Remove all references to `yapf` as it's no longer used (#26251)

Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>

Remove all references to `yapf` as it's no longer used (#26251)
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
4e256cad · Harry Mellor · GitHub · d6953beb · 4e256cad · 4e256cad
Unverified Commit 4e256cad authored Oct 05, 2025 by Harry Mellor Committed by GitHub Oct 05, 2025
18 changed files
--- a/vllm/model_executor/models/gemma3n_mm.py
+++ b/vllm/model_executor/models/gemma3n_mm.py
@@ -6,14 +6,16 @@ from typing import Annotated, Any, Literal, Optional, Union, cast
 import numpy as np
 import torch
-# yapf: disable
 from torch import nn
 from transformers import AutoModel, BatchFeature
-from transformers.models.gemma3n import (Gemma3nAudioConfig,
+from transformers.models.gemma3n import (
+    Gemma3nAudioConfig,
    Gemma3nAudioFeatureExtractor,
-                                         Gemma3nConfig, Gemma3nProcessor,
+    Gemma3nConfig,
+    Gemma3nProcessor,
    Gemma3nTextConfig,
-                                         Gemma3nVisionConfig)
+    Gemma3nVisionConfig,
+)
 from transformers.models.siglip import SiglipImageProcessorFast
 from vllm.config import ModelConfig, SpeechToTextConfig, VllmConfig
@@ -22,25 +24,32 @@ from vllm.inputs.data import PromptType
 from vllm.logger import init_logger
 from vllm.model_executor.layers.layernorm import RMSNorm
 from vllm.model_executor.layers.linear import RowParallelLinear
-from vllm.model_executor.layers.vocab_parallel_embedding import (
+from vllm.model_executor.layers.vocab_parallel_embedding import VocabParallelEmbedding
-    VocabParallelEmbedding)
 from vllm.model_executor.models.gemma3n import Gemma3nForCausalLM
 from vllm.model_executor.models.module_mapping import MultiModelKeys
 from vllm.model_executor.models.whisper import ISO639_1_SUPPORTED_LANGS
 from vllm.multimodal import MULTIMODAL_REGISTRY
-from vllm.multimodal.inputs import (MultiModalDataDict, MultiModalFieldConfig,
+from vllm.multimodal.inputs import (
-                                    MultiModalKwargsItems)
+    MultiModalDataDict,
-from vllm.multimodal.parse import (ImageProcessorItems, MultiModalDataItems,
+    MultiModalFieldConfig,
-                                   MultiModalDataParser)
+    MultiModalKwargsItems,
-from vllm.multimodal.processing import (BaseMultiModalProcessor,
+)
+from vllm.multimodal.parse import (
+    ImageProcessorItems,
+    MultiModalDataItems,
+    MultiModalDataParser,
+)
+from vllm.multimodal.processing import (
+    BaseMultiModalProcessor,
    BaseProcessingInfo,
    MultiModalPromptUpdates,
    MultiModalPromptUpdatesApplyResult,
    PlaceholderFeaturesInfo,
-                                        PromptReplacement, PromptUpdate,
+    PromptReplacement,
+    PromptUpdate,
    PromptUpdateDetails,
-                                        replace_token_matches)
+    replace_token_matches,
-# yapf: enable
+)
 from vllm.multimodal.profiling import BaseDummyInputsBuilder
 from vllm.sequence import IntermediateTensors
 from vllm.utils.tensor_schema import TensorSchema, TensorShape

--- a/vllm/model_executor/models/idefics3.py
+++ b/vllm/model_executor/models/idefics3.py
@@ -43,9 +43,6 @@ from vllm.multimodal.inputs import (
    MultiModalKwargsItems,
 )
 from vllm.multimodal.parse import ImageProcessorItems, ImageSize
-# yapf conflicts with isort for this block
-# yapf: disable
 from vllm.multimodal.processing import (
    BaseMultiModalProcessor,
    BaseProcessingInfo,
@@ -54,18 +51,13 @@ from vllm.multimodal.processing import (
    PromptUpdate,
    PromptUpdateDetails,
 )
-# yapf: enable
 from vllm.multimodal.profiling import BaseDummyInputsBuilder
 from vllm.sequence import IntermediateTensors
 from vllm.utils.tensor_schema import TensorSchema, TensorShape
-# yapf: disable
 from .idefics2_vision_model import (
    Idefics2VisionTransformer as Idefics3VisionTransformer,
 )
-# yapf: enable
 from .interfaces import MultiModalEmbeddings, SupportsLoRA, SupportsMultiModal
 from .llama import LlamaModel
 from .utils import AutoWeightsLoader, maybe_prefix

--- a/vllm/model_executor/models/phi3v.py
+++ b/vllm/model_executor/models/phi3v.py
@@ -45,9 +45,6 @@ from vllm.multimodal.parse import (
    ImageSize,
    MultiModalDataItems,
 )
-# yapf conflicts with isort for this block
-# yapf: disable
 from vllm.multimodal.processing import (
    BaseMultiModalProcessor,
    BaseProcessingInfo,
@@ -57,8 +54,6 @@ from vllm.multimodal.processing import (
    PromptUpdate,
    ResolvedPromptUpdate,
 )
-# yapf: enable
 from vllm.multimodal.profiling import BaseDummyInputsBuilder
 from vllm.sequence import IntermediateTensors
 from vllm.utils import is_list_of

--- a/vllm/model_executor/models/qwen2_5_vl.py
+++ b/vllm/model_executor/models/qwen2_5_vl.py
@@ -52,16 +52,12 @@ from vllm.distributed import utils as dist_utils
 from vllm.logger import init_logger
 from vllm.model_executor.layers.activation import get_act_and_mul_fn
 from vllm.model_executor.layers.layernorm import RMSNorm
-# yapf: disable
 from vllm.model_executor.layers.linear import (
    ColumnParallelLinear,
    MergedColumnParallelLinear,
    QKVParallelLinear,
    RowParallelLinear,
 )
-# yapf: enable
 from vllm.model_executor.layers.quantization import QuantizationConfig
 from vllm.model_executor.model_loader.weight_utils import default_weight_loader
 from vllm.model_executor.models.module_mapping import MultiModelKeys

--- a/vllm/model_executor/models/qwen3_next.py
+++ b/vllm/model_executor/models/qwen3_next.py
@@ -37,12 +37,7 @@ from vllm.model_executor.layers.fla.ops import (
    fused_recurrent_gated_delta_rule,
 )
 from vllm.model_executor.layers.fused_moe import FusedMoE
-# yapf conflicts with isort for this block
-# yapf: disable
 from vllm.model_executor.layers.layernorm import GemmaRMSNorm as Qwen3NextRMSNorm
-# yapf: enable
 from vllm.model_executor.layers.linear import (
    ColumnParallelLinear,
    QKVParallelLinear,

--- a/vllm/model_executor/models/registry.py
+++ b/vllm/model_executor/models/registry.py
@@ -54,7 +54,6 @@ from .interfaces_base import (
 logger = init_logger(__name__)
-# yapf: disable
 _TEXT_GENERATION_MODELS = {
    # [Decoder-only]
    "ApertusForCausalLM": ("apertus", "ApertusForCausalLM"),
@@ -127,7 +126,7 @@ _TEXT_GENERATION_MODELS = {
    "LongcatFlashForCausalLM": ("longcat_flash", "LongcatFlashForCausalLM"),
    "MambaForCausalLM": ("mamba", "MambaForCausalLM"),
    "FalconMambaForCausalLM": ("mamba", "MambaForCausalLM"),
-    "FalconH1ForCausalLM":("falcon_h1", "FalconH1ForCausalLM"),
+    "FalconH1ForCausalLM": ("falcon_h1", "FalconH1ForCausalLM"),
    "Mamba2ForCausalLM": ("mamba2", "Mamba2ForCausalLM"),
    "MiniCPMForCausalLM": ("minicpm", "MiniCPMForCausalLM"),
    "MiniCPM3ForCausalLM": ("minicpm3", "MiniCPM3ForCausalLM"),
@@ -184,7 +183,8 @@ _EMBEDDING_MODELS = {
    "LlamaModel": ("llama", "LlamaForCausalLM"),
    **{
        # Multiple models share the same architecture, so we include them all
-        k: (mod, arch) for k, (mod, arch) in _TEXT_GENERATION_MODELS.items()
+        k: (mod, arch)
+        for k, (mod, arch) in _TEXT_GENERATION_MODELS.items()
        if arch == "LlamaForCausalLM"
    },
    "MistralModel": ("llama", "LlamaForCausalLM"),
@@ -201,7 +201,10 @@ _EMBEDDING_MODELS = {
    "XLMRobertaModel": ("roberta", "RobertaEmbeddingModel"),
    # [Multimodal]
    "CLIPModel": ("clip", "CLIPEmbeddingModel"),
-    "LlavaNextForConditionalGeneration": ("llava_next", "LlavaNextForConditionalGeneration"),  # noqa: E501
+    "LlavaNextForConditionalGeneration": (
+        "llava_next",
+        "LlavaNextForConditionalGeneration",
+    ),  # noqa: E501
    "Phi3VForCausalLM": ("phi3v", "Phi3VForCausalLM"),
    "Qwen2VLForConditionalGeneration": ("qwen2_vl", "Qwen2VLForConditionalGeneration"),  # noqa: E501
    # Technically Terratorch models work on images, both in
@@ -214,14 +217,19 @@ _EMBEDDING_MODELS = {
 _CROSS_ENCODER_MODELS = {
    "BertForSequenceClassification": ("bert", "BertForSequenceClassification"),
    "BertForTokenClassification": ("bert", "BertForTokenClassification"),
-    "GteNewForSequenceClassification": ("bert_with_rope",
+    "GteNewForSequenceClassification": (
-                                        "GteNewForSequenceClassification"),
+        "bert_with_rope",
-    "ModernBertForSequenceClassification": ("modernbert",
+        "GteNewForSequenceClassification",
-                                            "ModernBertForSequenceClassification"),
+    ),
-    "RobertaForSequenceClassification": ("roberta",
+    "ModernBertForSequenceClassification": (
-                                         "RobertaForSequenceClassification"),
+        "modernbert",
-    "XLMRobertaForSequenceClassification": ("roberta",
+        "ModernBertForSequenceClassification",
-                                            "RobertaForSequenceClassification"),
+    ),
+    "RobertaForSequenceClassification": ("roberta", "RobertaForSequenceClassification"),
+    "XLMRobertaForSequenceClassification": (
+        "roberta",
+        "RobertaForSequenceClassification",
+    ),
    # [Auto-converted (see adapters.py)]
    "JinaVLForRanking": ("jina_vl", "JinaVLForSequenceClassification"),  # noqa: E501,
 }
@@ -229,64 +237,130 @@ _CROSS_ENCODER_MODELS = {
 _MULTIMODAL_MODELS = {
    # [Decoder-only]
    "AriaForConditionalGeneration": ("aria", "AriaForConditionalGeneration"),
-    "AyaVisionForConditionalGeneration": ("aya_vision", "AyaVisionForConditionalGeneration"),  # noqa: E501
+    "AyaVisionForConditionalGeneration": (
+        "aya_vision",
+        "AyaVisionForConditionalGeneration",
+    ),  # noqa: E501
    "Blip2ForConditionalGeneration": ("blip2", "Blip2ForConditionalGeneration"),
-    "ChameleonForConditionalGeneration": ("chameleon", "ChameleonForConditionalGeneration"),  # noqa: E501
+    "ChameleonForConditionalGeneration": (
-    "Cohere2VisionForConditionalGeneration": ("cohere2_vision", "Cohere2VisionForConditionalGeneration"),  # noqa: E501
+        "chameleon",
+        "ChameleonForConditionalGeneration",
+    ),  # noqa: E501
+    "Cohere2VisionForConditionalGeneration": (
+        "cohere2_vision",
+        "Cohere2VisionForConditionalGeneration",
+    ),  # noqa: E501
    "DeepseekVLV2ForCausalLM": ("deepseek_vl2", "DeepseekVLV2ForCausalLM"),
    "DotsOCRForCausalLM": ("dots_ocr", "DotsOCRForCausalLM"),
-    "Ernie4_5_VLMoeForConditionalGeneration": ("ernie45_vl", "Ernie4_5_VLMoeForConditionalGeneration"),  # noqa: E501
+    "Ernie4_5_VLMoeForConditionalGeneration": (
+        "ernie45_vl",
+        "Ernie4_5_VLMoeForConditionalGeneration",
+    ),  # noqa: E501
    "FuyuForCausalLM": ("fuyu", "FuyuForCausalLM"),
    "Gemma3ForConditionalGeneration": ("gemma3_mm", "Gemma3ForConditionalGeneration"),  # noqa: E501
-    "Gemma3nForConditionalGeneration": ("gemma3n_mm", "Gemma3nForConditionalGeneration"),    # noqa: E501
+    "Gemma3nForConditionalGeneration": (
+        "gemma3n_mm",
+        "Gemma3nForConditionalGeneration",
+    ),  # noqa: E501
    "GLM4VForCausalLM": ("glm4v", "GLM4VForCausalLM"),
    "Glm4vForConditionalGeneration": ("glm4_1v", "Glm4vForConditionalGeneration"),  # noqa: E501
    "Glm4vMoeForConditionalGeneration": ("glm4_1v", "Glm4vMoeForConditionalGeneration"),  # noqa: E501
-    "GraniteSpeechForConditionalGeneration": ("granite_speech", "GraniteSpeechForConditionalGeneration"),  # noqa: E501
+    "GraniteSpeechForConditionalGeneration": (
+        "granite_speech",
+        "GraniteSpeechForConditionalGeneration",
+    ),  # noqa: E501
    "H2OVLChatModel": ("h2ovl", "H2OVLChatModel"),
    "InternVLChatModel": ("internvl", "InternVLChatModel"),
    "NemotronH_Nano_VL_V2": ("nano_nemotron_vl", "NemotronH_Nano_VL_V2"),
-    "InternS1ForConditionalGeneration": ("interns1", "InternS1ForConditionalGeneration"),  # noqa: E501
+    "InternS1ForConditionalGeneration": (
-    "InternVLForConditionalGeneration": ("interns1", "InternS1ForConditionalGeneration"),  # noqa: E501
+        "interns1",
-    "Idefics3ForConditionalGeneration":("idefics3","Idefics3ForConditionalGeneration"),
+        "InternS1ForConditionalGeneration",
-    "SmolVLMForConditionalGeneration": ("smolvlm","SmolVLMForConditionalGeneration"),  # noqa: E501
+    ),  # noqa: E501
+    "InternVLForConditionalGeneration": (
+        "interns1",
+        "InternS1ForConditionalGeneration",
+    ),  # noqa: E501
+    "Idefics3ForConditionalGeneration": (
+        "idefics3",
+        "Idefics3ForConditionalGeneration",
+    ),
+    "SmolVLMForConditionalGeneration": ("smolvlm", "SmolVLMForConditionalGeneration"),  # noqa: E501
    "KeyeForConditionalGeneration": ("keye", "KeyeForConditionalGeneration"),
-    "KeyeVL1_5ForConditionalGeneration": ("keye_vl1_5", "KeyeVL1_5ForConditionalGeneration"), # noqa: E501
+    "KeyeVL1_5ForConditionalGeneration": (
+        "keye_vl1_5",
+        "KeyeVL1_5ForConditionalGeneration",
+    ),  # noqa: E501
    "RForConditionalGeneration": ("rvl", "RForConditionalGeneration"),
    "KimiVLForConditionalGeneration": ("kimi_vl", "KimiVLForConditionalGeneration"),  # noqa: E501
    "Llama_Nemotron_Nano_VL": ("nemotron_vl", "LlamaNemotronVLChatModel"),
    "Llama4ForConditionalGeneration": ("mllama4", "Llama4ForConditionalGeneration"),  # noqa: E501
    "LlavaForConditionalGeneration": ("llava", "LlavaForConditionalGeneration"),
-    "LlavaNextForConditionalGeneration": ("llava_next", "LlavaNextForConditionalGeneration"),  # noqa: E501
+    "LlavaNextForConditionalGeneration": (
-    "LlavaNextVideoForConditionalGeneration": ("llava_next_video", "LlavaNextVideoForConditionalGeneration"),  # noqa: E501
+        "llava_next",
-    "LlavaOnevisionForConditionalGeneration": ("llava_onevision", "LlavaOnevisionForConditionalGeneration"),  # noqa: E501
+        "LlavaNextForConditionalGeneration",
+    ),  # noqa: E501
+    "LlavaNextVideoForConditionalGeneration": (
+        "llava_next_video",
+        "LlavaNextVideoForConditionalGeneration",
+    ),  # noqa: E501
+    "LlavaOnevisionForConditionalGeneration": (
+        "llava_onevision",
+        "LlavaOnevisionForConditionalGeneration",
+    ),  # noqa: E501
    "MantisForConditionalGeneration": ("llava", "MantisForConditionalGeneration"),  # noqa: E501
    "MiDashengLMModel": ("midashenglm", "MiDashengLMModel"),
-    "MiniMaxVL01ForConditionalGeneration": ("minimax_vl_01", "MiniMaxVL01ForConditionalGeneration"),  # noqa: E501
+    "MiniMaxVL01ForConditionalGeneration": (
+        "minimax_vl_01",
+        "MiniMaxVL01ForConditionalGeneration",
+    ),  # noqa: E501
    "MiniCPMO": ("minicpmo", "MiniCPMO"),
    "MiniCPMV": ("minicpmv", "MiniCPMV"),
-    "Mistral3ForConditionalGeneration": ("mistral3", "Mistral3ForConditionalGeneration"),  # noqa: E501
+    "Mistral3ForConditionalGeneration": (
+        "mistral3",
+        "Mistral3ForConditionalGeneration",
+    ),  # noqa: E501
    "MolmoForCausalLM": ("molmo", "MolmoForCausalLM"),
    "NVLM_D": ("nvlm_d", "NVLM_D_Model"),
    "Ovis": ("ovis", "Ovis"),
    "Ovis2_5": ("ovis2_5", "Ovis2_5"),
-    "PaliGemmaForConditionalGeneration": ("paligemma", "PaliGemmaForConditionalGeneration"),  # noqa: E501
+    "PaliGemmaForConditionalGeneration": (
+        "paligemma",
+        "PaliGemmaForConditionalGeneration",
+    ),  # noqa: E501
    "Phi3VForCausalLM": ("phi3v", "Phi3VForCausalLM"),
    "Phi4MMForCausalLM": ("phi4mm", "Phi4MMForCausalLM"),
    "Phi4MultimodalForCausalLM": ("phi4_multimodal", "Phi4MultimodalForCausalLM"),  # noqa: E501
    "PixtralForConditionalGeneration": ("pixtral", "PixtralForConditionalGeneration"),  # noqa: E501
    "QwenVLForConditionalGeneration": ("qwen_vl", "QwenVLForConditionalGeneration"),  # noqa: E501
    "Qwen2VLForConditionalGeneration": ("qwen2_vl", "Qwen2VLForConditionalGeneration"),  # noqa: E501
-    "Qwen2_5_VLForConditionalGeneration": ("qwen2_5_vl", "Qwen2_5_VLForConditionalGeneration"),  # noqa: E501
+    "Qwen2_5_VLForConditionalGeneration": (
-    "Qwen2AudioForConditionalGeneration": ("qwen2_audio", "Qwen2AudioForConditionalGeneration"),  # noqa: E501
+        "qwen2_5_vl",
-    "Qwen2_5OmniModel": ("qwen2_5_omni_thinker", "Qwen2_5OmniThinkerForConditionalGeneration"),  # noqa: E501
+        "Qwen2_5_VLForConditionalGeneration",
-    "Qwen2_5OmniForConditionalGeneration": ("qwen2_5_omni_thinker", "Qwen2_5OmniThinkerForConditionalGeneration"),  # noqa: E501
+    ),  # noqa: E501
+    "Qwen2AudioForConditionalGeneration": (
+        "qwen2_audio",
+        "Qwen2AudioForConditionalGeneration",
+    ),  # noqa: E501
+    "Qwen2_5OmniModel": (
+        "qwen2_5_omni_thinker",
+        "Qwen2_5OmniThinkerForConditionalGeneration",
+    ),  # noqa: E501
+    "Qwen2_5OmniForConditionalGeneration": (
+        "qwen2_5_omni_thinker",
+        "Qwen2_5OmniThinkerForConditionalGeneration",
+    ),  # noqa: E501
    "Qwen3VLForConditionalGeneration": ("qwen3_vl", "Qwen3VLForConditionalGeneration"),  # noqa: E501
-    "Qwen3VLMoeForConditionalGeneration": ("qwen3_vl_moe", "Qwen3VLMoeForConditionalGeneration"),  # noqa: E501
+    "Qwen3VLMoeForConditionalGeneration": (
+        "qwen3_vl_moe",
+        "Qwen3VLMoeForConditionalGeneration",
+    ),  # noqa: E501
    "SkyworkR1VChatModel": ("skyworkr1v", "SkyworkR1VChatModel"),
    "Step3VLForConditionalGeneration": ("step3_vl", "Step3VLForConditionalGeneration"),  # noqa: E501
    "TarsierForConditionalGeneration": ("tarsier", "TarsierForConditionalGeneration"),  # noqa: E501
-    "Tarsier2ForConditionalGeneration": ("qwen2_vl", "Tarsier2ForConditionalGeneration"),  # noqa: E501
+    "Tarsier2ForConditionalGeneration": (
+        "qwen2_vl",
+        "Tarsier2ForConditionalGeneration",
+    ),  # noqa: E501
    "UltravoxModel": ("ultravox", "UltravoxModel"),
    "VoxtralForConditionalGeneration": ("voxtral", "VoxtralForConditionalGeneration"),  # noqa: E501
    # [Encoder-decoder]
@@ -324,13 +398,27 @@ _TRANSFORMERS_BACKEND_MODELS = {
    "TransformersForCausalLM": ("transformers", "TransformersForCausalLM"),
    "TransformersForMultimodalLM": ("transformers", "TransformersForMultimodalLM"),  # noqa: E501
    "TransformersMoEForCausalLM": ("transformers_moe", "TransformersMoEForCausalLM"),  # noqa: E501
-    "TransformersMoEForMultimodalLM": ("transformers_moe", "TransformersMoEForMultimodalLM"),  # noqa: E501
+    "TransformersMoEForMultimodalLM": (
-    "TransformersEmbeddingModel": ("transformers_pooling", "TransformersEmbeddingModel"),  # noqa: E501
+        "transformers_moe",
-    "TransformersForSequenceClassification": ("transformers_pooling", "TransformersForSequenceClassification"),  # noqa: E501
+        "TransformersMoEForMultimodalLM",
-    "TransformersMoEForSequenceClassification": ("transformers_pooling", "TransformersMoEForSequenceClassification"),  # noqa: E501
+    ),  # noqa: E501
-    "TransformersMoEEmbeddingModel": ("transformers_pooling", "TransformersMoEEmbeddingModel"),  # noqa: E501
+    "TransformersEmbeddingModel": (
+        "transformers_pooling",
+        "TransformersEmbeddingModel",
+    ),  # noqa: E501
+    "TransformersForSequenceClassification": (
+        "transformers_pooling",
+        "TransformersForSequenceClassification",
+    ),  # noqa: E501
+    "TransformersMoEForSequenceClassification": (
+        "transformers_pooling",
+        "TransformersMoEForSequenceClassification",
+    ),  # noqa: E501
+    "TransformersMoEEmbeddingModel": (
+        "transformers_pooling",
+        "TransformersMoEEmbeddingModel",
+    ),  # noqa: E501
 }
-# yapf: enable
 _VLLM_MODELS = {
    **_TEXT_GENERATION_MODELS,

--- a/vllm/model_executor/models/smolvlm.py
+++ b/vllm/model_executor/models/smolvlm.py
@@ -8,13 +8,10 @@ from transformers import SmolVLMProcessor
 from vllm.config import VllmConfig
 from vllm.multimodal import MULTIMODAL_REGISTRY
-# yapf: disable
 from .idefics3 import Idefics3DummyInputsBuilder as SmolVLMDummyInputsBuilder
 from .idefics3 import Idefics3ForConditionalGeneration, Idefics3ProcessingInfo
 from .idefics3 import Idefics3MultiModalProcessor as SmolVLMMultiModalProcessor
-# yapf: enable
 class SmolVLMProcessingInfo(Idefics3ProcessingInfo):
    def get_hf_processor(self, **kwargs: object) -> SmolVLMProcessor:

--- a/vllm/model_executor/models/voxtral.py
+++ b/vllm/model_executor/models/voxtral.py
@@ -32,11 +32,7 @@ from vllm.model_executor.layers.quantization import QuantizationConfig
 from vllm.model_executor.model_loader.weight_utils import default_weight_loader
 from vllm.model_executor.models import SupportsPP
 from vllm.model_executor.models.module_mapping import MultiModelKeys
-# yapf: disable
 from vllm.model_executor.models.whisper import WhisperEncoder
-# yapf: enable
 from vllm.multimodal import MULTIMODAL_REGISTRY
 from vllm.multimodal.inputs import (
    MultiModalDataDict,

--- a/vllm/transformers_utils/chat_templates/registry.py
+++ b/vllm/transformers_utils/chat_templates/registry.py
@@ -28,7 +28,6 @@ def _get_minicpmv_chat_template_fallback(tokenizer_name_or_path: str) -> Optiona
    return CHAT_TEMPLATES_DIR / "template_chatml.jinja"
-# yapf: disable
 _MODEL_TYPE_TO_CHAT_TEMPLATE_FALLBACK: dict[str, ChatTemplatePath] = {
    "blip-2": CHAT_TEMPLATES_DIR / "template_blip2.jinja",
    "clip": CHAT_TEMPLATES_DIR / "template_basic.jinja",
@@ -39,7 +38,6 @@ _MODEL_TYPE_TO_CHAT_TEMPLATE_FALLBACK: dict[str, ChatTemplatePath] = {
    "paligemma": CHAT_TEMPLATES_DIR / "template_basic.jinja",
    "qwen": _get_qwen_chat_template_fallback,
 }
-# yapf: enable
 def register_chat_template_fallback_path(

--- a/vllm/transformers_utils/configs/arctic.py
+++ b/vllm/transformers_utils/configs/arctic.py
--- a/vllm/transformers_utils/configs/nemotron_vl.py
+++ b/vllm/transformers_utils/configs/nemotron_vl.py
--- a/vllm/transformers_utils/configs/ovis.py
+++ b/vllm/transformers_utils/configs/ovis.py
--- a/vllm/transformers_utils/processors/deepseek_vl2.py
+++ b/vllm/transformers_utils/processors/deepseek_vl2.py
--- a/vllm/transformers_utils/processors/ovis.py
+++ b/vllm/transformers_utils/processors/ovis.py
--- a/vllm/v1/attention/backends/flashinfer.py
+++ b/vllm/v1/attention/backends/flashinfer.py
--- a/vllm/v1/attention/backends/mla/rocm_aiter_mla.py
+++ b/vllm/v1/attention/backends/mla/rocm_aiter_mla.py
--- a/vllm/v1/serial_utils.py
+++ b/vllm/v1/serial_utils.py
--- a/vllm/v1/worker/gpu_model_runner.py
+++ b/vllm/v1/worker/gpu_model_runner.py