3 people 6fe67cbbf3 [Spec][V2] Support MTP speculative decoding under pipeline parallelism (#46994) ...

Signed-off-by: Chris Eastwood <chris.eastwood@pwn4g3.dev>
Signed-off-by: Nick Hill <nickhill123@gmail.com>
Co-authored-by: Nick Hill <nickhill123@gmail.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>

2026-09-11 17:55:21 +00:00

..

transformers

[Core][Model] Unify XD-RoPE into M-RoPE and derive the channel count (#56078)

2026-09-09 09:23:30 -07:00

__init__.py

[Bugfix][Model Runner V2] Restore multimodal draft capability detection (#50417)

2026-08-03 14:18:14 -07:00

adapters.py

Reduce AutoWeightsLoader kwargs (#53106)

2026-08-20 15:40:11 +00:00

afmoe.py

[Refactor] Delete dead code in models (#51838)

2026-08-11 15:44:26 -04:00

aimv2.py

Reduce AutoWeightsLoader kwargs (#53106)

2026-08-20 15:40:11 +00:00

apertus.py

[Model] Remove ten deprecated model architectures (#53608)

2026-08-24 17:06:00 -07:00

arcee.py

Reduce AutoWeightsLoader kwargs (#53106)

2026-08-20 15:40:11 +00:00

aria.py

Reduce AutoWeightsLoader kwargs (#53106)

2026-08-20 15:40:11 +00:00

audioflamingo3.py

[MM] Simplify prompt updates: replace PromptSeq with list[int] (#53372)

2026-08-23 14:50:32 +08:00

AXK1.py

[MOE] Standardize and abstract fused shared expert optimization selection (#51695)

2026-08-18 12:56:56 -04:00

bagel.py

[MM] Remove text components from ProcessorInputs (#53093)

2026-08-20 20:42:53 -07:00

bailing_moe_linear.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[aA][bB]" (#48977)

2026-08-06 19:30:15 +00:00

bailing_moe_mtp.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[aA][bB]" (#48977)

2026-08-06 19:30:15 +00:00

bailing_moe_v3_mtp.py

[Model] Add Bailing V3 VL support (#55921)

2026-09-10 02:46:56 -07:00

bailing_moe_v3_vl.py

[Model] Add Bailing V3 VL support (#55921)

2026-09-10 02:46:56 -07:00

bailing_moe_v3.py

[Model] Add Bailing V3 VL support (#55921)

2026-09-10 02:46:56 -07:00

bailing_moe.py

Fix weight tying (#51665)

2026-08-20 10:31:55 +00:00

bee.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[aA][bB]" (#48977)

2026-08-06 19:30:15 +00:00

bert_with_rope.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[aA][bB]" (#48977)

2026-08-06 19:30:15 +00:00

bert.py

Reduce AutoWeightsLoader kwargs (#53106)

2026-08-20 15:40:11 +00:00

blip2.py

[Perf] Use SDPA for BLIP-2 Q-Former attention (#55285)

2026-09-06 07:04:19 +00:00

blip.py

Reduce AutoWeightsLoader kwargs (#53106)

2026-08-20 15:40:11 +00:00

bloom.py

Fix weight tying (#51665)

2026-08-20 10:31:55 +00:00

chatglm.py

[Mypy] Fix mypy typing for model interfaces and H/I models (#54079)

2026-08-28 16:49:19 -04:00

clip.py

[Bugfix][Multimodal] Encode text in mixed CLIP/SigLIP pooling batches (#53165)

2026-08-24 10:45:42 -07:00

cohere2_moe.py

Reduce AutoWeightsLoader kwargs (#53106)

2026-08-20 15:40:11 +00:00

cohere2_vision.py

[MM] Simplify prompt updates: replace PromptSeq with list[int] (#53372)

2026-08-23 14:50:32 +08:00

cohere_asr.py

[Bugfix][Model] Fix CohereASR streaming audio-token estimate (unit + subsampling) (#53829)

2026-09-02 08:55:16 -07:00

cohere_compass.py

[MM][CG] Enable encoder CUDA Graph for MiniCPM-V (#42785)

2026-09-10 06:28:48 +00:00

cohere_eagle.py

Reduce AutoWeightsLoader kwargs (#53106)

2026-08-20 15:40:11 +00:00

colbert.py

Reduce AutoWeightsLoader kwargs (#53106)

2026-08-20 15:40:11 +00:00

colmodernvbert.py

[MM] Simplify _apply_hf_processor_main (#53275)

2026-08-22 01:49:37 -07:00

colpali.py

[MM] Simplify _apply_hf_processor_main (#53275)

2026-08-22 01:49:37 -07:00

colqwen3_5.py

[Bugfix] Fix ColQwen3.5 pooler projector initialization (#54847)

2026-09-02 10:35:43 +08:00

colqwen3.py

[EPD] Allow KV consumers to omit MM embeddings (#52697)

2026-08-19 02:16:08 +00:00

commandr.py

[Mypy] Fix mypy typing for model interfaces and H/I models (#54079)

2026-08-28 16:49:19 -04:00

config.py

[Model] Support DeepSeek-V4.1-Flash (#56214)

2026-09-11 02:11:19 -07:00

conformer_encoder.py

[Perf] Accumulate Conformer attention scores with baddbmm (#55062)

2026-09-03 04:17:52 -07:00

cosmos3_edge.py

[Bugfix] Fix Qwen3-VL and Cosmos3-Edge text architectures for CPU and pipeline parallelism (#53699)

2026-09-11 05:58:15 -07:00

cosmos3.py

Cosmos3 FP8 ModelOpt/Diffusers remapping (#48952)

2026-07-20 11:31:48 -07:00

dbrx.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[cC][dD]" (#52003)

2026-08-13 11:33:55 +08:00

deepencoder2.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[cC][dD]" (#52003)

2026-08-13 11:33:55 +08:00

deepencoder.py

[Perf] Fuse DeepEncoder relative bias in Triton attention (#55629)

2026-09-08 13:22:04 +00:00

deepseek_eagle3.py

[Core][MRV2] Support eagle3 spec decode with pipeline parallel (#50514)

2026-09-05 11:45:34 +08:00

deepseek_eagle.py

Reduce AutoWeightsLoader kwargs (#53106)

2026-08-20 15:40:11 +00:00

deepseek_mtp.py

[Spec][V2] Support MTP speculative decoding under pipeline parallelism (#46994)

2026-09-11 17:55:21 +00:00

deepseek_ocr2.py

[MM] Simplify _apply_hf_processor_main (#53275)

2026-08-22 01:49:37 -07:00

deepseek_ocr.py

[MM][CG] Enable encoder CUDA Graph for MiniCPM-V (#42785)

2026-09-10 06:28:48 +00:00

deepseek_v2.py

[Hy4] support Hy4-preview model (#54160)

2026-08-29 12:07:28 +08:00

deepseek_vl2.py

[MM] Simplify _apply_hf_processor_main (#53275)

2026-08-22 01:49:37 -07:00

diffusion_gemma.py

[1/2][Model Runner V2] DBO support, eager mode (#50945)

2026-09-04 14:17:39 -07:00

dots_ocr.py

[Bugfix][LoRA] Use an explicit capability flag for tower connector LoRA (#53092)

2026-08-21 09:45:54 +00:00

eagle2_5_vl.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[eE][fF]" (#53381)

2026-08-22 15:41:27 -04:00

ernie45_moe.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[eE][fF]" (#53381)

2026-08-22 15:41:27 -04:00

ernie45_vl_moe.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[eE][fF]" (#53381)

2026-08-22 15:41:27 -04:00

ernie45_vl.py

[MM][CG] Enable encoder CUDA Graph for MiniCPM-V (#42785)

2026-09-10 06:28:48 +00:00

ernie45.py

…

ernie_mtp.py

Fix weight tying (#51665)

2026-08-20 10:31:55 +00:00

exaone4_5_mtp.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[eE][fF]" (#53381)

2026-08-22 15:41:27 -04:00

exaone4_5.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[eE][fF]" (#53381)

2026-08-22 15:41:27 -04:00

exaone4.py

Fix weight tying (#51665)

2026-08-20 10:31:55 +00:00

exaone_moe_mtp.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[eE][fF]" (#53381)

2026-08-22 15:41:27 -04:00

exaone_moe.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[eE][fF]" (#53381)

2026-08-22 15:41:27 -04:00

exaone.py

Fix weight tying (#51665)

2026-08-20 10:31:55 +00:00

extract_hidden_states.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[eE][fF]" (#53381)

2026-08-22 15:41:27 -04:00

falcon_h1.py

[XPU][Bugfix] Fix FalconH1 pipeline-parallel execution (#55913)

2026-09-10 17:47:29 +08:00

falcon.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[eE][fF]" (#53381)

2026-08-22 15:41:27 -04:00

fireredasr2.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[eE][fF]" (#53381)

2026-08-22 15:41:27 -04:00

funasr.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[eE][fF]" (#53381)

2026-08-22 15:41:27 -04:00

funaudiochat.py

[MM] Simplify prompt updates: replace PromptSeq with list[int] (#53372)

2026-08-23 14:50:32 +08:00

gemma2.py

Fix weight tying (#51665)

2026-08-20 10:31:55 +00:00

gemma3_mm.py

[MM][CG] Enable encoder CUDA Graph for MiniCPM-V (#42785)

2026-09-10 06:28:48 +00:00

gemma3.py

Fix weight tying (#51665)

2026-08-20 10:31:55 +00:00

gemma3n_audio_utils.py

[Bugfix] Fix integer overflow in Gemma3n audio processing (#31657)

2026-01-10 17:52:53 +08:00

gemma3n_mm.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[gG]" (#53616)

2026-08-25 09:47:16 -07:00

gemma3n.py

[Bugfix][Gemma] Conditionally create KV projections/norms on KV-shared layers (#54917)

2026-09-07 15:04:00 +01:00

gemma4_dspark.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[gG]" (#53616)

2026-08-25 09:47:16 -07:00

gemma4_mm.py

[MM][CG] Enable encoder CUDA Graph for MiniCPM-V (#42785)

2026-09-10 06:28:48 +00:00

gemma4_mtp.py

[Bugfix] Fix FlashInfer KV sharing with omitted K/V (#55864)

2026-09-10 23:07:52 -07:00

gemma4_unified.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[gG]" (#53616)

2026-08-25 09:47:16 -07:00

gemma4.py

[Bugfix][Gemma] Conditionally create KV projections/norms on KV-shared layers (#54917)

2026-09-07 15:04:00 +01:00

gemma.py

[Mypy] Fix mypy typing for model interfaces and H/I models (#54079)

2026-08-28 16:49:19 -04:00

glm4_1v.py

[MM][CG] Enable encoder CUDA Graph for MiniCPM-V (#42785)

2026-09-10 06:28:48 +00:00

glm4_moe_lite_mtp.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[gG]" (#53616)

2026-08-25 09:47:16 -07:00

glm4_moe_lite.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[gG]" (#53616)

2026-08-25 09:47:16 -07:00

glm4_moe_mtp.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[gG]" (#53616)

2026-08-25 09:47:16 -07:00

glm4_moe.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[gG]" (#53616)

2026-08-25 09:47:16 -07:00

glm4.py

Reduce AutoWeightsLoader kwargs (#53106)

2026-08-20 15:40:11 +00:00

glm4v.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[gG]" (#53616)

2026-08-25 09:47:16 -07:00

glm_ocr_mtp.py

[Bugfix] Fix GLM-OCR MTP position masking during CUDA graph capture (#56447)

2026-09-11 12:31:10 +00:00

glm_ocr.py

[Perf] Ensure async h2d copies are pinned in more places (#55202)

2026-09-04 17:17:01 -07:00

glm.py

…

glmasr_utils.py

[Refactor] Remove dead private func _fp8_perm and _extract_mask_for_item (#35068)

2026-02-23 05:05:20 -08:00

glmasr.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[gG]" (#53616)

2026-08-25 09:47:16 -07:00

gpt2.py

Reduce AutoWeightsLoader kwargs (#53106)

2026-08-20 15:40:11 +00:00

gpt_j.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[gG]" (#53616)

2026-08-25 09:47:16 -07:00

gpt_neox.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[gG]" (#53616)

2026-08-25 09:47:16 -07:00

gpt_oss.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[gG]" (#53616)

2026-08-25 09:47:16 -07:00

granite4_vision.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[gG]" (#53616)

2026-08-25 09:47:16 -07:00

granite_speech_plus.py

Add model support for granite speech plus (#43519)

2026-06-04 14:47:48 +00:00

granite_speech.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[gG]" (#53616)

2026-08-25 09:47:16 -07:00

granite.py

[Mypy] Fix mypy typing for model interfaces and H/I models (#54079)

2026-08-28 16:49:19 -04:00

granitemoe.py

Fix weight tying (#51665)

2026-08-20 10:31:55 +00:00

granitemoehybrid.py

[Bugfix][Model] Fix GraniteMoeHybrid per-expert quantized weight loading (#54052)

2026-09-01 12:38:53 +00:00

granitemoeshared.py

Fix weight tying (#51665)

2026-08-20 10:31:55 +00:00

h2ovl.py

[Mypy] Fix mypy typing for model interfaces and H/I models (#54079)

2026-08-28 16:49:19 -04:00

hrm_text.py

[Mypy] Fix mypy typing for model interfaces and H/I models (#54079)

2026-08-28 16:49:19 -04:00

hy_v3_mtp.py

[Mypy] Fix mypy typing for model interfaces and H/I models (#54079)

2026-08-28 16:49:19 -04:00

hy_v3.py

[Mypy] Fix mypy typing for model interfaces and H/I models (#54079)

2026-08-28 16:49:19 -04:00

hyperclovax_vision_v2.py

[Mypy] Fix mypy typing for model interfaces and H/I models (#54079)

2026-08-28 16:49:19 -04:00

hyperclovax.py

Fix weight tying (#51665)

2026-08-20 10:31:55 +00:00

idefics2_vision_model.py

[MM][CG] Enable encoder CUDA Graph for MiniCPM-V (#42785)

2026-09-10 06:28:48 +00:00

idefics3.py

[MM][CG] Enable encoder CUDA Graph for MiniCPM-V (#42785)

2026-09-10 06:28:48 +00:00

interfaces_base.py

[Mypy] Fix mypy typing for model interfaces and H/I models (#54079)

2026-08-28 16:49:19 -04:00

interfaces.py

[MM][CG] Enable encoder CUDA Graph for MiniCPM-V (#42785)

2026-09-10 06:28:48 +00:00

intern_vit.py

Remove even more unnecessary load_weights methods (#48496)

2026-07-18 08:40:27 +00:00

internlm2.py

[Mypy] Fix mypy typing for model interfaces and H/I models (#54079)

2026-08-28 16:49:19 -04:00

interns1_pro.py

[Mypy] Fix mypy typing for model interfaces and H/I models (#54079)

2026-08-28 16:49:19 -04:00

interns1_vit.py

Remove even more unnecessary load_weights methods (#48496)

2026-07-18 08:40:27 +00:00

interns1.py

[Mypy] Fix mypy typing for model interfaces and H/I models (#54079)

2026-08-28 16:49:19 -04:00

interns2_mobius.py

[Mypy] Fix mypy typing for model interfaces and H/I models (#54079)

2026-08-28 16:49:19 -04:00

interns2_preview.py

Reduce AutoWeightsLoader kwargs (#53106)

2026-08-20 15:40:11 +00:00

internvl.py

[MM][CG] Enable encoder CUDA Graph for MiniCPM-V (#42785)

2026-09-10 06:28:48 +00:00

iquest_loopcoder.py

[Mypy] Fix mypy typing for model interfaces and H/I models (#54079)

2026-08-28 16:49:19 -04:00

isaac.py

[Perf] Ensure async h2d copies are pinned in more places (#55202)

2026-09-04 17:17:01 -07:00

jais2.py

Fix weight tying (#51665)

2026-08-20 10:31:55 +00:00

jamba.py

[Mypy] Fix typing for J models (#54130)

2026-08-29 08:50:15 -04:00

jina_vl.py

[Bugfix][MM] Fix JinaVL processing cache order (#53553)

2026-08-25 00:03:05 -07:00

jina.py

[Mypy] Fix typing for J models (#54130)

2026-08-29 08:50:15 -04:00

k2_horizon.py

[Model] Add K2-Horizon model support (#55063)

2026-09-03 01:17:08 -07:00

kanana_v.py

[Perf] Pin CPU tensors before non_blocking H2D in three MM paths (#54292)

2026-08-28 20:13:40 -07:00

keye_vl1_5.py

[MM] Simplify _apply_hf_processor_main (#53275)

2026-08-22 01:49:37 -07:00

keye.py

[Perf] Ensure async h2d copies are pinned in more places (#55202)

2026-09-04 17:17:01 -07:00

kimi_audio.py

[MM] Simplify _apply_hf_processor_main (#53275)

2026-08-22 01:49:37 -07:00

kimi_k25_vit.py

[Perf] Pin CPU tensors before non_blocking H2D in three MM paths (#54292)

2026-08-28 20:13:40 -07:00

kimi_k25.py

[MM][CG] Enable encoder CUDA Graph for MiniCPM-V (#42785)

2026-09-10 06:28:48 +00:00

kimi_vl.py

[MM][CG] Enable encoder CUDA Graph for MiniCPM-V (#42785)

2026-09-10 06:28:48 +00:00

laguna_dflash.py

[Core][MRV2] Support eagle3 spec decode with pipeline parallel (#50514)

2026-09-05 11:45:34 +08:00

laguna.py

[Mypy] Fix mypy typing for L models (#54177)

2026-09-04 07:50:43 -04:00

lfm2_moe.py

Fix weight tying (#51665)

2026-08-20 10:31:55 +00:00

lfm2_siglip2.py

Reduce AutoWeightsLoader kwargs (#53106)

2026-08-20 15:40:11 +00:00

lfm2_vl.py

[MM][CG] Enable encoder CUDA Graph for MiniCPM-V (#42785)

2026-09-10 06:28:48 +00:00

lfm2.py

Fix weight tying (#51665)

2026-08-20 10:31:55 +00:00

lightonocr.py

[Mypy] Fix mypy typing for L models (#54177)

2026-09-04 07:50:43 -04:00

llama4_eagle.py

[Mypy] Fix mypy typing for L models (#54177)

2026-09-04 07:50:43 -04:00

llama4.py

[Mypy] Fix mypy typing for L models (#54177)

2026-09-04 07:50:43 -04:00

llama_eagle3.py

[Core][MRV2] Support eagle3 spec decode with pipeline parallel (#50514)

2026-09-05 11:45:34 +08:00

llama_eagle.py

[Mypy] Fix mypy typing for L models (#54177)

2026-09-04 07:50:43 -04:00

llama.py

[Core][MRV2] Support eagle3 spec decode with pipeline parallel (#50514)

2026-09-05 11:45:34 +08:00

llava_next_video.py

[Mypy] Fix mypy typing for L models (#54177)

2026-09-04 07:50:43 -04:00

llava_next.py

[Mypy] Fix mypy typing for L models (#54177)

2026-09-04 07:50:43 -04:00

llava_onevision2.py

[Mypy] Fix mypy typing for L models (#54177)

2026-09-04 07:50:43 -04:00

llava_onevision.py

[Mypy] Fix mypy typing for L models (#54177)

2026-09-04 07:50:43 -04:00

llava.py

[Mypy] Fix mypy typing for L models (#54177)

2026-09-04 07:50:43 -04:00

longcat_flash_mtp.py

[Mypy] Fix mypy typing for L models (#54177)

2026-09-04 07:50:43 -04:00

longcat_flash_ngram.py

[Mypy] Fix mypy typing for L models (#54177)

2026-09-04 07:50:43 -04:00

longcat_flash.py

[Mypy] Fix mypy typing for L models (#54177)

2026-09-04 07:50:43 -04:00

mamba2.py

Remove unnecessary load_weights methods (#44589)

2026-06-29 01:52:23 -07:00

mamba.py

[Mypy] Fix typing for M models (#54262)

2026-09-01 11:26:00 -04:00

medusa.py

[Mypy] Fix typing for M models (#54262)

2026-09-01 11:26:00 -04:00

mellum.py

[Mypy] Fix typing for M models (#54262)

2026-09-01 11:26:00 -04:00

midashenglm.py

[Mypy] Fix typing for M models (#54262)

2026-09-01 11:26:00 -04:00

mimo_audio.py

[Mypy] Fix typing for M models (#54262)

2026-09-01 11:26:00 -04:00

mimo_mtp.py

Remove more unnecessary load_weights methods (#47058)

2026-06-30 15:22:16 +01:00

mimo_v2_mtp.py

[Mypy] Fix typing for M models (#54262)

2026-09-01 11:26:00 -04:00

mimo_v2_omni.py

[Mypy] Fix typing for M models (#54262)

2026-09-01 11:26:00 -04:00

mimo_v2.py

[Mypy] Fix typing for M models (#54262)

2026-09-01 11:26:00 -04:00

mimo.py

[Core][MRV2] Support eagle3 spec decode with pipeline parallel (#50514)

2026-09-05 11:45:34 +08:00

minicpm3.py

[Perf] RMSNorm uncontiguous support, 1.2~3.1x kernel performance improvement (#49750)

2026-07-29 13:45:38 -07:00

minicpm_eagle.py

[Mypy] Fix typing for M models (#54262)

2026-09-01 11:26:00 -04:00

minicpm.py

[Mypy] Fix typing for M models (#54262)

2026-09-01 11:26:00 -04:00

minicpmo.py

[Mypy] Fix typing for M models (#54262)

2026-09-01 11:26:00 -04:00

minicpmv4_6.py

[Mypy] Fix typing for M models (#54262)

2026-09-01 11:26:00 -04:00

minicpmv.py

[MM][CG] Enable encoder CUDA Graph for MiniCPM-V (#42785)

2026-09-10 06:28:48 +00:00

minimax_m2.py

Reduce AutoWeightsLoader kwargs (#53106)

2026-08-20 15:40:11 +00:00

mistral3.py

[Mypy] Fix mypy typing for L models (#54177)

2026-09-04 07:50:43 -04:00

mistral_eagle.py

[Mypy] Fix typing for M models (#54262)

2026-09-01 11:26:00 -04:00

mistral_large_3_eagle.py

[Mypy] Fix typing for M models (#54262)

2026-09-01 11:26:00 -04:00

mistral_large_3.py

[MISC] Cleanup deprecated parameters (#53559)

2026-08-24 10:19:04 -07:00

mistral.py

[Mypy] Fix typing for M models (#54262)

2026-09-01 11:26:00 -04:00

mixtral.py

[CPU][Zen] Route Int8 MoE inference through zentorch on AMD (#44834)

2026-09-01 09:31:31 +00:00

mllama4.py

[MM][CG] Enable encoder CUDA Graph for MiniCPM-V (#42785)

2026-09-10 06:28:48 +00:00

mlp_speculator.py

[Refactor] Remove unused dead code (#38842)

2026-04-06 11:52:05 -04:00

modernbert.py

[Model] Add FP8 quantization support for ModernBERT (#53101)

2026-08-24 03:04:19 +00:00

module_mapping.py

[Mypy] Fix typing for M models (#54262)

2026-09-01 11:26:00 -04:00

molmo2.py

[Mypy] Fix typing for M models (#54262)

2026-09-01 11:26:00 -04:00

molmo.py

[Mypy] Fix typing for M models (#54262)

2026-09-01 11:26:00 -04:00

moondream3.py

[Bugfix] Keep the Moondream3 MoE all-reduce out of the fused-path try (#54152)

2026-08-28 15:06:42 -07:00

moonvit.py

[Mypy] Fix typing for M models (#54262)

2026-09-01 11:26:00 -04:00

moss_audio.py

[Mypy] Fix typing for M models (#54262)

2026-09-01 11:26:00 -04:00

moss_transcribe_diarize.py

[Mypy] Fix typing for M models (#54262)

2026-09-01 11:26:00 -04:00

muse_glimmer.py

[Mypy] Fix typing for M models (#54262)

2026-09-01 11:26:00 -04:00

nano_nemotron_vl.py

[MM] Simplify prompt updates: replace PromptSeq with list[int] (#53372)

2026-08-23 14:50:32 +08:00

nemotron_h_mtp.py

[Feature][Spec Decode] MTP with separate (possibly quantized) lm head for nemotron (#54574)

2026-09-10 08:26:31 -04:00

nemotron_h.py

[Mamba] Add FlashInfer ReplaySSM backend (#52506)

2026-09-01 10:26:43 -04:00

nemotron_nas.py

Fix weight tying (#51665)

2026-08-20 10:31:55 +00:00

nemotron_parse.py

[MM] Simplify _apply_hf_processor_main (#53275)

2026-08-22 01:49:37 -07:00

nemotron_vl.py

Reduce AutoWeightsLoader kwargs (#53106)

2026-08-20 15:40:11 +00:00

nemotron.py

Standardise weight tying on ParallelLMHead.tie_weights (#52147)

2026-08-13 15:10:09 +00:00

nvlm_d.py

[MM] Simplify prompt updates: replace PromptSeq with list[int] (#53372)

2026-08-23 14:50:32 +08:00

olmo_hybrid.py

Fix weight tying (#51665)

2026-08-20 10:31:55 +00:00

olmoe.py

[MoE Refactor] Rename FusedMoE to FusedMoEFactory (#44941)

2026-07-31 11:28:18 +01:00

openai_privacy_filter.py

[Model] Add support for openai/privacy-filter (#41026)

2026-06-28 16:18:22 -07:00

opencua.py

[MM] Remove text components from ProcessorInputs (#53093)

2026-08-20 20:42:53 -07:00

openpangu_mtp.py

[MoE Refactor] Rename FusedMoE.make_expert_params_mapping to fused_moe_make_expert_params_mapping (#40671)

2026-04-23 11:22:34 -04:00

openpangu_vl.py

[Bugfix] Fix OpenPangu multimodal embedding merge (#55941)

2026-09-08 23:53:59 +00:00

openpangu.py

Fix weight tying (#51665)

2026-08-20 10:31:55 +00:00

openvla.py

[MM] Simplify prompt updates: replace PromptSeq with list[int] (#53372)

2026-08-23 14:50:32 +08:00

opt.py

Fix weight tying (#51665)

2026-08-20 10:31:55 +00:00

orion.py

Standardise weight tying on ParallelLMHead.tie_weights (#52147)

2026-08-13 15:10:09 +00:00

ovis2_5.py

[MM] Simplify _apply_hf_processor_main (#53275)

2026-08-22 01:49:37 -07:00

ovis.py

[MM] Simplify prompt updates: replace PromptSeq with list[int] (#53372)

2026-08-23 14:50:32 +08:00

paddleocr_vl.py

[Perf] Ensure async h2d copies are pinned in more places (#55202)

2026-09-04 17:17:01 -07:00

paligemma.py

[Mypy] Fix mypy typing for P models (#54169)

2026-09-04 20:47:11 +00:00

parakeet.py

[Platform] Fill in the missing backend parameter for torch.compile (#51781)

2026-08-19 10:31:19 +00:00

param2moe.py

[Mypy] Fix mypy typing for P models (#54169)

2026-09-04 20:47:11 +00:00

phi3.py

…

phi3v.py

[Mypy] Fix mypy typing for P models (#54169)

2026-09-04 20:47:11 +00:00

phi4mm_audio.py

[Perf] Ensure async h2d copies are pinned in more places (#55202)

2026-09-04 17:17:01 -07:00

phi4mm_utils.py

[Mypy] Fix mypy typing for P models (#54169)

2026-09-04 20:47:11 +00:00

phi4mm.py

[Mypy] Fix mypy typing for P models (#54169)

2026-09-04 20:47:11 +00:00

phi4siglip.py

[MM] Simplify prompt updates: replace PromptSeq with list[int] (#53372)

2026-08-23 14:50:32 +08:00

phi.py

Remove unnecessary load_weights methods (#44589)

2026-06-29 01:52:23 -07:00

phimoe.py

[MoE Refactor] Rename FusedMoE to FusedMoEFactory (#44941)

2026-07-31 11:28:18 +01:00

pixtral.py

[Model] Pixtral: use packed multimodal encoder attention (#52185)

2026-08-26 14:14:40 +00:00

plamo3.py

[Mypy] Fix mypy typing for P models (#54169)

2026-09-04 20:47:11 +00:00

qianfan_ocr.py

[Model] support Qianfan-OCR model (#40136)

2026-05-05 10:51:25 +00:00

qwen2_5_omni_thinker.py

[Perf][Multimodal] Avoid duplicate text embedding in Qwen2.5-Omni (#55415)

2026-09-05 06:39:32 +00:00

qwen2_5_vl.py

[MM][CG] Enable encoder CUDA Graph for MiniCPM-V (#42785)

2026-09-10 06:28:48 +00:00

qwen2_audio.py

[MM] Simplify _apply_hf_processor_main (#53275)

2026-08-22 01:49:37 -07:00

qwen2_moe.py

[Model] Support Qwen3.8-Flash-Next (#53896)

2026-08-31 05:57:56 +00:00

qwen2_rm.py

Remove unnecessary load_weights methods (#44589)

2026-06-29 01:52:23 -07:00

qwen2_vl.py

[MM][CG] Enable encoder CUDA Graph for MiniCPM-V (#42785)

2026-09-10 06:28:48 +00:00

qwen2.py

[Core][MRV2] Support eagle3 spec decode with pipeline parallel (#50514)

2026-09-05 11:45:34 +08:00

qwen3_5_mtp.py

[Spec][V2] Support MTP speculative decoding under pipeline parallelism (#46994)

2026-09-11 17:55:21 +00:00

qwen3_5.py

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

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

qwen3_asr_forced_aligner.py

Reduce AutoWeightsLoader kwargs (#53106)

2026-08-20 15:40:11 +00:00

qwen3_asr_realtime.py

[MM] Cache common token sequences (#53560)

2026-08-24 07:48:10 -07:00

qwen3_asr.py

[Rust Frontend][gRPC] Add audio and video media inputs (#53760)

2026-08-30 11:31:37 +08:00

qwen3_dflash2.py

[Spec Decode] DFlash2: local convolution + candidate selector (#52816)

2026-08-20 22:27:22 -07:00

qwen3_dflash.py

[Core][MRV2] Support eagle3 spec decode with pipeline parallel (#50514)

2026-09-05 11:45:34 +08:00

qwen3_dspark.py

[Spec Decode] Add NVFP4 DSpark gathered top-k projection (#55713)

2026-09-09 17:33:46 +00:00

qwen3_eagle3.py

[Core][MRV2] Support eagle3 spec decode with pipeline parallel (#50514)

2026-09-05 11:45:34 +08:00

qwen3_moe.py

Standardise weight tying on ParallelLMHead.tie_weights (#52147)

2026-08-13 15:10:09 +00:00

qwen3_next_mtp.py

[MOE] Standardize and abstract fused shared expert optimization selection (#51695)

2026-08-18 12:56:56 -04:00

qwen3_next.py

[ROCm] [BugFix] Fix AITER MXFP4 ASM-GEMM crash on unfused shared experts (#55213)

2026-09-10 05:23:36 +00:00

qwen3_omni_moe_thinker.py

[ROCm][Perf] Fix Qwen3-vLLM audio encoder TP when heads are not divisible by TP size (#45900)

2026-09-09 13:28:24 +00:00

qwen3_vl_moe.py

[Bugfix] Qwen3-VL(-MoE): pass architectures to with_hf_config for pipeline parallelism (#43272)

2026-09-10 14:49:35 +01:00

qwen3_vl.py

[Bugfix] Qwen3-VL(-MoE): pass architectures to with_hf_config for pipeline parallelism (#43272)

2026-09-10 14:49:35 +01:00

qwen3.py

Fix weight tying (#51665)

2026-08-20 10:31:55 +00:00

radio.py

[Mypy] Fix typing for R/S models (#54157)

2026-09-10 09:43:08 -04:00

registry.py

[Model] Support DeepSeek-V4.1-Flash (#56214)

2026-09-11 02:11:19 -07:00

rnj1.py

Fix weight tying (#51665)

2026-08-20 10:31:55 +00:00

roberta.py

[Mypy] Fix typing for R/S models (#54157)

2026-09-10 09:43:08 -04:00

rvl.py

[Misc] Reorganize inputs (#35182)

2026-03-25 10:22:54 -07:00

sarvam.py

[Mypy] Fix typing for R/S models (#54157)

2026-09-10 09:43:08 -04:00

seed_oss.py

Fix weight tying (#51665)

2026-08-20 10:31:55 +00:00

siglip2navit.py

[Mypy] Fix typing for R/S models (#54157)

2026-09-10 09:43:08 -04:00

siglip.py

[Mypy] Fix typing for R/S models (#54157)

2026-09-10 09:43:08 -04:00

skyworkr1v.py

[Mypy] Fix typing for R/S models (#54157)

2026-09-10 09:43:08 -04:00

smolvlm.py

[Bugfix] Standardize getting number of image patches/tokens (#34358)

2026-02-12 20:47:01 -08:00

solar.py

[Mypy] Fix typing for R/S models (#54157)

2026-09-10 09:43:08 -04:00

stablelm.py

[Bugfix][Model] Enable torch.compile for StableLM (#54969)

2026-09-03 03:21:54 -07:00

step1.py

Remove unnecessary load_weights methods (#44589)

2026-06-29 01:52:23 -07:00

step3_text.py

[Mypy] Fix typing for R/S models (#54157)

2026-09-10 09:43:08 -04:00

step3_vl.py

[Mypy] Fix typing for R/S models (#54157)

2026-09-10 09:43:08 -04:00

step3p5_mtp.py

[Mypy] Fix typing for R/S models (#54157)

2026-09-10 09:43:08 -04:00

step3p5.py

[Mypy] Fix typing for R/S models (#54157)

2026-09-10 09:43:08 -04:00

step3p7.py

[Mypy] Fix typing for R/S models (#54157)

2026-09-10 09:43:08 -04:00

step_vl.py

[Mypy] Fix typing for R/S models (#54157)

2026-09-10 09:43:08 -04:00

telechat2.py

[Mypy Fix] Mypy fix for "vllm/model_executor/models/[tT]" (#53466)

2026-08-26 14:12:11 +00:00

teleflm.py

…

terratorch.py

[Bugfix] Reject tokenizer-less Qwen VL processor init (#54886)

2026-09-03 21:25:35 -07:00

ultravox.py

[MM] Simplify prompt updates: replace PromptSeq with list[int] (#53372)

2026-08-23 14:50:32 +08:00

unlimited_ocr.py

[MM] Simplify _apply_hf_processor_main (#53275)

2026-08-22 01:49:37 -07:00

utils.py

[Quant][Kernel] Remove GPTQ Group/Dynamic Activation Ordering (#54809)

2026-09-08 12:45:12 -07:00

vision.py

[XPU] Use fused_input_norm kernel in FusedInputNorm (#52945)

2026-09-07 13:44:31 +08:00

voxtral_realtime.py

[MM] Remove renderer_applies_updates flag (#53385)

2026-08-23 01:25:55 +08:00

voxtral.py

[Bugfix][Models] Register sleep-managed runtime buffers (#53507)

2026-08-29 10:32:24 +08:00

voyage.py

[Model] Use AutoWeightsLoader for Voyage (#42972)

2026-05-22 05:28:23 +00:00

whisper_causal.py

[Deprecation] Deprecate items scheduled for 0.29 (#55353)

2026-09-11 08:17:02 -07:00

whisper_utils.py

[Voxtral] Add new streaming arch (#32861)

2026-01-23 12:41:52 +01:00

whisper.py

[MM] Simplify _apply_hf_processor_main (#53275)

2026-08-22 01:49:37 -07:00

zamba2.py

Remove more unnecessary load_weights methods (#47058)

2026-06-30 15:22:16 +01:00