Skip to content

Instantly share code, notes, and snippets.

@eyaltrabelsi

eyaltrabelsi/test_sanity_metric_training.py

Last active June 21, 2024 10:55
Show Gist options
  • Save eyaltrabelsi/946b64f93d24a32e4875cc4400ffd17f to your computer and use it in GitHub Desktop.
Save eyaltrabelsi/946b64f93d24a32e4875cc4400ffd17f to your computer and use it in GitHub Desktop.
Download ZIP
Raw
test_sanity_metric_training.py
def test_sanity_metric_training(capfd):
input_df = get_production_sample(pct=0.1)
features_df = feature_engineering(input_df)
model = train_model(features_data)
stdout_of_training = capfd.readouterr()
training_losses_per_epoch = parse_training_loss_from_logs(stdout_of_training)
assert training_losses_per_epoch == sorted(training_losses_per_epoch), "training loss must is monotoncly decreasing"
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment