xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-01-02 05:00:53 +08:00
Code Issues Packages Projects Releases Wiki Activity
vllm/vllm/model_executor/models
History
Zhonghua Deng 969bbc7c61
[Model] Add MiMo-V2-Flash support (#30836) 
Signed-off-by: Abatom <abzhonghua@gmail.com>
Signed-off-by: Jumiar <liuanqim10@126.com>
Signed-off-by: Zyann7 <zyann7@outlook.com>
Co-authored-by: Jumiar <liuanqim10@126.com>
Co-authored-by: Zyann7 <zyann7@outlook.com>
Co-authored-by: Jee Jee Li <pandaleefree@gmail.com>
2025-12-19 17:17:03 +00:00
..
transformers
[ROCm][CI][Bugfix] Multi-Modal Model Support Fixes and Attention Backend Improvements (#30270)
2025-12-19 02:17:27 +00:00
__init__.py
…
adapters.py
[Model] Automatic conversion of TokenClassification model (#30666)
2025-12-15 08:13:00 +00:00
afmoe.py
[Bugfix][Model] Fix Afmoe rope_parameters issue (#30505)
2025-12-12 02:53:57 +00:00
aimv2.py
[MM Encoder]: Migrate legacy ViT MultiHeadAttention to new MMEncoderAttention interface (#30684)
2025-12-19 02:04:19 +08:00
apertus.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
arcee.py
…
arctic.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
aria.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
audioflamingo3.py
Add AudioFlamingo3 model support (#30539)
2025-12-14 02:14:55 -08:00
aya_vision.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
bagel.py
[New Model] BAGEL support (AR only) (#28439)
2025-12-15 14:58:23 +08:00
baichuan.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
bailing_moe.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
bamba.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
bee.py
…
bert_with_rope.py
…
bert.py
Don'e assume position_embedding_type will be present for BERT and RoBERTa models (#30770)
2025-12-16 13:40:26 +00:00
blip2.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
blip.py
[MM Encoder]: Migrate legacy ViT MultiHeadAttention to new MMEncoderAttention interface (#30684)
2025-12-19 02:04:19 +08:00
bloom.py
[Attention] Remove imports from vllm/attention/__init__.py (#29342)
2025-11-26 10:53:15 -07:00
chameleon.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
chatglm.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
clip.py
[MM Encoder]: Migrate legacy ViT MultiHeadAttention to new MMEncoderAttention interface (#30684)
2025-12-19 02:04:19 +08:00
cohere2_vision.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
commandr.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
config.py
[Mamba] Removed disable cascade attn in MambaModelConfig (#30712)
2025-12-17 08:48:53 +00:00
dbrx.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
deepencoder.py
[MM Encoder]: Migrate legacy ViT MultiHeadAttention to new MMEncoderAttention interface (#30684)
2025-12-19 02:04:19 +08:00
deepseek_eagle.py
[Misc] Remove redundant attention var constants (#29650)
2025-11-28 04:35:19 -08:00
deepseek_mtp.py
[refactor] Add prefix support to embed_tokens in DeepSeek MTP (#30788)
2025-12-18 04:45:56 +00:00
deepseek_ocr.py
Revert "[Renderer] Separate out RendererConfig from ModelConfig (#30145)" (#30199)
2025-12-07 00:00:22 -08:00
deepseek_v2.py
[Bugfix] deepseek-V3.2 self.weights_proj has no bias (#30841)
2025-12-17 03:32:34 -08:00
deepseek_vl2.py
Revert "[Renderer] Separate out RendererConfig from ModelConfig (#30145)" (#30199)
2025-12-07 00:00:22 -08:00
dots1.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
dots_ocr.py
[CustomOp] Extract ApplyRotaryEmb as CustomOp and unify the dispatch logic (#29873)
2025-12-15 19:08:16 -08:00
ernie45_moe.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
ernie45_vl_moe.py
[Attention] Remove imports from vllm/attention/__init__.py (#29342)
2025-11-26 10:53:15 -07:00
ernie45_vl.py
[CustomOp] Extract ApplyRotaryEmb as CustomOp and unify the dispatch logic (#29873)
2025-12-15 19:08:16 -08:00
ernie45.py
…
ernie_mtp.py
…
exaone4.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
exaone.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
fairseq2_llama.py
…
falcon_h1.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
falcon.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
flex_olmo.py
…
fuyu.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
gemma2.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
gemma3_mm.py
[Bugfix] Schedule failure due to wrong get_image_size_with_most_features (#29692)
2025-12-12 02:27:20 -08:00
gemma3.py
[v1] Add PrefixLM support to TritonAttention backend (#30386)
2025-12-17 16:05:24 -08:00
gemma3n_mm.py
Revert "[Renderer] Separate out RendererConfig from ModelConfig (#30145)" (#30199)
2025-12-07 00:00:22 -08:00
gemma3n.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
gemma.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
glm4_1v.py
[CustomOp] Extract ApplyRotaryEmb as CustomOp and unify the dispatch logic (#29873)
2025-12-15 19:08:16 -08:00
glm4_moe_mtp.py
…
glm4_moe.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
glm4.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
glm4v.py
[MM Encoder]: Migrate legacy ViT MultiHeadAttention to new MMEncoderAttention interface (#30684)
2025-12-19 02:04:19 +08:00
glm.py
Access partial_rotary_factor from rope_parameters (#29966)
2025-12-04 18:42:49 +00:00
gpt2.py
[Attention] Remove imports from vllm/attention/__init__.py (#29342)
2025-11-26 10:53:15 -07:00
gpt_bigcode.py
[Attention] Remove imports from vllm/attention/__init__.py (#29342)
2025-11-26 10:53:15 -07:00
gpt_j.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
gpt_neox.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
gpt_oss.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
granite_speech.py
Revert "[Renderer] Separate out RendererConfig from ModelConfig (#30145)" (#30199)
2025-12-07 00:00:22 -08:00
granite.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
granitemoe.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
granitemoehybrid.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
granitemoeshared.py
[LoRA] Cleanup LoRA unused code (#29611)
2025-11-28 22:52:58 -08:00
gritlm.py
Revert "[Renderer] Separate out RendererConfig from ModelConfig (#30145)" (#30199)
2025-12-07 00:00:22 -08:00
grok1.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
h2ovl.py
[Misc] Refactor tokenizer interface (#29693)
2025-11-29 04:02:21 -08:00
hunyuan_v1.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
hunyuan_vision.py
[MM Encoder]: Migrate legacy ViT MultiHeadAttention to new MMEncoderAttention interface (#30684)
2025-12-19 02:04:19 +08:00
hyperclovax_vision.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
idefics2_vision_model.py
[MM Encoder]: Migrate legacy ViT MultiHeadAttention to new MMEncoderAttention interface (#30684)
2025-12-19 02:04:19 +08:00
idefics3.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
interfaces_base.py
[Deprecation] Remove missed fallback for embed_input_ids (#30469)
2025-12-11 10:06:36 +00:00
interfaces.py
[Misc] Improve error message for is_multimodal (#30483)
2025-12-11 06:39:51 -08:00
intern_vit.py
[MM Encoder]: Migrate legacy ViT MultiHeadAttention to new MMEncoderAttention interface (#30684)
2025-12-19 02:04:19 +08:00
internlm2_ve.py
…
internlm2.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
interns1_vit.py
[MM Encoder]: Migrate legacy ViT MultiHeadAttention to new MMEncoderAttention interface (#30684)
2025-12-19 02:04:19 +08:00
interns1.py
Revert "[Renderer] Separate out RendererConfig from ModelConfig (#30145)" (#30199)
2025-12-07 00:00:22 -08:00
internvl.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
jais2.py
adds jais 2 support (#30188)
2025-12-18 15:46:58 +00:00
jais.py
[Attention] Remove imports from vllm/attention/__init__.py (#29342)
2025-11-26 10:53:15 -07:00
jamba.py
[LoRA] Cleanup LoRA unused code (#29611)
2025-11-28 22:52:58 -08:00
jina_vl.py
Fix some Transformers nightly tests (#29802)
2025-12-02 07:05:27 +00:00
keye_vl1_5.py
[Misc] Various cleanups for MM input processing (#29970)
2025-12-04 06:22:20 +00:00
keye.py
[CustomOp] Extract ApplyRotaryEmb as CustomOp and unify the dispatch logic (#29873)
2025-12-15 19:08:16 -08:00
kimi_linear.py
…
kimi_vl.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
lfm2_moe.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
lfm2.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
lightonocr.py
[Chore] Deprecate merge_by_field_config arg (#30035)
2025-12-04 17:21:24 +00:00
llama4_eagle.py
[Bugfix][llama4_eagle] Fix missing 'lm_head' attribute (#29926)
2025-12-05 19:57:26 +00:00
llama4.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
llama_eagle3.py
[Spec Decode] Add support for EAGLE3 heads that do not use_aux_hidden_states (#27688)
2025-11-24 20:13:12 -05:00
llama_eagle.py
[Bugfix] Fix spec decode memory regression after #28549 (#28819)
2025-11-20 19:05:50 +08:00
llama.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
llava_next_video.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
llava_next.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
llava_onevision.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
llava.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
longcat_flash_mtp.py
…
longcat_flash.py
…
mamba2.py
…
mamba.py
…
medusa.py
…
midashenglm.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
mimo_mtp.py
…
mimo_v2_flash.py
[Model] Add MiMo-V2-Flash support (#30836)
2025-12-19 17:17:03 +00:00
mimo.py
…
minicpm3.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
minicpm_eagle.py
[LoRA] Cleanup LoRA unused code (#29611)
2025-11-28 22:52:58 -08:00
minicpm.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
minicpmo.py
…
minicpmv.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
minimax_m2.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
minimax_text_01.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
minimax_vl_01.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
mistral3.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
mistral_large_3_eagle.py
[Deprecation] Remove fallbacks for embed_input_ids and embed_multimodal (#30458)
2025-12-11 06:58:23 +00:00
mistral_large_3.py
Add Mistral Large 3 and Ministral 3 (#29757)
2025-12-02 10:29:00 +00:00
mixtral.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
mllama4.py
[MM Encoder]: Migrate legacy ViT MultiHeadAttention to new MMEncoderAttention interface (#30684)
2025-12-19 02:04:19 +08:00
mlp_speculator.py
…
modernbert.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
module_mapping.py
…
molmo.py
[MM Encoder]: Migrate legacy ViT MultiHeadAttention to new MMEncoderAttention interface (#30684)
2025-12-19 02:04:19 +08:00
moonvit.py
[XPU]fix Kimi-VL-A3B-thinking on xpu (#29309)
2025-11-24 21:02:21 +00:00
mpt.py
[Attention] Remove imports from vllm/attention/__init__.py (#29342)
2025-11-26 10:53:15 -07:00
nano_nemotron_vl.py
Revert "[Renderer] Separate out RendererConfig from ModelConfig (#30145)" (#30199)
2025-12-07 00:00:22 -08:00
nemotron_h.py
Add latent MoE support (#30203)
2025-12-08 17:35:01 +00:00
nemotron_nas.py
Fix nemotron_nas intermediate_size computation (#30795)
2025-12-17 01:06:28 +00:00
nemotron_vl.py
Revert "[Renderer] Separate out RendererConfig from ModelConfig (#30145)" (#30199)
2025-12-07 00:00:22 -08:00
nemotron.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
nvlm_d.py
…
olmo2.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
olmo.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
olmoe.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
opencua.py
[CustomOp][MM] Extract MMEncoderAttention as CustomOp and replace the backend of QwenVisionAttention with it. (#30125)
2025-12-15 11:13:32 +08:00
openpangu_mtp.py
…
openpangu.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
opt.py
[Attention] Remove imports from vllm/attention/__init__.py (#29342)
2025-11-26 10:53:15 -07:00
orion.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
ouro.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
ovis2_5.py
[CustomOp][MM] Extract MMEncoderAttention as CustomOp and replace the backend of QwenVisionAttention with it. (#30125)
2025-12-15 11:13:32 +08:00
ovis.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
paddleocr_vl.py
[CustomOp] Extract ApplyRotaryEmb as CustomOp and unify the dispatch logic (#29873)
2025-12-15 19:08:16 -08:00
paligemma.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
persimmon.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
phi3.py
…
phi3v.py
[Misc] Improve error message for is_multimodal (#30483)
2025-12-11 06:39:51 -08:00
phi4mm_audio.py
…
phi4mm_utils.py
…
phi4mm.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
phi.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
phimoe.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
pixtral.py
[Chore] Adjust tokenizer import to avoid circular imports (#30601)
2025-12-13 04:42:39 -08:00
plamo2.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
plamo3.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
qwen2_5_omni_thinker.py
[Chore] Remove v0 dead code for Qwen2.5-omni (#30883)
2025-12-17 19:54:39 -08:00
qwen2_5_vl.py
[CustomOp] Extract ApplyRotaryEmb as CustomOp and unify the dispatch logic (#29873)
2025-12-15 19:08:16 -08:00
qwen2_audio.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
qwen2_moe.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
qwen2_rm.py
…
qwen2_vl.py
[CustomOp] Extract ApplyRotaryEmb as CustomOp and unify the dispatch logic (#29873)
2025-12-15 19:08:16 -08:00
qwen2.py
[New Model] BAGEL support (AR only) (#28439)
2025-12-15 14:58:23 +08:00
qwen3_moe.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
qwen3_next_mtp.py
…
qwen3_next.py
[Bugfix] Qwen3-next with --hf-overrides \{\"num_hidden_layers\":8\} (#30433)
2025-12-13 22:12:45 +08:00
qwen3_omni_moe_thinker.py
[CustomOp][MM] Extract MMEncoderAttention as CustomOp and replace the backend of QwenVisionAttention with it. (#30125)
2025-12-15 11:13:32 +08:00
qwen3_vl_moe.py
[Bugfix] Fix multimodal configuration for Qwen3VL MOE model (#30670)
2025-12-15 14:06:01 +00:00
qwen3_vl.py
[Core][MM] Optimize encoder cache manager by operating with embeddings only (#30475)
2025-12-16 14:18:17 -08:00
qwen3.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
qwen_vl.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
qwen.py
[BugFix] Add embed_input_ids method to make QWenLMHeadModel a vllm model (#30674)
2025-12-15 10:38:29 +00:00
radio.py
…
registry.py
[Model] Add MiMo-V2-Flash support (#30836)
2025-12-19 17:17:03 +00:00
roberta.py
Don'e assume position_embedding_type will be present for BERT and RoBERTa models (#30770)
2025-12-16 13:40:26 +00:00
rvl.py
…
seed_oss.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
siglip2navit.py
[CustomOp] Extract ApplyRotaryEmb as CustomOp and unify the dispatch logic (#29873)
2025-12-15 19:08:16 -08:00
siglip.py
[MM Encoder]: Migrate legacy ViT MultiHeadAttention to new MMEncoderAttention interface (#30684)
2025-12-19 02:04:19 +08:00
skyworkr1v.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
smolvlm.py
…
solar.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
stablelm.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
starcoder2.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
step3_text.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00
step3_vl.py
[MM Encoder]: Migrate legacy ViT MultiHeadAttention to new MMEncoderAttention interface (#30684)
2025-12-19 02:04:19 +08:00
swin.py
Remove head_mask from Ultravox and Swin (#30764)
2025-12-16 08:02:41 -08:00
tarsier.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
telechat2.py
…
teleflm.py
[LoRA][2/2]Remove LoRA extra vocab (#28545)
2025-11-21 09:46:43 +08:00
terratorch.py
[Chore] Deprecate SupportsMultiModal.merge_by_field_config (#30170)
2025-12-06 07:57:28 +00:00
ultravox.py
Remove head_mask from Ultravox and Swin (#30764)
2025-12-16 08:02:41 -08:00
utils.py
[Deprecation] Advance deprecation status (#29617)
2025-11-27 19:04:58 +00:00
vision.py
[Bugfix] Fix broken ViT attention selection for Blackwell device (#30731)
2025-12-16 05:24:32 +00:00
voxtral.py
[Chore] Adjust tokenizer import to avoid circular imports (#30601)
2025-12-13 04:42:39 -08:00
whisper.py
[MM Encoder]: Migrate legacy ViT MultiHeadAttention to new MMEncoderAttention interface (#30684)
2025-12-19 02:04:19 +08:00
zamba2.py
Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)
2025-12-11 20:45:23 +00:00