[Rebase] Rebase to vllm v0.19.0 (#2475)

This commit is contained in:
Zhou Taichang authored and GitHub committed 2026-04-04 15:03:32 +08:00
1 parent 4c031580cf
commit 2804a85e58
52 files changed
+437 -142

No files matched your search

@@ -8,6 +8,7 @@ import torch.nn as nn
from transformers import BatchFeature
from vllm.config import VllmConfig
from vllm.config.multimodal import BaseDummyOptions
from vllm.inputs import MultiModalDataDict
from vllm.model_executor.layers.layernorm import RMSNorm as VllmRMSNorm
from vllm.model_executor.layers.linear import (
QKVParallelLinear,
@@ -20,7 +21,6 @@ from vllm.model_executor.models.qwen2 import Qwen2DecoderLayer, Qwen2MLP
from vllm.model_executor.models.utils import AutoWeightsLoader
from vllm.multimodal import MULTIMODAL_REGISTRY
from vllm.multimodal.inputs import (
MultiModalDataDict,
MultiModalFieldConfig,
MultiModalKwargsItems,
)
@@ -10,10 +10,11 @@ import torch.nn as nn
from transformers.feature_extraction_utils import BatchFeature
from vllm.config import VllmConfig
from vllm.config.multimodal import BaseDummyOptions
from vllm.inputs import MultiModalDataDict
from vllm.logger import init_logger
from vllm.model_executor.models.interfaces import SupportsMultiModal
from vllm.multimodal import MULTIMODAL_REGISTRY
from vllm.multimodal.inputs import MultiModalDataDict, MultiModalFieldConfig, MultiModalKwargsItems
from vllm.multimodal.inputs import MultiModalFieldConfig, MultiModalKwargsItems
from vllm.multimodal.parse import MultiModalDataItems, MultiModalDataParser
from vllm.multimodal.processing import (
BaseDummyInputsBuilder,
@@ -40,6 +40,7 @@ from vllm.config import CacheConfig, MultiModalConfig, VllmConfig
from vllm.config.multimodal import BaseDummyOptions
from vllm.distributed import get_pp_group, get_tensor_model_parallel_world_size
from vllm.distributed import utils as dist_utils
from vllm.inputs import MultiModalDataDict
from vllm.logger import init_logger
from vllm.model_executor.layers.attention import Attention
from vllm.model_executor.layers.attention.mm_encoder_attention import (
@@ -73,7 +74,6 @@ from vllm.model_executor.models.utils import (
from vllm.model_executor.models.vision import get_vit_attn_backend
from vllm.multimodal import MULTIMODAL_REGISTRY
from vllm.multimodal.inputs import (
MultiModalDataDict,
MultiModalFeatureSpec,
MultiModalFieldConfig,
MultiModalKwargsItems,
@@ -20,6 +20,7 @@ from vllm.compilation.decorators import support_torch_compile
from vllm.config import VllmConfig
from vllm.config.multimodal import BaseDummyOptions
from vllm.distributed import get_pp_group
from vllm.inputs import MultiModalDataDict
from vllm.logger import init_logger
from vllm.model_executor.layers.fused_moe import SharedFusedMoE
from vllm.model_executor.layers.linear import (
@@ -58,7 +59,6 @@ from vllm.model_executor.models.utils import (
from vllm.multimodal import MULTIMODAL_REGISTRY
from vllm.multimodal.image import rgba_to_rgb
from vllm.multimodal.inputs import (
MultiModalDataDict,
MultiModalFeatureSpec,
MultiModalFieldConfig,
MultiModalKwargsItems,
@@ -10,6 +10,7 @@ from torch import nn
from transformers import BatchFeature, Qwen2Config
from vllm.config import VllmConfig
from vllm.config.multimodal import BaseDummyOptions
from vllm.inputs import ModalityData, MultiModalDataDict
from vllm.logger import init_logger
from vllm.model_executor.layers.logits_processor import LogitsProcessor
from vllm.model_executor.models import SupportsPP
@@ -18,9 +19,7 @@ from vllm.model_executor.models.utils import init_vllm_registered_model
from vllm.multimodal import MULTIMODAL_REGISTRY
from vllm.multimodal.inputs import (
AudioItem,
ModalityData,
MultiModalBatchedField,
MultiModalDataDict,
MultiModalFieldConfig,
MultiModalFieldElem,
MultiModalKwargsItem,
@@ -13,6 +13,7 @@ from transformers.models.qwen2.modeling_qwen2 import (
)
from vllm.config import VllmConfig
from vllm.forward_context import get_forward_context
from vllm.inputs import MultiModalDataDict
from vllm.model_executor.layers.linear import ColumnParallelLinear
from vllm.model_executor.layers.logits_processor import LogitsProcessor
from vllm.model_executor.model_loader.weight_utils import default_weight_loader, maybe_remap_kv_scale_name
@@ -34,7 +35,6 @@ from vllm.model_executor.models.utils import (
)
from vllm.multimodal import MULTIMODAL_REGISTRY
from vllm.multimodal.inputs import (
MultiModalDataDict,
MultiModalFieldConfig,
MultiModalKwargsItems,
)
@@ -18,7 +18,7 @@ from transformers.models.qwen3_omni_moe.configuration_qwen3_omni_moe import (
Qwen3OmniMoeThinkerConfig,
)
from vllm.config import ModelConfig, VllmConfig
from vllm.inputs.data import PromptType, TokensPrompt
from vllm.inputs import PromptType, TokensPrompt
from vllm.logger import init_logger
from vllm.model_executor.layers.rotary_embedding import MRotaryEmbedding
from vllm.model_executor.models.interfaces import SupportsMRoPE, SupportsMultiModal, SupportsPP, SupportsRealtime
@@ -46,7 +46,7 @@ from transformers.models.whisper import WhisperFeatureExtractor
from vllm.compilation.decorators import support_torch_compile
from vllm.config import ModelConfig, SpeechToTextConfig, VllmConfig
from vllm.distributed import get_pp_group, get_tensor_model_parallel_world_size
from vllm.inputs.data import PromptType
from vllm.inputs import PromptType
from vllm.logger import init_logger
from vllm.model_executor.layers.activation import _ACTIVATION_REGISTRY
from vllm.model_executor.layers.attention.mm_encoder_attention import (
@@ -29,6 +29,7 @@ from mistral_common.tokens.tokenizers.audio import Audio, AudioEncoder
from transformers import BatchFeature
from transformers.tokenization_utils_base import TextInput
from vllm.config import VllmConfig
from vllm.inputs import MultiModalDataDict
from vllm.logger import init_logger
from vllm.model_executor.model_loader.weight_utils import default_weight_loader
from vllm.model_executor.models.interfaces import SupportsMultiModal
@@ -39,7 +40,6 @@ from vllm.model_executor.models.utils import (
)
from vllm.multimodal import MULTIMODAL_REGISTRY
from vllm.multimodal.inputs import (
MultiModalDataDict,
MultiModalFieldConfig,
MultiModalKwargsItems,
NestedTensors,