Skip to content

Fix loss computation in TFWav2Vec2ForCTC #18014

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

Merged
merged 1 commit into from
Jul 4, 2022
Merged

Fix loss computation in TFWav2Vec2ForCTC #18014

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
11 changes: 10 additions & 1 deletion src/transformers/models/wav2vec2/modeling_tf_wav2vec2.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -25,7 +25,13 @@

from ...activations_tf import get_tf_activation
from ...modeling_tf_outputs import TFBaseModelOutput, TFCausalLMOutput
from ...modeling_tf_utils import TFPreTrainedModel, booleans_processing, get_initializer, keras_serializable
from ...modeling_tf_utils import (
TFPreTrainedModel,
booleans_processing,
get_initializer,
keras_serializable,
unpack_inputs,
)
from ...tf_utils import shape_list, stable_softmax
from ...utils import (
ModelOutput,
Expand Down Expand Up @@ -1580,6 +1586,7 @@ def freeze_feature_encoder(self):
"""
self.wav2vec2.feature_extractor.trainable = False

@unpack_inputs
@add_start_docstrings_to_model_forward(WAV_2_VEC_2_INPUTS_DOCSTRING)
@replace_return_docstrings(output_type=TFCausalLMOutput, config_class=_CONFIG_FOR_DOC)
def call(
Expand Down Expand Up @@ -1702,6 +1709,8 @@ def call(
loss = tf.reduce_sum(loss)
if self.config.ctc_loss_reduction == "mean":
loss = tf.reduce_mean(loss)

loss = tf.reshape(loss, (1,))
else:
loss = None

Expand Down