Skip to content
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

Jump to bottom

ViViTMultiHeadClassificationModel - Add support for tasks with less than 5 classes. #1056

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open

ViViTMultiHeadClassificationModel - Add support for tasks with less than 5 classes. #1056

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
20 changes: 13 additions & 7 deletions scenic/projects/vivit/model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -916,6 +916,9 @@ def get_metrics_fn(self, split: Optional[str] = None) -> base_model.MetricFn:
"""
del split # for all splits, we return the same metric functions

num_classes_in_each_head = (
self.dataset_meta_data.get('class_splits', [-1]))
minimal_num_classes = min(num_classes_in_each_head)
def classification_metrics_function(logits, batch, metrics, class_splits,
split_names):

Expand Down Expand Up @@ -947,18 +950,21 @@ def classification_metrics_function(logits, batch, metrics, class_splits,
(model_utils.joint_accuracy(logits, one_hot_targets, class_splits,
weights),
base_model_utils.num_examples(logits, one_hot_targets, weights)))
pairwise_top_five = base_model_utils.psum_metric_normalizer(
(model_utils.joint_top_k(
logits, one_hot_targets, class_splits, k=5, weights=weights),
base_model_utils.num_examples(logits, one_hot_targets, weights)))
eval_name = f'{split_names[0]}-{split_names[1]}'
evaluated_metrics[f'{eval_name}_accuracy'] = pairwise_acc
evaluated_metrics[f'{eval_name}_accuracy_top_5'] = pairwise_top_five
if minimal_num_classes > 5:
pairwise_top_five = base_model_utils.psum_metric_normalizer(
(model_utils.joint_top_k(
logits, one_hot_targets, class_splits, k=5, weights=weights),
base_model_utils.num_examples(logits, one_hot_targets, weights)))
evaluated_metrics[f'{eval_name}_accuracy_top_5'] = pairwise_top_five

return evaluated_metrics

metrics = ViViT_CLASSIFICATION_METRICS
if minimal_num_classes <= 5:
metrics = ViViT_CLASSIFICATION_METRICS_BASIC
return functools.partial(
classification_metrics_function,
metrics=ViViT_CLASSIFICATION_METRICS,
metrics=metrics,
class_splits=self.class_splits,
split_names=self.split_names)