Roy WangandClaude Fable 5 b74ac0bc61 [Bugfix] Allocate packed outputs in fused_q_kv_rmsnorm so q_b_proj keeps its low-latency GEMM path at decode (#53109) ...

Signed-off-by: yasong.wang <yasong.wang@inferact.ai>
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>

2026-08-28 02:08:27 +00:00

..

common

[Bugfix] Allocate packed outputs in fused_q_kv_rmsnorm so q_b_proj keeps its low-latency GEMM path at decode (#53109)

2026-08-28 02:08:27 +00:00

deepseek_v4

[ROCm][Perf] Fuse SWA q/kv RMSNorm and q FP8 group quant for DeepSeek-V4 (#53540)

2026-08-27 09:37:54 -07:00

deepseek_v32

[Attention] Enable masked MHA for GLM-5 head dimensions (#53785)

2026-08-27 21:17:12 +00:00

dots3_note

[Model] Fix KV cache layout and optimize Dots3 NOTE Omni encoders (#53460)

2026-08-23 10:02:35 -07:00

inkling

[Model Runner V2] batch-sharded sample (#50465)

2026-08-24 15:49:57 -07:00

kimi_k3

[Kimi-K3] Merge MLA gate into QKV-A projection (#54015)

2026-08-27 17:47:18 -07:00

minimax_m3

[Model Runner V2] batch-sharded sample (#50465)

2026-08-24 15:49:57 -07:00

__init__.py

[Model Refactoring] Migrate DeepSeek V4 to vllm/models/ [1/N] (#43004)

2026-05-18 19:50:02 -07:00