Skip to content

Using synthetic dataset_type for more cases #3591

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

Open
charlesli640 wants to merge 1 commit into
base: main
Choose a base branch
from
+4 −4
Open

Using synthetic dataset_type for more cases #3591

Show file tree
Hide file tree
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
4 changes: 2 additions & 2 deletions tests/integration/simple_decoder_layer_test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -35,7 +35,7 @@ def test_simple_decoder_layer(self):
get_test_config_path(),
"base_output_directory=gs://runner-maxtext-logs",
"run_name=runner_simple_decoder_layer_test",
"dataset_path=gs://maxtext-dataset",
"dataset_type=synthetic",
"decoder_block=simple",
"enable_checkpointing=False",
"enable_goodput_recording=False",
Expand All @@ -52,7 +52,7 @@ def test_mlp_decoder_layer(self):
get_test_config_path(),
"base_output_directory=gs://runner-maxtext-logs",
"run_name=runner_simple_decoder_layer_test",
"dataset_path=gs://maxtext-dataset",
"dataset_type=synthetic",
"decoder_block=simple_mlp",
"enable_checkpointing=False",
"enable_goodput_recording=False",
Expand Down
4 changes: 2 additions & 2 deletions tests/integration/train_tests.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -225,7 +225,7 @@ def test_tpu_base(self):
@pytest.mark.integration_test
@pytest.mark.tpu_only
def test_tpu_tokamax(self):
train_main(TrainTests.CONFIGS["base"] + ["use_tokamax_splash=true"])
train_main(TrainTests.CONFIGS["synthetic"] + ["use_tokamax_splash=true"])
Copy link
Copy Markdown
Collaborator

@bvandermoon bvandermoon Apr 7, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why does the synthetic config set dataset_path?

Copy link
Copy Markdown
Collaborator Author

@charlesli640 charlesli640 Apr 8, 2026 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

dataset_path will be ignored if dataset_type is synthetic. As the data will be actually created in memory instead of loaded from the path. I guess it was set by copy-and-paste. But it does no harm.

Copy link
Copy Markdown
Collaborator

@bvandermoon bvandermoon Apr 8, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks @charlesli640. Mind removing dataset_path from the synthetic one for clarity?


@pytest.mark.integration_test
@pytest.mark.gpu_only
Expand Down Expand Up @@ -490,7 +490,7 @@ def test_gpu_cudnn_flash_jax(self):

@pytest.mark.integration_test
def test_base_model_shardy_false(self):
train_main(TrainTests.CONFIGS["base"] + ["shardy=False"])
train_main(TrainTests.CONFIGS["synthetic"] + ["shardy=False"])

@pytest.mark.integration_test
@pytest.mark.tpu_only
Expand Down
Loading