frozenleaves/transformers
1
0
Fork 0
mirror of https://github.com/huggingface/transformers.git synced 2025-10-20 17:13:56 +08:00
Code Issues Packages Projects Releases Wiki Activity

[serve] guard imports (#39825)

Browse Source 
guard imports
This commit is contained in:
Joao Gante
2025-08-18 16:28:10 +01:00
committed by GitHub
parent 7a0ba0d7d8
commit d6fad86d23
1 changed files with 11 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
src/transformers/commands/serving.py
View File

@ -32,7 +32,6 @@ from typing import Optional, Union
from huggingface_hub import model_info
from huggingface_hub.constants import HF_HUB_OFFLINE
from PIL import Image
import transformers
from transformers.models.auto.modeling_auto import (
@ -45,6 +44,7 @@ from transformers.utils.import_utils import (
is_openai_available,
is_pydantic_available,
is_uvicorn_available,
is_vision_available,
)
from .. import (
@ -54,7 +54,6 @@ from .. import (
ProcessorMixin,
TextIteratorStreamer,
)
from ..generation.continuous_batching import ContinuousBatchingManager, RequestStatus
from ..utils import is_torch_available, logging
from . import BaseTransformersCLICommand
@ -69,9 +68,14 @@ if is_torch_available():
PreTrainedModel,
)
from ..generation.continuous_batching import ContinuousBatchingManager, RequestStatus
if is_librosa_available():
import librosa
if is_vision_available():
from PIL import Image
serve_dependencies_available = (
is_pydantic_available() and is_fastapi_available() and is_uvicorn_available() and is_openai_available()
)
@ -811,7 +815,7 @@ class ServeCommand(BaseTransformersCLICommand):
return stream_chat_completion(inputs[0])
@staticmethod
def get_model_modality(model: PreTrainedModel) -> Modality:
def get_model_modality(model: "PreTrainedModel") -> Modality:
model_classname = model.__class__.__name__
if model_classname in MODEL_FOR_IMAGE_TEXT_TO_TEXT_MAPPING_NAMES.values():
modality = Modality.VLM
@ -1545,7 +1549,9 @@ class ServeCommand(BaseTransformersCLICommand):
logger.info(f"Loaded model {model_id_and_revision}")
return model, data_processor
def load_model_and_processor(self, model_id_and_revision: str) -> tuple[PreTrainedModel, PreTrainedTokenizerFast]:
def load_model_and_processor(
self, model_id_and_revision: str
) -> tuple["PreTrainedModel", PreTrainedTokenizerFast]:
"""
Loads the text model and processor from the given model ID and revision into the ServeCommand instance.
@ -1570,7 +1576,7 @@ class ServeCommand(BaseTransformersCLICommand):
return model, processor
def load_audio_model_and_processor(self, model_id_and_revision: str) -> tuple[PreTrainedModel, ProcessorMixin]:
def load_audio_model_and_processor(self, model_id_and_revision: str) -> tuple["PreTrainedModel", ProcessorMixin]:
"""
Loads the audio model and processor from the given model ID and revision into the ServeCommand instance.