biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Misc] Rename MultiModalInputsV2 -> MultiModalInputs (#12244)

Browse Source 
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
This commit is contained in:
Cyrus Leung
2025-01-21 15:31:19 +08:00
committed by GitHub
parent 2fc6944c5e
commit 96912550c8
12 changed files with 31 additions and 31 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
vllm/model_executor/models/blip2.py
View File

@@ -15,7 +15,7 @@ from vllm.model_executor.layers.sampler import SamplerOutput, get_sampler
from vllm.model_executor.sampling_metadata import SamplingMetadata
from vllm.multimodal import MULTIMODAL_REGISTRY
from vllm.multimodal.inputs import (MultiModalDataDict, MultiModalFieldConfig,
MultiModalInputsV2, MultiModalKwargs,
MultiModalInputs, MultiModalKwargs,
NestedTensors, PlaceholderRange)
from vllm.multimodal.parse import MultiModalDataItems
from vllm.multimodal.processing import (BaseMultiModalProcessor,
@@ -490,7 +490,7 @@ class Blip2MultiModalProcessor(BaseMultiModalProcessor[Blip2ProcessingInfo]):
prompt: Union[str, list[int]],
mm_data: MultiModalDataDict,
hf_processor_mm_kwargs: Mapping[str, object],
) -> MultiModalInputsV2:
) -> MultiModalInputs:
result = super().apply(prompt, mm_data, hf_processor_mm_kwargs)
# Only <image> tokens should be considered as placeholders,

4
vllm/model_executor/models/chameleon.py
View File

@@ -29,7 +29,7 @@ from vllm.model_executor.sampling_metadata import SamplingMetadata
from vllm.model_executor.utils import set_weight_attrs
from vllm.multimodal import MULTIMODAL_REGISTRY
from vllm.multimodal.inputs import (MultiModalDataDict, MultiModalFieldConfig,
MultiModalInputsV2, MultiModalKwargs,
MultiModalInputs, MultiModalKwargs,
NestedTensors, PlaceholderRange)
from vllm.multimodal.parse import MultiModalDataItems
from vllm.multimodal.processing import (BaseMultiModalProcessor,
@@ -159,7 +159,7 @@ class ChameleonMultiModalProcessor(
prompt: Union[str, list[int]],
mm_data: MultiModalDataDict,
hf_processor_mm_kwargs: Mapping[str, object],
) -> MultiModalInputsV2:
) -> MultiModalInputs:
result = super().apply(prompt, mm_data, hf_processor_mm_kwargs)
# Only <image> tokens should be considered as placeholders,

4
vllm/model_executor/models/fuyu.py
View File

@@ -31,7 +31,7 @@ from vllm.model_executor.models.persimmon import PersimmonForCausalLM
from vllm.model_executor.sampling_metadata import SamplingMetadata
from vllm.multimodal import MULTIMODAL_REGISTRY
from vllm.multimodal.inputs import (MultiModalDataDict, MultiModalFieldConfig,
MultiModalInputsV2, MultiModalKwargs,
MultiModalInputs, MultiModalKwargs,
NestedTensors, PlaceholderRange)
from vllm.multimodal.parse import (ImageProcessorItems, ImageSize,
MultiModalDataItems)
@@ -232,7 +232,7 @@ class FuyuMultiModalProcessor(BaseMultiModalProcessor[FuyuProcessingInfo]):
prompt: Union[str, list[int]],
mm_data: MultiModalDataDict,
hf_processor_mm_kwargs: Mapping[str, object],
) -> MultiModalInputsV2:
) -> MultiModalInputs:
result = super().apply(prompt, mm_data, hf_processor_mm_kwargs)
# Only |SPEAKER| (image) tokens should be considered as placeholders,

6
vllm/model_executor/models/llava.py
View File

@@ -24,7 +24,7 @@ from vllm.model_executor.layers.sampler import SamplerOutput, get_sampler
from vllm.model_executor.sampling_metadata import SamplingMetadata
from vllm.multimodal import MULTIMODAL_REGISTRY
from vllm.multimodal.inputs import (MultiModalDataDict, MultiModalFieldConfig,
MultiModalInputsV2, MultiModalKwargs,
MultiModalInputs, MultiModalKwargs,
NestedTensors)
from vllm.multimodal.parse import (ImageEmbeddingItems, ImageProcessorItems,
ImageSize, MultiModalDataItems)
@@ -746,7 +746,7 @@ class MantisMultiModalProcessor(LlavaMultiModalProcessor):
prompt: Union[str, list[int]],
mm_data: MultiModalDataDict,
hf_processor_mm_kwargs: Mapping[str, object],
) -> MultiModalInputsV2:
) -> MultiModalInputs:
hf_config = self.info.get_hf_config()
image_token_id = hf_config.image_token_index
@@ -805,7 +805,7 @@ class MantisMultiModalProcessor(LlavaMultiModalProcessor):
for modality, placeholders in mm_placeholders.items()
}
return MultiModalInputsV2(
return MultiModalInputs(
type="multimodal",
prompt=prompt,
prompt_token_ids=prompt_ids,

4
vllm/model_executor/models/phi3v.py
View File

@@ -31,7 +31,7 @@ from vllm.model_executor.layers.vocab_parallel_embedding import (
from vllm.model_executor.sampling_metadata import SamplingMetadata
from vllm.multimodal import MULTIMODAL_REGISTRY
from vllm.multimodal.inputs import (MultiModalDataDict, MultiModalFieldConfig,
MultiModalInputsV2, MultiModalKwargs,
MultiModalInputs, MultiModalKwargs,
NestedTensors, PlaceholderRange)
from vllm.multimodal.parse import (ImageEmbeddingItems, ImageProcessorItems,
ImageSize, MultiModalDataItems)
@@ -484,7 +484,7 @@ class Phi3VMultiModalProcessor(BaseMultiModalProcessor[Phi3VProcessingInfo]):
prompt: Union[str, list[int]],
mm_data: MultiModalDataDict,
hf_processor_mm_kwargs: Mapping[str, object],
) -> MultiModalInputsV2:
) -> MultiModalInputs:
result = super().apply(prompt, mm_data, hf_processor_mm_kwargs)
# Only <|image|> tokens should be considered as placeholders,

4
vllm/model_executor/models/qwen2_audio.py
View File

@@ -37,7 +37,7 @@ from vllm.model_executor.layers.sampler import SamplerOutput, get_sampler
from vllm.model_executor.sampling_metadata import SamplingMetadata
from vllm.multimodal import MULTIMODAL_REGISTRY
from vllm.multimodal.inputs import (MultiModalDataDict, MultiModalFieldConfig,
MultiModalInputsV2, MultiModalKwargs,
MultiModalInputs, MultiModalKwargs,
NestedTensors, PlaceholderRange)
from vllm.multimodal.parse import (AudioProcessorItems, MultiModalDataItems,
MultiModalDataParser)
@@ -245,7 +245,7 @@ class Qwen2AudioMultiModalProcessor(
prompt: Union[str, list[int]],
mm_data: MultiModalDataDict,
hf_processor_mm_kwargs: Mapping[str, object],
) -> MultiModalInputsV2:
) -> MultiModalInputs:
result = super().apply(prompt, mm_data, hf_processor_mm_kwargs)
# Only <|AUDIO|> tokens should be considered as placeholders,