Cyrus Leung ecfa7bb373 [MM] Cache common token sequences (#53560) ...

Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
Signed-off-by: Cyrus Leung <tlleungac@connect.ust.hk>

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

..

transformers

[MM] Cache common token sequences (#53560)

2026-08-24 07:48:10 -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

Fix weight tying (#51665)

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

arcee.py

Reduce AutoWeightsLoader kwargs (#53106)

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

arctic.py

Fix weight tying (#51665)

2026-08-20 10:31:55 +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] [Quantization] Add Ling hybrid MXFP4 routed experts support (#52114)

2026-08-13 21:46:53 +08:00

bailing_moe_v3.py

[Model] [Quantization] Add Ling hybrid MXFP4 routed experts support (#52114)

2026-08-13 21:46:53 +08:00

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

[MM] Simplify _apply_hf_processor_main (#53275)

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

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

chameleon.py

[MM] Simplify _apply_hf_processor_main (#53275)

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

chatglm.py

Remove unnecessary load_weights methods (#44589)

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

cheers.py

[MM] Simplify _apply_hf_processor_main (#53275)

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

clip.py

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

2026-08-20 20:42:53 -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

[MM] Simplify _apply_hf_processor_main (#53275)

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

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

Reduce AutoWeightsLoader kwargs (#53106)

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

colqwen3.py

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

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

commandr.py

Reduce AutoWeightsLoader kwargs (#53106)

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

config.py

[DCP] Default query replication for GLM sparse attention (#50382)

2026-08-21 09:07:05 -07:00

conformer_encoder.py

[model] support FireRedLID (#39290)

2026-04-10 08:43:58 +00:00

cosmos3_edge.py

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

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

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

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

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

deepseek_eagle3.py

Reduce AutoWeightsLoader kwargs (#53106)

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

deepseek_eagle.py

Reduce AutoWeightsLoader kwargs (#53106)

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

deepseek_mtp.py

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

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

deepseek_ocr2.py

[MM] Simplify _apply_hf_processor_main (#53275)

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

deepseek_ocr.py

[MM] Simplify _apply_hf_processor_main (#53275)

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

deepseek_v2.py

[6/N][KV-Cache Layout Refactor] Standardize KV cache layout (#51718)

2026-08-21 17:47:49 -07:00

deepseek_vl2.py

[MM] Simplify _apply_hf_processor_main (#53275)

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

diffusion_gemma.py

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

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

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

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

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

fairseq2_llama.py

Fix weight tying (#51665)

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

falcon_h1.py

Fix weight tying (#51665)

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

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

fireredlid.py

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

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

flex_olmo.py

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

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

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

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

gemma3.py

Fix weight tying (#51665)

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

gemma3n_audio_utils.py

…

gemma3n_mm.py

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

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

gemma3n.py

Reduce AutoWeightsLoader kwargs (#53106)

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

gemma4_dspark.py

Fix Gemma 4 for upcoming Transformers version (#49797)

2026-08-10 07:48:26 +00:00

gemma4_mm.py

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

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

gemma4_mtp.py

Standardise weight tying on ParallelLMHead.tie_weights (#52147)

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

gemma4_unified.py

[Bugfix][Model] Gemma3n/Gemma4: pad variable-length audio batches (#50958)

2026-08-05 12:46:00 +08:00

gemma4.py

Reduce AutoWeightsLoader kwargs (#53106)

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

gemma.py

[Bugfix] Load untied Gemma LM head weights (#53170)

2026-08-21 00:47:08 +00:00

glm4_1v.py

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

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

glm4_moe_lite_mtp.py

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

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

glm4_moe_lite.py

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

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

glm4_moe_mtp.py

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

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

glm4_moe.py

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

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

glm4.py

Reduce AutoWeightsLoader kwargs (#53106)

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

glm4v.py

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

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

glm_ocr_mtp.py

Remove even more unnecessary load_weights methods (#48496)

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

glm_ocr.py

[Refactor] Remove multiple dead codes (#50285)

2026-08-03 10:59:08 -04:00

glm.py

…

glmasr_utils.py

…

glmasr.py

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

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

gpt2.py

Reduce AutoWeightsLoader kwargs (#53106)

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

gpt_j.py

Reduce AutoWeightsLoader kwargs (#53106)

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

gpt_neox.py

Reduce AutoWeightsLoader kwargs (#53106)

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

gpt_oss.py

Add routed expert loading for gpt-oss (#52209)

2026-08-23 17:59:08 +08:00

granite4_vision.py

[MM] Keep more metadata tensors on CPU (#52827)

2026-08-18 19:56:33 -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

[MM] Simplify _apply_hf_processor_main (#53275)

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

granite.py

Fix weight tying (#51665)

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

granitemoe.py

Fix weight tying (#51665)

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

granitemoehybrid.py

Standardise weight tying on ParallelLMHead.tie_weights (#52147)

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

granitemoeshared.py

Fix weight tying (#51665)

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

gritlm.py

[Perf] Remove redundant device copies for CPU-only pooling token IDs, 48.9% E2E throughput improvement (#38139)

2026-03-29 18:12:50 +00:00

h2ovl.py

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

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

hrm_text.py

Fix weight tying (#51665)

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

hunyuan_v1.py

Revert "Remove native Hunyuan V1 and VL implementations" (#53296)

2026-08-21 13:39:58 -07:00

hunyuan_vision.py

[MM] Simplify _apply_hf_processor_main (#53275)

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

hy_v3_mtp.py

Remove even more unnecessary load_weights methods (#48496)

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

hy_v3.py

[Bugfix] Fix HYV3 shared_mlp prefix for compressed-tensors ignore matching (#48682)

2026-08-21 09:07:01 -07:00

hyperclovax_vision_v2.py

[MM] Simplify _apply_hf_processor_main (#53275)

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

hyperclovax_vision.py

[MM] Simplify _apply_hf_processor_main (#53275)

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

hyperclovax.py

Fix weight tying (#51665)

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

idefics2_vision_model.py

Reduce AutoWeightsLoader kwargs (#53106)

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

idefics3.py

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

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

interfaces_base.py

…

interfaces.py

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

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

intern_vit.py

Remove even more unnecessary load_weights methods (#48496)

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

internlm2.py

Fix weight tying (#51665)

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

interns1_pro.py

Reduce AutoWeightsLoader kwargs (#53106)

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

interns1_vit.py

Remove even more unnecessary load_weights methods (#48496)

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

interns1.py

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

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

interns2_mobius.py

[Bugfix][Refactor] Keep Qwen3Next layer boundaries sequence parallel (#50685)

2026-08-14 11:58:32 +08:00

interns2_preview.py

Reduce AutoWeightsLoader kwargs (#53106)

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

internvl.py

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

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

iquest_loopcoder.py

Fix weight tying (#51665)

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

isaac.py

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

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

jais2.py

Fix weight tying (#51665)

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

jamba.py

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

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

jina_vl.py

[MM] Simplify _apply_hf_processor_main (#53275)

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

jina.py

[Pooling] Improve BGE-M3 sync pooling throughput by up to 3.13% (#53464)

2026-08-24 08:33:32 -04:00

kanana_v.py

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

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

keye_vl1_5.py

[MM] Simplify _apply_hf_processor_main (#53275)

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

keye.py

[MM] Simplify _apply_hf_processor_main (#53275)

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

kimi_audio.py

[MM] Simplify _apply_hf_processor_main (#53275)

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

kimi_k25_vit.py

[Kernel][Kimi] fused vision q/k roper kernel (#50400)

2026-08-20 09:15:47 +00:00

kimi_k25.py

[MM] Simplify _apply_hf_processor_main (#53275)

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

kimi_vl.py

[MM] Simplify _apply_hf_processor_main (#53275)

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

laguna_dflash.py

Add Laguna XS.2.1 DFlash drafter support (#46853)

2026-07-02 18:09:27 -07:00

laguna.py

Fix weight tying (#51665)

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

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

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

lfm2.py

Fix weight tying (#51665)

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

lightonocr.py

[Bugfix][Model] Mistral3: fix image placeholder grid for processor size overrides (#52874)

2026-08-22 21:34:38 -07:00

llama4_eagle.py

Reduce AutoWeightsLoader kwargs (#53106)

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

llama4.py

Fix weight tying (#51665)

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

llama_eagle3.py

Reduce AutoWeightsLoader kwargs (#53106)

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

llama_eagle.py

Reduce AutoWeightsLoader kwargs (#53106)

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

llama.py

Reduce AutoWeightsLoader kwargs (#53106)

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

llava_next_video.py

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

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

llava_next.py

[MM] Simplify _apply_hf_processor_main (#53275)

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

llava_onevision2.py

[MM] Cache common token sequences (#53560)

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

llava_onevision.py

[MM] Simplify _apply_hf_processor_main (#53275)

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

llava.py

[MM] Simplify _apply_hf_processor_main (#53275)

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

longcat_flash_mtp.py

[Model] Add LongCat-Flash-Lite (n-gram embedding) (#47857)

2026-07-10 07:17:50 -07:00

longcat_flash_ngram.py

Reduce AutoWeightsLoader kwargs (#53106)

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

longcat_flash.py

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

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

mamba2.py

Remove unnecessary load_weights methods (#44589)

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

mamba.py

Standardise weight tying on ParallelLMHead.tie_weights (#52147)

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

medusa.py

Add Medusa speculative decoding e2e test (#41396)

2026-06-30 18:02:22 +00:00

mellum.py

Standardise weight tying on ParallelLMHead.tie_weights (#52147)

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

midashenglm.py

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

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

mimo_audio.py

[Model] Add MiMo-V2.5 support (#40967)

2026-04-27 13:26:51 +00:00

mimo_mtp.py

Remove more unnecessary load_weights methods (#47058)

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

mimo_v2_mtp.py

[SpecDecoding] extend mtp support for mimo 2.5 (#41905)

2026-05-09 18:22:59 +00:00

mimo_v2_omni.py

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

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

mimo_v2.py

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

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

mimo.py

Reduce AutoWeightsLoader kwargs (#53106)

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

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

Fix weight tying (#51665)

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

minicpm.py

Fix weight tying (#51665)

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

minicpmo.py

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

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

minicpmv4_6.py

[MM] Cache common token sequences (#53560)

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

minicpmv.py

[MM] Cache common token sequences (#53560)

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

minimax_m2.py

Reduce AutoWeightsLoader kwargs (#53106)

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

mistral3.py

[Bugfix][Model] Mistral3: fix image placeholder grid for processor size overrides (#52874)

2026-08-22 21:34:38 -07:00

mistral_eagle.py

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

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

mistral_large_3_eagle.py

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

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

mistral_large_3.py

[Model] Migrate MistralLarge3ForCausalLM to AutoWeightsLoader (#48153)

2026-07-10 12:27:58 +00:00

mistral.py

Fix weight tying (#51665)

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

mixtral.py

Standardise weight tying on ParallelLMHead.tie_weights (#52147)

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

mllama4.py

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

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

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

…

molmo2.py

[MM] Simplify _apply_hf_processor_main (#53275)

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

molmo.py

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

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

moondream3.py

[MM] Cache common token sequences (#53560)

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

moonvit.py

[Core] Check for GPU<->CPU syncs during CI (#43107)

2026-08-14 17:42:05 -07:00

moss_audio.py

[MM] Cache common token sequences (#53560)

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

moss_transcribe_diarize.py

[MM] Cache common token sequences (#53560)

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

mpt.py

Fix weight tying (#51665)

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

muse_glimmer.py

[Bugfix][LoRA] Add multimodal module mapping for Muse-Glimmer (#53513)

2026-08-24 05:51:43 +00:00

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

Add MTP support for Nemotron VL models (#53121)

2026-08-24 00:59:56 +00:00

nemotron_h.py

Reduce AutoWeightsLoader kwargs (#53106)

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

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

olmo3.py

Fix weight tying (#51665)

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

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

[Quant] Consolidate GPTQ: rename gptq_marlin.py to auto_gptq.py (#38288)

2026-05-15 08:25:52 +08:00

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

[MM] Simplify _apply_hf_processor_main (#53275)

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

paligemma.py

[MM] Cache common token sequences (#53560)

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

parakeet.py

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

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

param2moe.py

Standardise weight tying on ParallelLMHead.tie_weights (#52147)

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

phi3.py

…

phi3v.py

[MM] Cache common token sequences (#53560)

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

phi4mm_audio.py

[Perf][4/n] Eliminate various GPU<->CPU syncs (#42347)

2026-05-19 10:35:54 -04:00

phi4mm_utils.py

[Refactor] Remove multiple dead codes (#50285)

2026-08-03 10:59:08 -04:00

phi4mm.py

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

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

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

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

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

plamo3.py

Fix weight tying (#51665)

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

qianfan_ocr.py

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

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

qwen2_5_omni_thinker.py

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

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

qwen2_5_vl.py

[MM] Simplify _apply_hf_processor_main (#53275)

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

qwen2_audio.py

[MM] Simplify _apply_hf_processor_main (#53275)

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

qwen2_moe.py

Standardise weight tying on ParallelLMHead.tie_weights (#52147)

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

qwen2_rm.py

Remove unnecessary load_weights methods (#44589)

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

qwen2_vl.py

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

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

qwen2.py

Fix weight tying (#51665)

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

qwen3_5_mtp.py

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

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

qwen3_5.py

Reduce AutoWeightsLoader kwargs (#53106)

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

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

[MM] Cache common token sequences (#53560)

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

qwen3_dflash2.py

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

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

qwen3_dflash.py

[Model] Add Qwen3-Omni DSpark support (#52560)

2026-08-22 15:35:22 -07:00

qwen3_dspark.py

[Model] Add Qwen3-Omni DSpark support (#52560)

2026-08-22 15:35:22 -07:00

qwen3_eagle3.py

Reduce AutoWeightsLoader kwargs (#53106)

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

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

Reduce AutoWeightsLoader kwargs (#53106)

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

qwen3_omni_moe_thinker.py

[LoRA] Add Qwen3-Omni multimodal LoRA support (#52786)

2026-08-24 05:25:23 +00:00

qwen3_vl_moe.py

Standardise weight tying on ParallelLMHead.tie_weights (#52147)

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

qwen3_vl.py

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

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

qwen3.py

Fix weight tying (#51665)

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

radio.py

[Refactor] Remove dead code in multiple files (#49745)

2026-07-27 15:58:26 -04:00

registry.py

[Model] Add Qwen3-Omni DSpark support (#52560)

2026-08-22 15:35:22 -07:00

rnj1.py

Fix weight tying (#51665)

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

roberta.py

Reduce AutoWeightsLoader kwargs (#53106)

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

rvl.py

[Misc] Reorganize inputs (#35182)

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

sarvam.py

Fix weight tying (#51665)

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

seed_oss.py

Fix weight tying (#51665)

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

siglip2navit.py

Remove more unnecessary load_weights methods (#47058)

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

siglip.py

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

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

skyworkr1v.py

Reduce AutoWeightsLoader kwargs (#53106)

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

smolvlm.py

…

solar.py

Standardise weight tying on ParallelLMHead.tie_weights (#52147)

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

stablelm.py

Standardise weight tying on ParallelLMHead.tie_weights (#52147)

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

step1.py

Remove unnecessary load_weights methods (#44589)

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

step3_text.py

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

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

step3_vl.py

[MM] Keep more metadata tensors on CPU (#52827)

2026-08-18 19:56:33 -07:00

step3p5_mtp.py

fix(step3p5-mtp): honor exclude_modules for the MTP head via prefix (#48883)

2026-07-29 16:58:29 +00:00

step3p5.py

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

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

step3p7.py

[Model]Support Step-3.7-Flash (#43859)

2026-05-28 17:01:48 -07:00

step_vl.py

[MM][CG] Enable encoder Cudagraph for Step3VL (#42224)

2026-05-17 20:19:13 -07:00

telechat2.py

Fix weight tying (#51665)

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

teleflm.py

…

terratorch.py

[Core] Check for GPU<->CPU syncs during CI (#43107)

2026-08-14 17:42:05 -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

Reduce AutoWeightsLoader kwargs (#53106)

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

vision.py

replace batch_norm to numerically identical without cudnn (#51734)

2026-08-11 12:43:20 +08:00

voxtral_realtime.py

[MM] Remove renderer_applies_updates flag (#53385)

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

voxtral.py

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

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

voyage.py

[Model] Use AutoWeightsLoader for Voyage (#42972)

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

whisper_causal.py

[6/N][KV-Cache Layout Refactor] Standardize KV cache layout (#51718)

2026-08-21 17:47:49 -07:00

whisper_utils.py

…

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