3 people ca97f7b9bb Fix Gemma4 MoE expert weight remapping (#41206) ...

Signed-off-by: sunghoon.baek <sunghoon.baek@connectfy.cloud>
Co-authored-by: sunghoon.baek <sunghoon.baek@connectfy.cloud>
Co-authored-by: OpenAI Codex <codex@openai.com>

2026-04-30 00:12:42 -07:00

..

transformers

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

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

__init__.py

[Chore] Remove SupportsV0Only interface and update supported models docs (#26783)

2025-10-14 04:55:10 -07:00

adapters.py

[Bugfix][Pooling] Fix silent weight corruption with buffer-reusing iterators (#39650)

2026-04-13 19:37:27 +00:00

afmoe.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

aimv2.py

[MM] Align the prefix of MMEncoderAttention with Attention (#33750)

2026-02-04 04:07:30 +00:00

apertus.py

[Refactor] Consolidate SupportsEagle (#36063)

2026-03-13 23:22:40 +00:00

arcee.py

[Refactor] Consolidate SupportsEagle (#36063)

2026-03-13 23:22:40 +00:00

arctic.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

aria.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

audioflamingo3.py

[Misc] Reorganize inputs (#35182)

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

AXK1.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

aya_vision.py

[Misc] Reorganize inputs (#35182)

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

bagel.py

[Misc] Reorganize inputs (#35182)

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

baichuan.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

bailing_moe_linear.py

[Bugfix] BailingMoeV2.5: rotate full qk_rope_head_dim in MLA RoPE (#41185)

2026-04-29 18:08:55 +08:00

bailing_moe.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

bamba.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

bee.py

[Misc] Reorganize inputs (#35182)

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

bert_with_rope.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

bert.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

blip2.py

[Misc] Reorganize inputs (#35182)

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

blip.py

[MM] Align the prefix of MMEncoderAttention with Attention (#33750)

2026-02-04 04:07:30 +00:00

bloom.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

chameleon.py

[Misc] Reorganize inputs (#35182)

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

chatglm.py

[Misc] Cleanup more configs and processors (#37560)

2026-03-19 15:45:23 +00:00

cheers.py

[Model] Add support for Cheers multimodal model (#38788)

2026-04-02 21:01:40 +08:00

clip.py

[Misc] Reorganize inputs (#35182)

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

cohere2_vision.py

[Models] Cohere MoE (#40817)

2026-04-29 15:54:54 +00:00

cohere_asr.py

Fix Cohere ASR after HF upgrade (#40582)

2026-04-29 23:39:04 -07:00

cohere_moe.py

[Models] Cohere MoE (#40817)

2026-04-29 15:54:54 +00:00

colbert.py

[Model] Add LFM2-ColBERT-350M support (#37528)

2026-03-20 14:57:57 +00:00

colmodernvbert.py

[Model] Update ColModernVBERT to support latest HF checkpoint (#39307)

2026-04-09 10:48:51 +08:00

colpali.py

[Model] Add ColPali late interaction model for multi-modal retrieval (#36818)

2026-03-13 02:18:57 +00:00

colqwen3_5.py

[Model] Add ColQwen3.5 4.5B support (#36887)

2026-03-17 21:17:02 +00:00

colqwen3.py

[Model] Consolidate score logic by introduce score_type (#36479)

2026-03-10 13:32:25 +00:00

commandr.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

config.py

[Examples] Resettle features examples. (#40995)

2026-04-28 00:33:41 -07:00

conformer_encoder.py

[model] support FireRedLID (#39290)

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

dbrx.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

deepencoder2.py

[PluggableLayer][MM] Add PluggableLayer for CustomQwen2Decoder (#37293)

2026-03-20 06:24:07 +00:00

deepencoder.py

[PluggableLayer][MM] Add PluggableLayer for RelPosAttention (#33753)

2026-03-03 19:41:27 -08:00

deepseek_eagle3.py

Kimi k2.5 MLA based eagle3 (#36361)

2026-03-11 11:36:11 -04:00

deepseek_eagle.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

deepseek_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

deepseek_ocr2.py

[Misc] Reorganize inputs (#35182)

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

deepseek_ocr.py

[Misc] Reorganize inputs (#35182)

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

deepseek_v2.py

[Feature]: IndexCache support for DSA models (#37735)

2026-04-29 15:15:35 -04:00

deepseek_v4_mtp.py

[DSV4] Align aux stream API with DeepseekV4DecoderLayer (#41171)

2026-04-28 17:22:03 -07:00

deepseek_v4.py

[Bugfix] Fix repeated DSv4 RoPE cache initialization (#41148)

2026-04-29 20:29:55 +08:00

deepseek_vl2.py

[Misc] Reorganize inputs (#35182)

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

dots1.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

dots_ocr.py

[Misc] Reorganize inputs (#35182)

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

eagle2_5_vl.py

[2/3] Refactor InternVL-based processors (#37324)

2026-03-18 22:22:19 +08:00

ernie45_moe.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

ernie45_vl_moe.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

ernie45_vl.py

[Model] Use mm_features for Ernie-4.5 VL M-RoPE (#39753)

2026-04-14 01:11:52 -07:00

ernie45.py

[Model] Add PaddleOCR-VL Model Support (#27758)

2025-11-03 19:04:22 +08:00

ernie_mtp.py

[Chore] Update type annotation of input_ids in model forward (#33063)

2026-01-26 06:02:10 -08:00

ernie.py

[Model] Add support for BERT-like Chinese ERNIE pooling models (#36385)

2026-03-13 03:23:53 +00:00

exaone4_5_mtp.py

[Bugfix] add SupportsMultiModal to Exaone4_5_MTP (#39526)

2026-04-10 22:57:26 -07:00

exaone4_5.py

Add EXAONE-4.5 (#39388)

2026-04-09 20:53:26 -07:00

exaone4.py

Add EXAONE-4.5 (#39388)

2026-04-09 20:53:26 -07:00

exaone_moe_mtp.py

Add EXAONE-4.5 (#39388)

2026-04-09 20:53:26 -07:00

exaone_moe.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

exaone.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

extract_hidden_states.py

[Attention][V0 Deprecation] Deprecate accept output buffer (#39125)

2026-04-07 17:14:58 -04:00

fairseq2_llama.py

Convert formatting to use ruff instead of yapf + isort (#26247)

2025-10-05 07:06:22 -07:00

falcon_h1.py

[Model] Use AutoWeightsLoader for FalconH1 (#39092)

2026-04-07 16:25:17 +08:00

falcon.py

[Misc] Cleanup more configs and processors (#37560)

2026-03-19 15:45:23 +00:00

fireredasr2.py

[Audio] Bundle get_generation_prompt() params into SpeechToTextParams (#36268)

2026-04-22 12:24:18 +08:00

fireredlid.py

[model] support FireRedLID (#39290)

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

flex_olmo.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

funasr.py

support hotwords for FunASR model (#39674)

2026-04-22 02:25:06 -07:00

funaudiochat.py

[Misc] Reorganize inputs (#35182)

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

fuyu.py

[Misc] Reorganize inputs (#35182)

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

gemma2.py

Make Gemma and Gemma 2 accept inputs_embeds like Gemma 3 (#36787)

2026-03-11 18:12:43 +00:00

gemma3_mm.py

[Misc] Reorganize inputs (#35182)

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

gemma3.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

gemma3n_audio_utils.py

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

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

gemma3n_mm.py

[Audio] Bundle get_generation_prompt() params into SpeechToTextParams (#36268)

2026-04-22 12:24:18 +08:00

gemma3n.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

gemma4_mm.py

[Model] Gemma4: add bidirectional vision attention for sliding layers with window guard (#40534)

2026-04-24 08:27:46 +00:00

gemma4.py

Fix Gemma4 MoE expert weight remapping (#41206)

2026-04-30 00:12:42 -07:00

gemma.py

Make Gemma and Gemma 2 accept inputs_embeds like Gemma 3 (#36787)

2026-03-11 18:12:43 +00:00

glm4_1v.py

[bug-fix] GLM OCR Patch Merger context_dim (#37962)

2026-03-26 05:11:21 -07:00

glm4_moe_lite_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

glm4_moe_lite.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

glm4_moe_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

glm4_moe.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

glm4.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

glm4v.py

[Misc] Reorganize inputs (#35182)

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

glm_ocr_mtp.py

[GLM-OCR] GLM-OCR with MTP Support (#33005)

2026-01-26 06:24:43 -08:00

glm_ocr.py

[bug-fix] GLM OCR Patch Merger context_dim (#37962)

2026-03-26 05:11:21 -07:00

glm.py

Access partial_rotary_factor from rope_parameters (#29966)

2025-12-04 18:42:49 +00:00

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

[Audio] Bundle get_generation_prompt() params into SpeechToTextParams (#36268)

2026-04-22 12:24:18 +08:00

gpt2.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

gpt_bigcode.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

gpt_j.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

gpt_neox.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

gpt_oss.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

granite4_vision.py

[Bugfix][Granite4Vision] Fix deepstack buffer causing decode slowdown in compiled mode (#40917)

2026-04-28 07:43:30 +00:00

granite_speech.py

[Audio] Bundle get_generation_prompt() params into SpeechToTextParams (#36268)

2026-04-22 12:24:18 +08:00

granite.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

granitemoe.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

granitemoehybrid.py

fix: Apply embedding_multiplier to inputs_embeds (#34813)

2026-02-23 00:42:46 -08:00

granitemoeshared.py

Remove padding_index from models that don't use it for better Transformers v5 compatibility (#35189)

2026-02-24 08:04:46 -08:00

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

grok1.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

h2ovl.py

[Misc] Clean up processing logic (#37541)

2026-03-19 13:30:20 +00:00

hunyuan_v1.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

hunyuan_vision.py

[Misc] Reorganize inputs (#35182)

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

hy_v3_mtp.py

[Model] Support Hy3 preview (#40681)

2026-04-23 22:08:26 +08:00

hy_v3.py

[Model] Support Hy3 preview (#40681)

2026-04-23 22:08:26 +08:00

hyperclovax_vision_v2.py

[Misc] Reorganize inputs (#35182)

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

hyperclovax_vision.py

[Misc] Reorganize inputs (#35182)

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

hyperclovax.py

[Model] Add HyperCLOVAX-SEED-Think-14B language model support (#37107)

2026-03-16 06:40:05 +00:00

idefics2_vision_model.py

[CI][MM] Gate vision encoder attention mask to MiniCPM only, fixing Aria regression (#36206)

2026-03-06 01:17:08 -08:00

idefics3.py

[Misc] Reorganize inputs (#35182)

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

interfaces_base.py

[Model] Deprecate the score task (this will not affect users). (#37537)

2026-03-20 08:07:56 +00:00

interfaces.py

[Audio] Bundle get_generation_prompt() params into SpeechToTextParams (#36268)

2026-04-22 12:24:18 +08:00

intern_vit.py

[Model] Add torch.compile support for InternVL vision encoder (#38049)

2026-03-25 23:52:29 -07:00

internlm2_ve.py

[Chore] Update type annotation of input_ids in model forward (#33063)

2026-01-26 06:02:10 -08:00

internlm2.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

interns1_pro.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

interns1_vit.py

[MM] Align the prefix of MMEncoderAttention with Attention (#33750)

2026-02-04 04:07:30 +00:00

interns1.py

[Misc] Reorganize inputs (#35182)

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

internvl.py

[Misc] Reorganize inputs (#35182)

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

iquest_loopcoder.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

isaac.py

[Misc] Reorganize inputs (#35182)

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

jais2.py

Remove padding_index from models that don't use it for better Transformers v5 compatibility (#35189)

2026-02-24 08:04:46 -08:00

jais.py

[Model] Add AutoWeightsLoader support for jais (#38074)

2026-03-25 12:38:40 +00:00

jamba.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

jina_vl.py

[Chore] Update type annotation of input_ids in model forward (#33063)

2026-01-26 06:02:10 -08:00

jina.py

Add Jina Embeddings v5 model support (fixes #38633) (#39575)

2026-04-16 06:37:10 +00:00

kanana_v.py

[Misc] Reorganize inputs (#35182)

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

keye_vl1_5.py

[Model] Use mm_features for Keye-VL and Keye-1.5-VL M-RoPE (#39869)

2026-04-16 02:16:06 -07:00

keye.py

[Model] Use mm_features for Keye-VL and Keye-1.5-VL M-RoPE (#39869)

2026-04-16 02:16:06 -07:00

kimi_audio.py

[Audio] Bundle get_generation_prompt() params into SpeechToTextParams (#36268)

2026-04-22 12:24:18 +08:00

kimi_k25_vit.py

[Bugfix] Add quant_config in ViT of Kimi-K2.5 (#34501)

2026-02-13 16:05:34 +00:00

kimi_k25.py

fix(kimi_k25): resolve media_placeholder_token_id from tokenizer (#39344)

2026-04-11 21:10:24 -07:00

kimi_linear.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

kimi_vl.py

[Misc] Reorganize inputs (#35182)

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

laguna.py

[New Model] Laguna XS.2 implementation (#41129)

2026-04-28 14:23:00 -04:00

lfm2_moe.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

lfm2_siglip2.py

[torch.compile][BE][Multimodal] Remove requirement to set_model_tag to avoid cache conflict (#37345)

2026-03-19 17:26:12 +00:00

lfm2_vl.py

[Misc] Reorganize inputs (#35182)

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

lfm2.py

Fix Plamo 2/3 & LFM2 for Transformers v5 (#38090)

2026-03-25 12:29:49 +00:00

lightonocr.py

[Misc] Cleanup more configs and processors (#37560)

2026-03-19 15:45:23 +00:00

llama4_eagle.py

[Spec Decode] Reduce TP communication for speculative decoding draft token generation (#34049)

2026-02-22 14:59:16 -08:00

llama4.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

llama_eagle3.py

[Bugfix] Pass drafter quant_config to ParallelLMHead in Eagle3 (#37280)

2026-03-25 11:42:58 +00:00

llama_eagle.py

Add llmcompressor fp8 kv-cache quant (per-tensor and per-attn_head) (#30141)

2026-01-22 13:29:57 -07:00

llama.py

Replace shape_invariants with simpler apprach in dynamic_arg_dims utilizing shape_id property. (#36194)

2026-04-29 18:32:15 +00:00

llava_next_video.py

[Misc] Reorganize inputs (#35182)

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

llava_next.py

Increase Flexibility for OOV Multimodal Token Handling (#34858)

2026-03-08 20:30:49 -07:00

llava_onevision.py

[Misc] Reorganize inputs (#35182)

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

llava.py

[Misc] Reorganize inputs (#35182)

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

longcat_flash_mtp.py

[Chore] Update type annotation of input_ids in model forward (#33063)

2026-01-26 06:02:10 -08:00

longcat_flash.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

mamba2.py

[Model][Spec Decode] Nemotron-H MTP and Mamba Speculative Decoding Support (#33726)

2026-02-24 09:49:56 -08:00

mamba.py

[Chore] Update type annotation of input_ids in model forward (#33063)

2026-01-26 06:02:10 -08:00

medusa.py

[LoRA][1/N]Remove LoRA extra vocab (#28382)

2025-11-11 11:06:21 -08:00

midashenglm.py

[Misc] Reorganize inputs (#35182)

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

mimo_audio.py

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

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

mimo_mtp.py

[Chore] Update type annotation of input_ids in model forward (#33063)

2026-01-26 06:02:10 -08:00

mimo_v2_mtp.py

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

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

mimo_v2_omni.py

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

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

mimo_v2.py

[Model] update for mimo v25 (#41029)

2026-04-27 21:52:54 -07:00

mimo.py

[Chore] Update type annotation of input_ids in model forward (#33063)

2026-01-26 06:02:10 -08:00

minicpm3.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

minicpm_eagle.py

[Misc][Model][Refactor] Pass the prefix into Linear layers (#31669)

2026-01-05 20:03:18 +00:00

minicpm.py

[Refactor] Consolidate SupportsEagle (#36063)

2026-03-13 23:22:40 +00:00

minicpmo.py

[Misc] Reorganize inputs (#35182)

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

minicpmv.py

fix online fp8 for MiniCPM models (#39862)

2026-04-15 09:09:20 +00:00

minimax_m2.py

[Bugfix] Fix k_norm weight sharding in MiniMaxM2Attention when total_num_kv_heads < tp_size (#38191)

2026-04-27 05:57:13 -07:00

minimax_text_01.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

minimax_vl_01.py

[Bugfix] Fix models and tests for transformers v5 (#33977)

2026-02-06 21:47:41 +08:00

mistral3.py

[Bugfix][Model] Fix Devstral Small 2 HF format weight loading (#39293)

2026-04-14 10:11:18 +00:00

mistral_eagle.py

[FEATURE] Add EagleMistralForCausalLM (#41024)

2026-04-28 12:22:20 -07:00

mistral_large_3_eagle.py

Fix EagleMistralLarge3Model initialization (#37232)

2026-03-16 15:41:00 -07:00

mistral_large_3.py

Add Mistral Large 3 and Ministral 3 (#29757)

2025-12-02 10:29:00 +00:00

mistral.py

Remove unused logic in models/mistral.py (#33095)

2026-01-26 09:01:52 -08:00

mixtral.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

mllama4.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

mlp_speculator.py

[Refactor] Remove unused dead code (#38842)

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

modernbert.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

module_mapping.py

[Misc] Remove unused ModelKeys (#32608)

2026-01-19 17:34:59 +00:00

molmo2.py

[Misc] Reorganize inputs (#35182)

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

molmo.py

[Misc] Reorganize inputs (#35182)

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

moonvit.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

mpt.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

musicflamingo.py

Update to transformers v5 (#30566)

2026-04-15 16:29:15 -07:00

nano_nemotron_vl.py

Fix Nano Nemotron VL static image inputs (#40724)

2026-04-24 09:18:55 +00:00

nemotron_h_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

nemotron_h.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

nemotron_nas.py

Remove padding_index from models that don't use it for better Transformers v5 compatibility (#35189)

2026-02-24 08:04:46 -08:00

nemotron_parse.py

[Misc] Reorganize inputs (#35182)

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

nemotron_vl.py

[2/3] Refactor InternVL-based processors (#37324)

2026-03-18 22:22:19 +08:00

nemotron.py

[Misc] Cleanup more configs and processors (#37560)

2026-03-19 15:45:23 +00:00

nvlm_d.py

[Misc] Reorganize inputs (#35182)

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

olmo2.py

Various Transformers v5 fixes (#38127)

2026-03-26 00:10:08 +00:00

olmo_hybrid.py

[Kernel] Mamba support different layout for Conv state (#37416)

2026-04-03 01:50:09 +02:00

olmo.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

olmoe.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

opencua.py

[Refactor] Define MM data parser in processing info instead of processor itself (#33260)

2026-01-29 13:55:17 +08:00

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

[Model] Apply #32631 for recent models (#33785)

2026-02-04 12:23:01 +00:00

openpangu.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

opt.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

orion.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

ouro.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

ovis2_5.py

[Misc] Reorganize inputs (#35182)

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

ovis.py

[Misc] Reorganize inputs (#35182)

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

paddleocr_vl.py

[Model] Use mm_features to compute mrope positions for PaddleOCR-VL (#39888)

2026-04-16 06:14:00 -07:00

paligemma.py

[Misc] Reorganize inputs (#35182)

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

parakeet.py

Bugfix: Parakeet: .conv.pointwise/depthwise_conv1/2.bias weigths can exist even if convolution_bias=False (#40007)

2026-04-16 23:22:05 +00:00

param2moe.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

persimmon.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

phi3.py

Convert formatting to use ruff instead of yapf + isort (#26247)

2025-10-05 07:06:22 -07:00

phi3v.py

[Misc] Reorganize inputs (#35182)

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

phi4mm_audio.py

[Bugfix][CI] fix typos (#34934)

2026-03-05 17:05:46 +00:00

phi4mm_utils.py

[Bugfix][CI] fix typos (#34934)

2026-03-05 17:05:46 +00:00

phi4mm.py

[Misc] Reorganize inputs (#35182)

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

phi4siglip.py

[Model] Add Phi4ForCausalLMV for microsoft/Phi-4-reasoning-vision-15B (#38306)

2026-04-02 21:14:57 -07:00

phi.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

phimoe.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

pixtral.py

[BUGFIX] Fix Pixtral consolidated format vision weight loading (#39916)

2026-04-19 22:25:03 -07:00

plamo2.py

[Mamba] Flashinfer selective_state_update (#36162)

2026-04-14 15:10:58 -04:00

plamo3.py

Fix Plamo 2/3 & LFM2 for Transformers v5 (#38090)

2026-03-25 12:29:49 +00:00

qwen2_5_omni_thinker.py

[Refactor] Remove unused param (#39750)

2026-04-21 14:59:20 -07:00

qwen2_5_vl.py

[Model] Use helper function to run MM processors with token inputs (where applicable) (#38018)

2026-03-26 16:44:04 +08:00

qwen2_audio.py

[Misc] Reorganize inputs (#35182)

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

qwen2_moe.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

qwen2_rm.py

[Chore] Update type annotation of input_ids in model forward (#33063)

2026-01-26 06:02:10 -08:00

qwen2_vl.py

[Misc] Reorganize inputs (#35182)

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

qwen2.py

Replace shape_invariants with simpler apprach in dynamic_arg_dims utilizing shape_id property. (#36194)

2026-04-29 18:32:15 +00:00

qwen3_5_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

qwen3_5.py

[Bugfix] LoRA: extend expert base_layer loading to Qwen3.5 and Step3.x (#37114)

2026-04-21 14:17:03 +00:00

qwen3_asr_forced_aligner.py

[Feature] Add Qwen3-ForcedAligner support via token classification pooling (#35367)

2026-03-29 00:27:52 +00:00

qwen3_asr_realtime.py

[Misc] Reorganize inputs (#35182)

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

qwen3_asr.py

[Audio] Bundle get_generation_prompt() params into SpeechToTextParams (#36268)

2026-04-22 12:24:18 +08:00

qwen3_dflash.py

[Speculative Decoding] Add DFlash speculators config parsing (#38300)

2026-04-15 16:22:15 -04:00

qwen3_moe.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

qwen3_next_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

qwen3_next.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

qwen3_omni_moe_thinker.py

[Bugfix] Remove invalid deepstack boundary check for Qwen3-VL (#40932)

2026-04-27 07:27:06 +00:00

qwen3_vl_moe.py

[Bugfix] LoRA: extend expert base_layer loading to Qwen3.5 and Step3.x (#37114)

2026-04-21 14:17:03 +00:00

qwen3_vl.py

[Bugfix] Remove invalid deepstack boundary check for Qwen3-VL (#40932)

2026-04-27 07:27:06 +00:00

qwen3.py

[Feat][Spec Decode] DFlash (#36847)

2026-03-30 15:03:15 -04:00

qwen_vl.py

[Misc] Reorganize inputs (#35182)

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

qwen.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

radio.py

Fix Nano Nemotron VL regressions (#38655)

2026-04-03 15:22:06 +08:00

registry.py

[Models] Cohere MoE (#40817)

2026-04-29 15:54:54 +00:00

rnj1.py

[Model] Add block-local attention and YaRN for local layers to Gemma3 (#39823)

2026-04-21 23:34:50 -07:00

roberta.py

[Bugfix] Fix RoBERTa position_ids accumulation on CUDA graph padding (#37884)

2026-03-23 15:15:12 +00:00

rvl.py

[Misc] Reorganize inputs (#35182)

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

sarvam.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

seed_oss.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

siglip2navit.py

[New Model] support new model ovis2.6 (#34426)

2026-02-13 00:12:45 -08:00

siglip.py

[Misc] Reorganize inputs (#35182)

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

skyworkr1v.py

[Misc] Reorganize inputs (#35182)

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

smolvlm.py

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

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

solar.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

stablelm.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

starcoder2.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00

step1.py

[Refactor] Consolidate SupportsEagle (#36063)

2026-03-13 23:22:40 +00:00

step3_text.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

step3_vl.py

[Misc] Reorganize inputs (#35182)

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

step3p5_mtp.py

[Bugfix] LoRA: extend expert base_layer loading to Qwen3.5 and Step3.x (#37114)

2026-04-21 14:17:03 +00:00

step3p5.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

step_vl.py

[MM] Align the prefix of MMEncoderAttention with Attention (#33750)

2026-02-04 04:07:30 +00:00

tarsier.py

[Misc] Cleanup more configs and processors (#37560)

2026-03-19 15:45:23 +00:00

telechat2.py

Convert formatting to use ruff instead of yapf + isort (#26247)

2025-10-05 07:06:22 -07:00

teleflm.py

[LoRA][2/2]Remove LoRA extra vocab (#28545)

2025-11-21 09:46:43 +08:00

terratorch.py

[Misc] Reorganize inputs (#35182)

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

ultravox.py

[Misc] Reorganize inputs (#35182)

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

utils.py

feat(models): implement Google Gemma 4 architecture support (MoE, Multimodal, Reasoning, Tool-Use) (#38826)

2026-04-02 11:13:28 -07:00

vision.py

[Perf] FP8 FlashInfer Attn for ViT (#38065)

2026-04-27 13:44:15 +08:00

voxtral_realtime.py

[Audio] Bundle get_generation_prompt() params into SpeechToTextParams (#36268)

2026-04-22 12:24:18 +08:00

voxtral.py

[Audio] Bundle get_generation_prompt() params into SpeechToTextParams (#36268)

2026-04-22 12:24:18 +08:00

voyage.py

Fix main pre-commit (#33975)

2026-02-06 00:08:05 -08:00

whisper_causal.py

[Attention][V0 Deprecation] Deprecate accept output buffer (#39125)

2026-04-07 17:14:58 -04:00

whisper_utils.py

[Voxtral] Add new streaming arch (#32861)

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

whisper.py

[Audio] Bundle get_generation_prompt() params into SpeechToTextParams (#36268)

2026-04-22 12:24:18 +08:00

zamba2.py

[5/N][Attention] Finish eliminating vllm/attention folder (#32064)

2026-01-27 10:02:51 -05:00