Skip to content

A lightweight yet powerful audio-to-MIDI converter with pitch bend detection

Split enum

Benjie Genchel435b5fa4a114
diff --git a/basic_pitch/constants.py b/basic_pitch/constants.py
index 0123dec..69a2020 100644
--- a/basic_pitch/constants.py
+++ b/basic_pitch/constants.py
@@ -62,4 +62,4 @@ def _freq_bins(bins_per_semitone: int, base_frequency: float, n_semitones: int)
FREQ_BINS_NOTES = _freq_bins(NOTES_BINS_PER_SEMITONE, ANNOTATIONS_BASE_FREQUENCY, ANNOTATIONS_N_SEMITONES)
FREQ_BINS_CONTOURS = _freq_bins(CONTOURS_BINS_PER_SEMITONE, ANNOTATIONS_BASE_FREQUENCY, ANNOTATIONS_N_SEMITONES)
-Splits = Enum("Splits", ["train", "validation", "test"])
+Split = Enum("Split", ["train", "validation", "test"])
diff --git a/basic_pitch/data/tf_example_deserialization.py b/basic_pitch/data/tf_example_deserialization.py
index ffaeb73..b59667d 100644
--- a/basic_pitch/data/tf_example_deserialization.py
+++ b/basic_pitch/data/tf_example_deserialization.py
@@ -33,6 +33,7 @@ from basic_pitch.constants import (
AUDIO_WINDOW_LENGTH,
N_FREQ_BINS_NOTES,
N_FREQ_BINS_CONTOURS,
+ Split,
)
N_SAMPLES_PER_TRACK = 20
@@ -59,13 +60,13 @@ def prepare_datasets(
# init both
ds_train = sample_datasets(
- "train",
+ Split.train,
datasets_base_path,
datasets=datasets_to_use,
dataset_sampling_frequency=dataset_sampling_frequency,
)
ds_validation = sample_datasets(
- "validation",
+ Split.validation,
datasets_base_path,
datasets=datasets_to_use,
dataset_sampling_frequency=dataset_sampling_frequency,
@@ -118,14 +119,14 @@ def prepare_visualization_datasets(
assert validation_steps is not None and validation_steps > 0
ds_train = sample_datasets(
- "train",
+ Split.train,
datasets_base_path,
datasets=datasets_to_use,
dataset_sampling_frequency=dataset_sampling_frequency,
n_samples_per_track=1,
)
ds_validation = sample_datasets(
- "validation",
+ Split.validation,
datasets_base_path,
datasets=datasets_to_use,
dataset_sampling_frequency=dataset_sampling_frequency,
@@ -153,7 +154,7 @@ def prepare_visualization_datasets(
def sample_datasets(
- split: str,
+ split: Split,
datasets_base_path: str,
datasets: List[str],
dataset_sampling_frequency: np.ndarray,
@@ -161,8 +162,7 @@ def sample_datasets(
n_samples_per_track: int = N_SAMPLES_PER_TRACK,
pairs: bool = False,
) -> tf.data.Dataset:
- assert split in ["train", "validation"]
- if split == "validation":
+ if split == Split.validation:
n_shuffle = 0
pairs = False
if n_samples_per_track != 1:
@@ -209,7 +209,7 @@ def sample_datasets(
def transcription_file_generator(
- split: str,
+ split: Split,
dataset_names: List[str],
datasets_base_path: str,
sample_weights: np.ndarray,
@@ -219,12 +219,12 @@ def transcription_file_generator(
"""
file_dict = {
dataset_name: tf.data.Dataset.list_files(
- os.path.join(datasets_base_path, dataset_name, "splits", split, "*tfrecord")
+ os.path.join(datasets_base_path, dataset_name, "splits", split.name, "*tfrecord")
)
for dataset_name in dataset_names
}
- if split == "train":
+ if split == Split.train:
return lambda: _train_file_generator(file_dict, sample_weights), False
return lambda: _validation_file_generator(file_dict), True
diff --git a/tests/data/test_tf_example_deserialization.py b/tests/data/test_tf_example_deserialization.py
index 4a2fd0e..379f5bc 100644
--- a/tests/data/test_tf_example_deserialization.py
+++ b/tests/data/test_tf_example_deserialization.py
@@ -24,6 +24,7 @@ import tensorflow as tf
from apache_beam.testing.test_pipeline import TestPipeline
from typing import List
+from basic_pitch.constants import Split
from basic_pitch.data.datasets.guitarset import GuitarSetToTfExample
from basic_pitch.data.pipeline import WriteBatchToTfRecord
from basic_pitch.data.tf_example_deserialization import (
@@ -135,7 +136,7 @@ def test_sample_datasets(tmp_path: pathlib.Path) -> None:
datasets_home = setup_test_resources(tmp_path)
ds = sample_datasets(
- split="train",
+ split=Split.train,
datasets_base_path=str(datasets_home),
datasets=["guitarset"],
dataset_sampling_frequency=np.array([1]),
@@ -148,12 +149,12 @@ def test_sample_datasets(tmp_path: pathlib.Path) -> None:
def test_transcription_file_generator_train(tmp_path: pathlib.Path) -> None:
- dataset_path = tmp_path / "test_ds" / "splits" / "train"
+ dataset_path = tmp_path / "test_ds" / "splits" / Split.train.name
dataset_path.mkdir(parents=True)
create_empty_tfrecord(dataset_path / "test.tfrecord")
file_gen, random_seed = transcription_file_generator(
- "train", ["test_ds"], datasets_base_path=str(tmp_path), sample_weights=np.array([1])
+ Split.train, ["test_ds"], datasets_base_path=str(tmp_path), sample_weights=np.array([1])
)
assert random_seed is False
@@ -167,12 +168,12 @@ def test_transcription_file_generator_train(tmp_path: pathlib.Path) -> None:
def test_transcription_file_generator_valid(tmp_path: pathlib.Path) -> None:
- dataset_path = tmp_path / "test_ds" / "splits" / "valid"
+ dataset_path = tmp_path / "test_ds" / "splits" / Split.validation.name
dataset_path.mkdir(parents=True)
create_empty_tfrecord(dataset_path / "test.tfrecord")
file_gen, random_seed = transcription_file_generator(
- "valid", ["test_ds"], datasets_base_path=str(tmp_path), sample_weights=np.array([1])
+ Split.validation, ["test_ds"], datasets_base_path=str(tmp_path), sample_weights=np.array([1])
)
assert random_seed is True