3 people 38db529f66 [feat]: Create interface for model-specific M-RoPE (#24194) ...

Signed-off-by: AzizCode92 <azizbenothman76@gmail.com>
Signed-off-by: Aziz <azizbenothman76@gmail.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
Co-authored-by: Cyrus Leung <tlleungac@connect.ust.hk>

2025-09-18 19:18:56 +00:00

..

__init__.py

[feat]: Create interface for model-specific M-RoPE (#24194)

2025-09-18 19:18:56 +00:00

adapters.py

Enable conversion of multimodal models to pooling tasks (#24451)

2025-09-12 03:30:41 +00:00

aimv2.py

[Misc] Use config definitions from Transformers library (#21913)

2025-08-08 23:10:51 -07:00

apertus.py

Fix weights loading for Apertus (#24100)

2025-09-02 18:34:28 +00:00

arcee.py

[Doc]: fixing doc typos (#24635)

2025-09-10 23:19:28 -07:00

arctic.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

aria.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

aya_vision.py

[Bugfix] Ensure correctness of Cohere2Vision processing (#23245)

2025-08-20 11:09:18 +00:00

baichuan.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

bailing_moe.py

[Model]: support Ling2.0 (#24627)

2025-09-15 05:09:30 -07:00

bamba.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

bert_with_rope.py

[Kernel] Delegate construction of FusedMoEQuantConfig to FusedMoEMethodBase subclasses (#22537)

2025-09-17 17:43:31 -06:00

bert.py

[New Model] Support BertForTokenClassification / Named Entity Recognition (NER) task (#24872)

2025-09-18 23:22:01 +08:00

blip2.py

[Docs] Fix warnings in mkdocs build (continued) (#24740)

2025-09-12 06:43:15 -07:00

blip.py

[Misc] Add SPDX-FileCopyrightText (#19100)

2025-06-03 11:20:17 -07:00

bloom.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

chameleon.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

chatglm.py

[Models] Improve iteration over layers (#19497)

2025-08-29 09:26:34 +08:00

clip.py

[Misc] Add SPDX-FileCopyrightText (#19100)

2025-06-03 11:20:17 -07:00

cohere2_vision.py

[CI/Build] Split out mm processor tests (#23260)

2025-08-20 20:05:20 -07:00

commandr.py

[Models] Improve iteration over layers (#19497)

2025-08-29 09:26:34 +08:00

config.py

[Chore] Cleanup guided namespace, move to structured outputs config (#22772)

2025-09-18 09:20:27 +00:00

constant_size_cache.py

[Misc] Add SPDX-FileCopyrightText (#19100)

2025-06-03 11:20:17 -07:00

dbrx.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

deepseek_eagle.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

deepseek_mtp.py

[Bugfix][Wide EP] Fix redundant work when using DeepEP, TP Attn, and EP MoE (#24134)

2025-09-08 19:01:51 -07:00

deepseek_v2.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

deepseek_vl2.py

[Core] Simplify and unify mm uuid handling & auto-generated mm hash overrides processing. (#24271)

2025-09-09 21:36:09 -07:00

deepseek.py

[Kernel] Delegate construction of FusedMoEQuantConfig to FusedMoEMethodBase subclasses (#22537)

2025-09-17 17:43:31 -06:00

dots1.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

ernie45_moe.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

ernie45_vl_moe.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

ernie45_vl.py

[Doc]: fix typos in various files (#24798)

2025-09-13 00:43:33 -07:00

ernie45.py

[Ernie 4.5] Name Change for Base 0.3B Model (#21735)

2025-07-28 12:22:32 +00:00

ernie_mtp.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

exaone4.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

exaone.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

fairseq2_llama.py

[Misc] Add SPDX-FileCopyrightText (#19100)

2025-06-03 11:20:17 -07:00

falcon_h1.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

falcon.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

fuyu.py

[Refactor] Define MultiModalKwargsItems separate from MultiModalKwargs (#23053)

2025-08-18 09:52:00 +00:00

gemma2.py

[Models] Improve iteration over layers (#19497)

2025-08-29 09:26:34 +08:00

gemma3_mm.py

Retrieve sliding_window from text config in Gemma3 MM (#25085)

2025-09-18 06:29:05 +00:00

gemma3.py

[New Model]: google/embeddinggemma-300m (#24318)

2025-09-05 22:58:36 -07:00

gemma3n_mm.py

Retrieve sliding_window from text config in Gemma3 MM (#25085)

2025-09-18 06:29:05 +00:00

gemma3n.py

Revert gemma3n fast prefill changes (#23897)

2025-08-29 12:16:57 -07:00

gemma.py

[Models] Improve iteration over layers (#19497)

2025-08-29 09:26:34 +08:00

glm4_1v.py

[Bugfix] Fix GLM4.1V multimodal processor with compatability for Transformers v4.56 (#24822)

2025-09-15 20:45:06 +08:00

glm4_moe_mtp.py

[Bugfix] Fix DeepSeek MTP (#22934)

2025-08-16 01:25:06 +00:00

glm4_moe.py

[Model] Apply SharedFusedMoE to glm4_moe. (#24849)

2025-09-17 16:02:31 +08:00

glm4.py

[Misc] Add SPDX-FileCopyrightText (#19100)

2025-06-03 11:20:17 -07:00

glm4v.py

[Refactor] Define MultiModalKwargsItems separate from MultiModalKwargs (#23053)

2025-08-18 09:52:00 +00:00

glm.py

[Misc] Add SPDX-FileCopyrightText (#19100)

2025-06-03 11:20:17 -07:00

gpt2.py

[Model] Systematic support for fp32 head, pooling models part (#23810)

2025-09-09 07:29:50 -07:00

gpt_bigcode.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

gpt_j.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

gpt_neox.py

[Models] Improve iteration over layers (#19497)

2025-08-29 09:26:34 +08:00

gpt_oss.py

[fix]: remove data type hardcoding from gptoss model implementation (#23807)

2025-09-18 18:15:23 +00:00

granite_speech.py

[CI/Build] Improve Tensor Schema tests speed by avoid engine core initialization (#23357)

2025-08-31 22:52:20 -07:00

granite.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

granitemoe.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

granitemoehybrid.py

[Models] Improve iteration over layers (#19497)

2025-08-29 09:26:34 +08:00

granitemoeshared.py

[Models] Improve iteration over layers (#19497)

2025-08-29 09:26:34 +08:00

gritlm.py

[V0 Deprecation] Remove pooling model support in V0 (#23434)

2025-08-29 00:04:02 -07:00

grok1.py

[Models] Improve iteration over layers (#19497)

2025-08-29 09:26:34 +08:00

h2ovl.py

[Core] Simplify and unify mm uuid handling & auto-generated mm hash overrides processing. (#24271)

2025-09-09 21:36:09 -07:00

hunyuan_v1.py

[EPLB] Add EPLB support for hunyuan_v1 (#23078)

2025-09-18 04:51:35 +00:00

hyperclovax_vision.py

[Core] Use key-only cache for BaseMultiModalProcessor (#23018)

2025-08-27 14:19:13 +08:00

idefics2_vision_model.py

[VLM] Migrate remain DP-supported ViT models to use disable_tp (#24363)

2025-09-11 18:30:41 +00:00

idefics3.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

interfaces_base.py

[Model] Explicit default_pooling_type interface (#23736)

2025-08-27 13:24:09 +00:00

interfaces.py

[feat]: Create interface for model-specific M-RoPE (#24194)

2025-09-18 19:18:56 +00:00

intern_vit.py

[Model] enable data parallel for InternVL vision encoder (#23909)

2025-09-17 21:11:46 -07:00

internlm2_ve.py

[Models] Improve iteration over layers (#19497)

2025-08-29 09:26:34 +08:00

internlm2.py

[Model] Systematic support for fp32 head, pooling models part (#23810)

2025-09-09 07:29:50 -07:00

interns1_vit.py

Feature/vit attention unification# 23880 (#23978)

2025-09-10 06:10:14 -07:00

interns1.py

[Doc]: fix typos in various files (#24821)

2025-09-15 01:08:52 -07:00

internvl.py

[Model] enable data parallel for InternVL vision encoder (#23909)

2025-09-17 21:11:46 -07:00

jais.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

jamba.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

jina_vl.py

[Model] Systematic support for fp32 head, pooling models part (#23810)

2025-09-09 07:29:50 -07:00

keye_vl1_5.py

[Docs] Fix warnings in mkdocs build (continued) (#24791)

2025-09-13 00:13:44 -07:00

keye.py

[Docs] Fix warnings in mkdocs build (continued) (#24791)

2025-09-13 00:13:44 -07:00

kimi_vl.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

lfm2.py

[Models] Improve iteration over layers (#19497)

2025-08-29 09:26:34 +08:00

llama4_eagle.py

[Meta] Official Eagle mm support, first enablement on llama4 (#20788)

2025-07-31 10:35:07 -07:00

llama4.py

[Kernels] Overlap shared experts with send/recv (#23273)

2025-09-03 12:35:18 -04:00

llama_eagle3.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

llama_eagle.py

[Meta] Official Eagle mm support, first enablement on llama4 (#20788)

2025-07-31 10:35:07 -07:00

llama.py

[Models][Quantization] Add quantization configuration update in Voxtral model (#24122)

2025-09-10 19:13:56 -07:00

llava_next_video.py

[Refactor] Define MultiModalKwargsItems separate from MultiModalKwargs (#23053)

2025-08-18 09:52:00 +00:00

llava_next.py

[Docs] Fix warnings in mkdocs build (continued) (#24791)

2025-09-13 00:13:44 -07:00

llava_onevision.py

[Doc]: fixing doc typos (#24635)

2025-09-10 23:19:28 -07:00

llava.py

[Docs] Fix warnings in mkdocs build (continued) (#24791)

2025-09-13 00:13:44 -07:00

mamba2.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

mamba_cache.py

[V1] [Hybrid] Support using float32 for state in Hybrid Models (Mamba2, Mamba1, Minimax) (#22928)

2025-08-15 12:57:06 +00:00

mamba.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

medusa.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

midashenglm.py

[Bugfix] MiDashengLM model contact error under concurrent testing (#24738)

2025-09-15 06:38:12 +00:00

mimo_mtp.py

[spec decode] Fix MTP inference path for MiMo-7B model (#25136)

2025-09-18 09:12:19 -07:00

mimo.py

[Models] Improve iteration over layers (#19497)

2025-08-29 09:26:34 +08:00

minicpm3.py

[Misc] Add SPDX-FileCopyrightText (#19100)

2025-06-03 11:20:17 -07:00

minicpm_eagle.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

minicpm.py

[Kernel] Delegate construction of FusedMoEQuantConfig to FusedMoEMethodBase subclasses (#22537)

2025-09-17 17:43:31 -06:00

minicpmo.py

Migrate MiniCPMOAudioInputs to TensorSchema (#21847)

2025-08-22 16:43:29 +08:00

minicpmv.py

[Doc]: fix typos in various files (#24726)

2025-09-12 06:43:12 -07:00

minimax_cache.py

[Misc] Add SPDX-FileCopyrightText (#19100)

2025-06-03 11:20:17 -07:00

minimax_text_01.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

minimax_vl_01.py

[Bugfix] Fix broken Minimax-01-VL model (#22116)

2025-08-19 08:49:29 +00:00

mistral3.py

[Doc] Fix cross-reference warnings (#25058)

2025-09-18 02:05:16 -07:00

mixtral.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

mllama4.py

[Docs] Fix warnings in mkdocs build (continued) (#24791)

2025-09-13 00:13:44 -07:00

mlp_speculator.py

[Misc] Add SPDX-FileCopyrightText (#19100)

2025-06-03 11:20:17 -07:00

modernbert.py

[Model] Systematic support for fp32 head, pooling models part (#23810)

2025-09-09 07:29:50 -07:00

module_mapping.py

[Misc] Add SPDX-FileCopyrightText (#19100)

2025-06-03 11:20:17 -07:00

molmo.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

moonvit.py

[Docs] Fix warnings in mkdocs build (continued) (#24791)

2025-09-13 00:13:44 -07:00

motif.py

[Model] New model support for Motif-1-Tiny (#23414)

2025-09-10 23:29:40 -07:00

mpt.py

[Models] Improve iteration over layers (#19497)

2025-08-29 09:26:34 +08:00

nano_nemotron_vl.py

Add RADIO Vision Encoder Support to vLLM (#24595)

2025-09-17 05:53:30 -07:00

nemotron_h.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

nemotron_nas.py

[Models] Improve iteration over layers (#19497)

2025-08-29 09:26:34 +08:00

nemotron_vl.py

[Doc]: fix typos in various files (#24798)

2025-09-13 00:43:33 -07:00

nemotron.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

nvlm_d.py

[Refactor] Define MultiModalKwargsItems separate from MultiModalKwargs (#23053)

2025-08-18 09:52:00 +00:00

olmo2.py

[Model] Add Olmo3 model implementation (#24534)

2025-09-13 03:26:21 +00:00

olmo.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

olmoe.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

opt.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

orion.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

ovis2_5.py

[CI/Build] Improve Tensor Schema tests speed by avoid engine core initialization (#23357)

2025-08-31 22:52:20 -07:00

ovis.py

Migrate OvisImagePatchInputs to TensorSchema (#22024)

2025-09-02 12:01:36 +08:00

paligemma.py

[Core] Simplify and unify mm uuid handling & auto-generated mm hash overrides processing. (#24271)

2025-09-09 21:36:09 -07:00

persimmon.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

phi3.py

[Misc] Add SPDX-FileCopyrightText (#19100)

2025-06-03 11:20:17 -07:00

phi3v.py

[Core] Use key-only cache for BaseMultiModalProcessor (#23018)

2025-08-27 14:19:13 +08:00

phi4_multimodal.py

[Docs] Fix warnings in mkdocs build (continued) (#24791)

2025-09-13 00:13:44 -07:00

phi4flash.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

phi4mm_audio.py

[Doc]: fixing doc typos (#24635)

2025-09-10 23:19:28 -07:00

phi4mm_utils.py

[Docs] Fix warnings in mkdocs build (continued) (#24092)

2025-09-10 06:23:28 -07:00

phi4mm.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

phi.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

phimoe.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

pixtral.py

[Core] Simplify and unify mm uuid handling & auto-generated mm hash overrides processing. (#24271)

2025-09-09 21:36:09 -07:00

plamo2.py

[Model] Clean up and simplify Mamba2 Metadata Usage in both V0 and V1 (#24331)

2025-09-16 14:53:43 +00:00

qwen2_5_omni_thinker.py

[Models] Optimise and simplify _validate_and_reshape_mm_tensor (#24742)

2025-09-12 15:37:37 +00:00

qwen2_5_vl.py

[Multi Modal][Performance] Fused Q,K's apply_rope into one (#24511)

2025-09-14 08:10:21 +00:00

qwen2_audio.py

[Models] Optimise and simplify _validate_and_reshape_mm_tensor (#24742)

2025-09-12 15:37:37 +00:00

qwen2_moe.py

[Bugfix][Qwen3-Next] add prefixes to shared_expert in qwen3-next and mlp in qwen2moe to successfully load ignored params in quantized models (#24960)

2025-09-18 12:08:50 +08:00

qwen2_rm.py

[Model] Systematic support for fp32 head, pooling models part (#23810)

2025-09-09 07:29:50 -07:00

qwen2_vl.py

[feat]: Create interface for model-specific M-RoPE (#24194)

2025-09-18 19:18:56 +00:00

qwen2.py

[Model] Support Qwen3-VL Model Series (#24727)

2025-09-17 05:01:04 +00:00

qwen3_moe.py

[Kernel] Delegate construction of FusedMoEQuantConfig to FusedMoEMethodBase subclasses (#22537)

2025-09-17 17:43:31 -06:00

qwen3_next_mtp.py

[Qwen] Add fp8 checkpoint support for qwen3-next. (#25079)

2025-09-18 08:16:04 +00:00

qwen3_next.py

[PERF] Add conv1d metadata to GDN attn (#25105)

2025-09-18 14:27:49 +00:00

qwen3_vl_moe.py

[MM Encoder] Apply DP ViT for Qwen3-VL model series (#24955)

2025-09-17 21:04:21 -07:00

qwen3_vl.py

[MM Encoder] Apply DP ViT for Qwen3-VL model series (#24955)

2025-09-17 21:04:21 -07:00

qwen3.py

[mypy] Fix incorrect type hint for EAGLE3 support (#23617)

2025-08-25 23:50:17 -07:00

qwen_vl.py

Migrate Qwen inputs to TensorSchema (#23473)

2025-08-28 10:43:26 +08:00

qwen.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

radio.py

Add RADIO Vision Encoder Support to vLLM (#24595)

2025-09-17 05:53:30 -07:00

registry.py

[New Model] Support BertForTokenClassification / Named Entity Recognition (NER) task (#24872)

2025-09-18 23:22:01 +08:00

roberta.py

[Model] Systematic support for fp32 head, pooling models part (#23810)

2025-09-09 07:29:50 -07:00

rvl.py

[Model][VLM] Support R-4B Model (#23246)

2025-08-21 04:08:52 +00:00

seed_oss.py

[Models] Improve iteration over layers (#19497)

2025-08-29 09:26:34 +08:00

siglip2navit.py

[Docs] Fix warnings in mkdocs build (continued) (#24791)

2025-09-13 00:13:44 -07:00

siglip.py

[Misc] Add SPDX-FileCopyrightText (#19100)

2025-06-03 11:20:17 -07:00

skyworkr1v.py

Migrate skyworkr1v inputs to TensorSchema (#23499)

2025-08-25 04:43:21 +00:00

smolvlm.py

[Misc] Automatically resolve HF processor init kwargs (#22005)

2025-07-31 22:44:10 -07:00

solar.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

stablelm.py

[Models] Improve iteration over layers (#19497)

2025-08-29 09:26:34 +08:00

starcoder2.py

[Models] Improve iteration over layers (#19497)

2025-08-29 09:26:34 +08:00

step3_text.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00

step3_vl.py

Feature/vit attention unification# 23880 (#23978)

2025-09-10 06:10:14 -07:00

swin.py

[New Model]Donut model (#23229)

2025-08-24 12:52:24 +00:00

tarsier.py

[Core] Use key-only cache for BaseMultiModalProcessor (#23018)

2025-08-27 14:19:13 +08:00

telechat2.py

[CI/Build] Fix registry tests (#21934)

2025-07-30 09:10:41 -07:00

teleflm.py

[Misc] Add SPDX-FileCopyrightText (#19100)

2025-06-03 11:20:17 -07:00

terratorch.py

[Core] Simplify and unify mm uuid handling & auto-generated mm hash overrides processing. (#24271)

2025-09-09 21:36:09 -07:00

transformers.py

Move MultiModalConfig from config/__init__.py to config/multimodal.py (#24659)

2025-09-15 17:43:16 +00:00

ultravox.py

[Doc]: fix typos in various files (#24821)

2025-09-15 01:08:52 -07:00

utils.py

[Bugfix] Fix hidden_size for multimodal classification model (#24501)

2025-09-09 10:37:25 -07:00

vision.py

[Multi Modal] Add FA3 in VIT (#24347)

2025-09-12 21:27:24 +08:00

voxtral.py

Invert pattern order to make sure that out_proj layers are identified (#24781)

2025-09-12 14:45:29 -07:00

whisper.py

[v1] Add Whisper model support (encoder-decoder) (#21088)

2025-09-10 13:53:35 -07:00

zamba2.py

[Model] Pass param prefix to LLMHead (#24862)

2025-09-17 16:01:27 +08:00