Skip to content

Fix audio pipelines #42920

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
zucchini-nlp wants to merge 1 commit into
base: main
Choose a base branch
from
+8 −12
Open

Fix audio pipelines #42920

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 0 additions & 1 deletion src/transformers/models/wav2vec2_bert/convert_wav2vec2_seamless_checkpoint.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -150,7 +150,6 @@ def convert_wav2vec2_bert_checkpoint(

# save feature extractor
fe = SeamlessM4TFeatureExtractor(padding_value=1)
fe._processor_class = "Wav2Vec2BertProcessor"
fe.save_pretrained(pytorch_dump_folder_path)

if repo_id:
Expand Down
13 changes: 7 additions & 6 deletions src/transformers/pipelines/__init__.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -22,7 +22,7 @@

from ..configuration_utils import PreTrainedConfig
from ..dynamic_module_utils import get_class_from_dynamic_module
from ..feature_extraction_utils import PreTrainedFeatureExtractor
from ..feature_extraction_utils import FeatureExtractionMixin, PreTrainedFeatureExtractor
from ..image_processing_utils import BaseImageProcessor
from ..models.auto.configuration_auto import AutoConfig
from ..models.auto.feature_extraction_auto import FEATURE_EXTRACTOR_MAPPING, AutoFeatureExtractor
Expand Down Expand Up @@ -986,12 +986,13 @@ def pipeline(
feature_extractor = AutoFeatureExtractor.from_pretrained(
feature_extractor, _from_pipeline=task, **hub_kwargs, **model_kwargs
)
config_dict, _ = FeatureExtractionMixin.get_feature_extractor_dict(
pretrained_model_name_or_path,
**hub_kwargs,
)
processor_class = config_dict.get("processor_class", None)

if (
feature_extractor._processor_class
and feature_extractor._processor_class.endswith("WithLM")
and isinstance(model_name, str)
):
if processor_class is not None and processor_class.endswith("WithLM") and isinstance(model_name, str):
try:
import kenlm # to trigger `ImportError` if not installed
from pyctcdecode import BeamSearchDecoderCTC
Expand Down
6 changes: 1 addition & 5 deletions src/transformers/pipelines/automatic_speech_recognition.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -198,11 +198,7 @@ def __init__(
self.type = "seq2seq_whisper"
elif model.__class__.__name__ in MODEL_FOR_SPEECH_SEQ_2_SEQ_MAPPING_NAMES.values():
self.type = "seq2seq"
elif (
feature_extractor._processor_class
and feature_extractor._processor_class.endswith("WithLM")
and decoder is not None
):
elif decoder is not None:
self.decoder = decoder
self.type = "ctc_with_lm"
else:
Expand Down