Management datasets

The instance created by the Main class allow to manage the datasets, in this section we will see the CRUD for datasets.

Create dataset

The add_dataset method is used to define a dataset. Has the following parameters:

• name (string): Name of the dataset.
• epochs (int): Number of epochs to train the model.
• train_ds (tf.data.Dataset | List): Dataset to train the model.
• val_ds (tf.data.Dataset | List, optional): Dataset to validate the model.
• test_ds (tf.data.Dataset | List, optional): Dataset to test the model.
• batch_size (int, optional): Batch size to train the model.
• shuffle_buffer (int, optional): Buffer size to shuffle the dataset.

Example

all_data = tf.data.Dataset.from_tensor_slices((
  tf.random.uniform([1000, 2]),
  tf.random.uniform([1000, 1])
))

train_ds = all_data.take(5)
val_ds = all_data.skip(5)

ia_maker.add_dataset(
  name='dataset_1',
  epochs=10,
  batch_size=32,
  shuffle_buffer=512,
  train_ds=train_ds,
  val_ds=val_ds
)

Output

Dataset dataset_1 was added

Update dataset

The update_dataset method is used to update a dataset. Has the following parameters:

• name (string): Name of the dataset to update.
• epochs (int): New number of epochs to train the model.
• train_ds (tf.data.Dataset | List): New dataset to train the model.
• val_ds (tf.data.Dataset | List): New dataset to validate the model.
• test_ds (tf.data.Dataset | List): New dataset to test the model.
• batch_size (int): New batch size to train the model.
• shuffle_buffer (int): New buffer size to shuffle the dataset.

Example

ia_maker.update_dataset(
  name='dataset_1',
  epochs=50,
  batch_size=128,
  shuffle_buffer=1024,
  train_ds=train_ds_2,
  val_ds=val_ds_2
)

Output

Dataset dataset_1 was updated

Delete dataset

The delete_dataset method is used to delete a dataset. Has the following parameters:

• name (string): Name of the dataset to delete.

Example

ia_maker.delete_dataset(name='dataset_1')

Output

Dataset dataset_1 was deleted