def __init__(self,
inputs,
outputs,
name='auto_model',
max_trials=100,
directory=None,
objective='val_loss',
tuner='greedy',
overwrite=False,
seed=None):
self.inputs = nest.flatten(inputs)
self.outputs = nest.flatten(outputs)
self.seed = seed
if seed:
np.random.seed(seed)
tf.random.set_seed(seed)
# TODO: Support passing a tuner instance.
if isinstance(tuner, str):
tuner = tuner_module.get_tuner_class(tuner)
self.tuner = tuner(
hypermodel=lambda hp: None,
overwrite=overwrite,
objective=objective,
max_trials=max_trials,
directory=directory,
seed=self.seed,
project_name=name)
self._split_dataset = False
if all([isinstance(output_node, base.Head)
for output_node in self.outputs]):
self.heads = self.outputs
else:
self.heads = [output_node.in_blocks[0] for output_node in self.outputs]
def __init__(self,
inputs,
outputs,
name='auto_model',
max_trials=100,
directory=None,
objective='val_loss',
tuner='greedy',
overwrite=False,
seed=None):
self.inputs = nest.flatten(inputs)
self.outputs = nest.flatten(outputs)
self.name = name
self.max_trials = max_trials
self.directory = directory
self.seed = seed
self.hyper_graph = None
self.objective = objective
# TODO: Support passing a tuner instance.
self.tuner = tuner_module.get_tuner_class(tuner)
self.overwrite = overwrite
self._split_dataset = False
if all([
isinstance(output_node, base.Head)
for output_node in self.outputs
]):
self.heads = self.outputs
else:
self.heads = [
output_node.in_blocks[0] for output_node in self.outputs
]