def paths_and_labels_to_dataset(
file_paths,
labels,
label_mode,
num_classes,
sampling_rate,
output_sequence_length,
ragged,
):
"""Constructs a fixed-size dataset of audio and labels."""
path_ds = tf.data.Dataset.from_tensor_slices(file_paths)
audio_ds = path_ds.map(
lambda x: read_and_decode_audio(x, sampling_rate,
output_sequence_length),
num_parallel_calls=tf.data.AUTOTUNE,
)
if ragged:
audio_ds = audio_ds.map(
lambda x: tf.RaggedTensor.from_tensor(x),
num_parallel_calls=tf.data.AUTOTUNE,
)
if label_mode:
label_ds = dataset_utils.labels_to_dataset(labels, label_mode,
num_classes)
audio_ds = tf.data.Dataset.zip((audio_ds, label_ds))
return audio_ds
def paths_and_labels_to_dataset(file_paths, labels, label_mode, num_classes,
max_length):
"""Constructs a dataset of text strings and labels."""
path_ds = tf.data.Dataset.from_tensor_slices(file_paths)
string_ds = path_ds.map(lambda x: path_to_string_content(x, max_length),
num_parallel_calls=tf.data.AUTOTUNE)
if label_mode:
label_ds = dataset_utils.labels_to_dataset(labels, label_mode,
num_classes)
string_ds = tf.data.Dataset.zip((string_ds, label_ds))
return string_ds
def paths_and_labels_to_dataset(image_paths,
image_size,
num_channels,
labels,
label_mode,
num_classes,
interpolation,
crop_to_aspect_ratio=False):
"""Constructs a dataset of images and labels."""
# TODO(fchollet): consider making num_parallel_calls settable
path_ds = tf.data.Dataset.from_tensor_slices(image_paths)
args = (image_size, num_channels, interpolation, crop_to_aspect_ratio)
img_ds = path_ds.map(lambda x: load_image(x, *args),
num_parallel_calls=tf.data.AUTOTUNE)
if label_mode:
label_ds = dataset_utils.labels_to_dataset(labels, label_mode,
num_classes)
img_ds = tf.data.Dataset.zip((img_ds, label_ds))
return img_ds