Skip to content

[WIP] add has_attentions as done in PyTorch side #16259

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
ydshieh merged 1 commit into from
Mar 19, 2022
Merged

[WIP] add has_attentions as done in PyTorch side #16259

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
7 changes: 2 additions & 5 deletions tests/test_modeling_tf_common.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -114,6 +114,7 @@ class TFModelTesterMixin:
test_resize_embeddings = True
test_head_masking = True
is_encoder_decoder = False
has_attentions = True

def _prepare_for_class(self, inputs_dict, model_class, return_labels=False) -> dict:
inputs_dict = copy.deepcopy(inputs_dict)
Expand Down Expand Up @@ -539,9 +540,7 @@ def check_pt_tf_models(tf_model, pt_model):

# Output all for aggressive testing
config.output_hidden_states = True
# Pure convolutional models have no attention
# TODO: use a better and general criteria
if "TFConvNext" not in model_class.__name__:
if self.has_attentions:
config.output_attentions = True

for k in ["attention_mask", "encoder_attention_mask", "decoder_attention_mask"]:
Expand All @@ -567,8 +566,6 @@ def check_pt_tf_models(tf_model, pt_model):
pt_model_class_name = model_class.__name__[2:] # Skip the "TF" at the beginning
pt_model_class = getattr(transformers, pt_model_class_name)

config.output_hidden_states = True

tf_model = model_class(config)
pt_model = pt_model_class(config)

Expand Down