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

set min seq len by default #621

Merged
merged 1 commit into from
Jan 18, 2025
Merged

set min seq len by default #621

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
6 changes: 3 additions & 3 deletions sub-packages/bionemo-esm2/src/bionemo/esm2/data/datamodule.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -65,8 +65,8 @@ def __init__(
valid_cluster_path: A path to the parquet files containing UniRef50 validation clusters.
valid_database_path: A path to the sqlite file mapping UniRef50 cluster IDs to sequences.
seed: Input random seed. If None, initializes randomly. Defaults to 42.
min_seq_length: Whether to pad sequences to a minimum length. If None, no extra padding is added. Defaults
to None.
min_seq_length: Whether to pad sequences to a minimum length. If None, sequences are padded to the maximum
sequence length. Defaults to None.
max_seq_length: The maximum context length for the ESM transformer. Defaults to 1024.
micro_batch_size: Passed to MegatronDataSampler. Defaults to 4.
global_batch_size: Passed to MegatronDataSampler.. Defaults to 8.
Expand All @@ -87,7 +87,7 @@ def __init__(
self._valid_cluster_path = valid_cluster_path
self._valid_database_path = valid_database_path
self._seed = seed
self._min_seq_length = min_seq_length
self._min_seq_length = min_seq_length if min_seq_length is not None else max_seq_length
self._max_seq_length = max_seq_length
self._mask_prob = mask_prob
self._mask_token_prob = mask_token_prob
Expand Down
15 changes: 15 additions & 0 deletions sub-packages/bionemo-esm2/tests/bionemo/esm2/data/test_datamodule.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -39,6 +39,21 @@ def test_create_esm_datamodule_raises_without_trainer(dummy_protein_dataset, dum
data_module.setup()


def test_esm_datamodule_sets_min_seq_len_to_max_seq_len(dummy_protein_dataset, dummy_parquet_train_val_inputs):
train_cluster_path, valid_cluster_path = dummy_parquet_train_val_inputs

# Initialize the data module.
data_module = ESMDataModule(
train_cluster_path=train_cluster_path,
train_database_path=dummy_protein_dataset,
valid_cluster_path=valid_cluster_path,
valid_database_path=dummy_protein_dataset,
max_seq_length=36,
)

assert data_module._min_seq_length == 36


def test_create_esm_datamodule_raises_without_trainer_max_steps(dummy_protein_dataset, dummy_parquet_train_val_inputs):
train_cluster_path, valid_cluster_path = dummy_parquet_train_val_inputs

Expand Down
Loading