Loading deepchem/models/tensorgraph/robust_multitask.py +83 −77 Original line number Diff line number Diff line Loading @@ -21,6 +21,26 @@ class RobustMultitaskClassifier(TensorGraph): Key idea is to have bypass layers that feed directly from features to task output. Hopefully will allow tasks to route around bad multitasking. """ def __init__(self, n_tasks, n_features, layer_sizes=[1000], weight_init_stddevs=0.02, bias_init_consts=1.0, weight_decay_penalty=0.0, weight_decay_penalty_type="l2", dropouts=0.5, activation_fns=tf.nn.relu, n_classes=2, bypass_layer_sizes=[100], bypass_weight_init_stddevs=[.02], bypass_bias_init_consts=[1.], bypass_dropouts=[.5], **kwargs): """ Create a RobustMultitaskClassifier. Parameters ---------- n_tasks: int Loading Loading @@ -61,23 +81,6 @@ class RobustMultitaskClassifier(TensorGraph): the dropout probablity to use for bypass layers. same requirements as dropouts """ def __init__(self, n_tasks, n_features, layer_sizes=[1000], weight_init_stddevs=0.02, bias_init_consts=1.0, weight_decay_penalty=0.0, weight_decay_penalty_type="l2", dropouts=0.5, activation_fns=tf.nn.relu, n_classes=2, bypass_layer_sizes=[100], bypass_weight_init_stddevs=[.02], bypass_bias_init_consts=[1.], bypass_dropouts=[.5], **kwargs): super(RobustMultitaskClassifier, self).__init__(**kwargs) self.n_tasks = n_tasks self.n_features = n_features Loading Loading @@ -207,6 +210,25 @@ class RobustMultitaskRegressor(TensorGraph): Key idea is to have bypass layers that feed directly from features to task output. Hopefully will allow tasks to route around bad multitasking. """ def __init__(self, n_tasks, n_features, layer_sizes=[1000], weight_init_stddevs=0.02, bias_init_consts=1.0, weight_decay_penalty=0.0, weight_decay_penalty_type="l2", dropouts=0.5, activation_fns=tf.nn.relu, bypass_layer_sizes=[100], bypass_weight_init_stddevs=[.02], bypass_bias_init_consts=[1.], bypass_dropouts=[.5], **kwargs): """ Create a RobustMultitaskRegressor. Parameters ---------- n_tasks: int Loading Loading @@ -245,22 +267,6 @@ class RobustMultitaskRegressor(TensorGraph): the dropout probablity to use for bypass layers. same requirements as dropouts """ def __init__(self, n_tasks, n_features, layer_sizes=[1000], weight_init_stddevs=0.02, bias_init_consts=1.0, weight_decay_penalty=0.0, weight_decay_penalty_type="l2", dropouts=0.5, activation_fns=tf.nn.relu, bypass_layer_sizes=[100], bypass_weight_init_stddevs=[.02], bypass_bias_init_consts=[1.], bypass_dropouts=[.5], **kwargs): super(RobustMultitaskRegressor, self).__init__(**kwargs) self.n_tasks = n_tasks self.n_features = n_features Loading Loading
deepchem/models/tensorgraph/robust_multitask.py +83 −77 Original line number Diff line number Diff line Loading @@ -21,6 +21,26 @@ class RobustMultitaskClassifier(TensorGraph): Key idea is to have bypass layers that feed directly from features to task output. Hopefully will allow tasks to route around bad multitasking. """ def __init__(self, n_tasks, n_features, layer_sizes=[1000], weight_init_stddevs=0.02, bias_init_consts=1.0, weight_decay_penalty=0.0, weight_decay_penalty_type="l2", dropouts=0.5, activation_fns=tf.nn.relu, n_classes=2, bypass_layer_sizes=[100], bypass_weight_init_stddevs=[.02], bypass_bias_init_consts=[1.], bypass_dropouts=[.5], **kwargs): """ Create a RobustMultitaskClassifier. Parameters ---------- n_tasks: int Loading Loading @@ -61,23 +81,6 @@ class RobustMultitaskClassifier(TensorGraph): the dropout probablity to use for bypass layers. same requirements as dropouts """ def __init__(self, n_tasks, n_features, layer_sizes=[1000], weight_init_stddevs=0.02, bias_init_consts=1.0, weight_decay_penalty=0.0, weight_decay_penalty_type="l2", dropouts=0.5, activation_fns=tf.nn.relu, n_classes=2, bypass_layer_sizes=[100], bypass_weight_init_stddevs=[.02], bypass_bias_init_consts=[1.], bypass_dropouts=[.5], **kwargs): super(RobustMultitaskClassifier, self).__init__(**kwargs) self.n_tasks = n_tasks self.n_features = n_features Loading Loading @@ -207,6 +210,25 @@ class RobustMultitaskRegressor(TensorGraph): Key idea is to have bypass layers that feed directly from features to task output. Hopefully will allow tasks to route around bad multitasking. """ def __init__(self, n_tasks, n_features, layer_sizes=[1000], weight_init_stddevs=0.02, bias_init_consts=1.0, weight_decay_penalty=0.0, weight_decay_penalty_type="l2", dropouts=0.5, activation_fns=tf.nn.relu, bypass_layer_sizes=[100], bypass_weight_init_stddevs=[.02], bypass_bias_init_consts=[1.], bypass_dropouts=[.5], **kwargs): """ Create a RobustMultitaskRegressor. Parameters ---------- n_tasks: int Loading Loading @@ -245,22 +267,6 @@ class RobustMultitaskRegressor(TensorGraph): the dropout probablity to use for bypass layers. same requirements as dropouts """ def __init__(self, n_tasks, n_features, layer_sizes=[1000], weight_init_stddevs=0.02, bias_init_consts=1.0, weight_decay_penalty=0.0, weight_decay_penalty_type="l2", dropouts=0.5, activation_fns=tf.nn.relu, bypass_layer_sizes=[100], bypass_weight_init_stddevs=[.02], bypass_bias_init_consts=[1.], bypass_dropouts=[.5], **kwargs): super(RobustMultitaskRegressor, self).__init__(**kwargs) self.n_tasks = n_tasks self.n_features = n_features Loading
