SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

Fix tests after huggingface_hub 0.24 (#32054)

Browse Source 
* adapt tests

* style

* comment
This commit is contained in:
Lucain
2024-07-19 20:32:39 +02:00
committed by GitHub
parent fe008d6ebe
commit 0fdea8607d
1 changed files with 36 additions and 23 deletions
Download Patch File Download Diff File Expand all files Collapse all files

21
tests/trainer/test_trainer.py
View File

@@ -3739,6 +3739,7 @@ class TrainerIntegrationWithHubTester(unittest.TestCase):
def test_push_to_hub_with_saves_each_epoch(self): def test_push_to_hub_with_saves_each_epoch(self):
with tempfile.TemporaryDirectory() as tmp_dir: with tempfile.TemporaryDirectory() as tmp_dir:
with self.assertLogs(level="WARNING") as logs:
trainer = get_regression_trainer( trainer = get_regression_trainer(
output_dir=os.path.join(tmp_dir, "test-trainer-epoch"), output_dir=os.path.join(tmp_dir, "test-trainer-epoch"),
push_to_hub=True, push_to_hub=True,
@@ -3752,8 +3753,10 @@ class TrainerIntegrationWithHubTester(unittest.TestCase):
commits = list_repo_commits(f"{USER}/test-trainer-epoch", token=self._token) commits = list_repo_commits(f"{USER}/test-trainer-epoch", token=self._token)
commits = [c.title for c in commits] commits = [c.title for c in commits]
self.assertIn("initial commit", commits) self.assertIn("initial commit", commits)
for i in range(1, 4): self.assertIn("Training in progress, epoch 1", commits)
self.assertIn(f"Training in progress, epoch {i}", commits) self.assertIn("Training in progress, epoch 2", commits)
# Epochs 3 and 4 are not guaranteed to be present (empty commits)
self.assertTrue(any("Skipping to prevent empty commit." in record.message for record in logs.records))
def test_push_to_hub_with_saves_each_n_steps(self): def test_push_to_hub_with_saves_each_n_steps(self):
num_gpus = max(1, backend_device_count(torch_device)) num_gpus = max(1, backend_device_count(torch_device))
@@ -3761,6 +3764,7 @@ class TrainerIntegrationWithHubTester(unittest.TestCase):
self.skipTest(reason="More than 2 GPUs available") self.skipTest(reason="More than 2 GPUs available")
with tempfile.TemporaryDirectory() as tmp_dir: with tempfile.TemporaryDirectory() as tmp_dir:
with self.assertLogs(level="WARNING") as logs:
trainer = get_regression_trainer( trainer = get_regression_trainer(
output_dir=os.path.join(tmp_dir, "test-trainer-step"), output_dir=os.path.join(tmp_dir, "test-trainer-step"),
push_to_hub=True, push_to_hub=True,
@@ -3776,10 +3780,19 @@ class TrainerIntegrationWithHubTester(unittest.TestCase):
commits = [c.title for c in commits] commits = [c.title for c in commits]
self.assertIn("initial commit", commits) self.assertIn("initial commit", commits)
# Some commits are skipped if nothing has changed
# We expect 1 commit per 5 epochs + 1 commit at the end
nb_empty_commits = len(
[record for record in logs.records if "Skipping to prevent empty commit." in record.message]
)
nb_epoch_commits = len([commit for commit in commits if "Training in progress, step" in commit])
# max_steps depend on the number of available GPUs # max_steps depend on the number of available GPUs
max_steps = math.ceil(trainer.args.num_train_epochs * len(trainer.get_train_dataloader())) max_steps = math.ceil(trainer.args.num_train_epochs * len(trainer.get_train_dataloader()))
for i in range(5, max_steps, 5): nb_expected_commits = len(range(5, max_steps, 5))
self.assertIn(f"Training in progress, step {i}", commits)
# '>=' since final commit might be an empty commit as well (not deterministic)
self.assertGreaterEqual(nb_empty_commits + nb_epoch_commits, nb_expected_commits)
@require_tensorboard @require_tensorboard
def test_push_to_hub_with_tensorboard_logs(self): def test_push_to_hub_with_tensorboard_logs(self):