Roy Wang committed
R
Add JetBrains Mellum2 recipes (Thinking + Instruct) (#503)
Roy Wang committed
V
MiniMax scope fuse_minimax_qk_norm to NVIDIA (#502)
vllmellm committed
May 31, 2026
R
Fix restricted hardware leaking into global preference (#496)
Roy Wang committed
May 30, 2026
R
Step-3.7-Flash: add DGX Station (GB300) support (#495)
Roy Wang committed
May 29, 2026
S
Add DGX Station guidance for DeepSeek V4 Flash (#493)
Serge Panev committed
R
Step-3.7-Flash: pin dedicated Docker image, prefer over pip (#494)
Roy Wang committed
L
[Model] add Step-3.7-Flash (#492)
ltd0924 committed
May 26, 2026
F
Fix DGX station memory (#488)
Faradawn Yang committed
M
mi300x verified tag for Ministral-3-14B-Instruct-2512 (#476)
matti-palomaki-amd committed
May 22, 2026
R
Render `vllm serve --omni` with task tabs for omni recipes (#487)
Roy Wang committed
R
Improve omni recipe UX and add brand-filtered dependencies (#486)
Roy Wang committed
R
Add platform self-host dialog with per-recipe opt-in (#485)
Roy Wang committed
May 21, 2026
S
Add DGX Station GB300 recipe support (#470)
Serge Panev committed
P
Remove --no-async-scheduling from pd_cluster strategy (#483)
Peter Pan committed
M
Add GLM-5.1 NVFP4 variant (#482)
Mohammad Miadh Angkad committed
L
update Gemma4 recipe for Intel cpu (#463)
Louie Tsai committed
H
[ROCm] update the minimax recipe to explicitly specify the attn backend (#481)
Hongxia Yang committed
May 20, 2026
A
Update MiniMax M2.5 H200 recipe (#474)
Anish Shanbhag committed
May 18, 2026
R
Update Qwen3.5 PD recipe (#473)
Roy Wang committed
May 17, 2026
R
Promote variants to HF-URL JSON endpoints + PD fitness check (#472)
Roy Wang committed
R
Enable pd_cluster strategy for large MoE models (#471)
Roy Wang committed
May 15, 2026
R
Add internlm/Intern-S2-Preview recipe (#469)
Roy Wang committed