from __future__ import print_function, division
from __future__ import absolute_import
import os
def optimizer_factory(job):
import keras.optimizers
if isinstance(job, dict):
# old simple code
optimizer = job['$value']
optimizer_settings = job[optimizer]
else:
optimizer = job.get_parameter('keras_optimizer')
optimizer_settings = job.get_parameter('keras_optimizer', return_group=True)
if 'sgd' == optimizer:
return keras.optimizers.SGD(lr=optimizer_settings['learning_rate'] or 0.01, momentum=optimizer_settings['momentum'] or 0, nesterov=optimizer_settings['nesterov'], decay=optimizer_settings['decay'] or 0.0)
if 'rmsprop' == optimizer:
return keras.optimizers.RMSprop(lr=optimizer_settings['learning_rate'] or 0.001, rho=optimizer_settings['rho'] or 0.9, epsilon=optimizer_settings['epsilon'] or 1e-08, decay=optimizer_settings['decay'] or 0.0)
if 'adagrad' == optimizer:
return keras.optimizers.Adagrad(lr=optimizer_settings['learning_rate'] or 0.01, epsilon=optimizer_settings['epsilon'] or 1e-08, decay=optimizer_settings['decay'] or 0.0)
if 'adadelta' == optimizer:
return keras.optimizers.Adadelta(lr=optimizer_settings['learning_rate'] or 1.0, rho=optimizer_settings['rho'] or 0.95, epsilon=optimizer_settings['epsilon'] or 1e-08, decay=optimizer_settings['decay'] or 0.0)
if 'adam' == optimizer:
return keras.optimizers.Adam(lr=optimizer_settings['learning_rate'] or 0.001, beta_1=optimizer_settings['beta_1'] or 0.9, beta_2=optimizer_settings['beta_2'] or 0.999, epsilon=optimizer_settings['epsilon'] or 1e-08, decay=optimizer_settings['decay'] or 0.0)
if 'adamax' == optimizer:
return keras.optimizers.Adamax(lr=optimizer_settings['learning_rate'] or 0.002, beta_1=optimizer_settings['beta_1'] or 0.9, beta_2=optimizer_settings['beta_2'] or 0.999, epsilon=optimizer_settings['epsilon'] or 1e-08, decay=optimizer_settings['decay'] or 0.0)
def load_weights(model, weights_path):
from keras import backend as K
if not os.path.isfile(weights_path):
raise Exception("File does not exist.")
import h5py
f = h5py.File(weights_path, mode='r')
# new file format
layer_names = [n.decode('utf8') for n in f.attrs['layer_names']]
if len(layer_names) != len(model.layers):
print("Warning: Layer count different")
# we batch weight value assignments in a single backend call
# which provides a speedup in TensorFlow.
weight_value_tuples = []
for k, name in enumerate(layer_names):
g = f[name]
weight_names = [n.decode('utf8') for n in g.attrs['weight_names']]
layer = model.get_layer(name=name)
if layer and len(weight_names):
weight_values = [g[weight_name] for weight_name in weight_names]
if not hasattr(layer, 'trainable_weights'):
print("Layer %s (%s) has no trainable weights, but we tried to load it." % (
name, type(layer).__name__))
else:
symbolic_weights = layer.trainable_weights + layer.non_trainable_weights
if len(weight_values) != len(symbolic_weights):
raise Exception('Layer #' + str(k) +
' (named "' + layer.name +
'" in the current model) was found to '
'correspond to layer ' + name +
' in the save file. '
'However the new layer ' + layer.name +
' expects ' + str(len(symbolic_weights)) +
' weights, but the saved weights have ' +
str(len(weight_values)) +
' elements.')
weight_value_tuples += list(zip(symbolic_weights, weight_values))
K.batch_set_value(weight_value_tuples)
f.close()
def model_to_graph(model):
from keras.engine import InputLayer
from keras.layers import Convolution2D, MaxPooling2D, Dropout, Activation, Dense, Embedding, RepeatVector, Merge
graph = {
'nodes': [],
'links': [],
'groups': []
}
map = {
'idx': {},
'flatten': [],
'group_pointer': -1
}
def layer_to_dict(layer):
info = {}
if isinstance(layer, Dropout):
info['dropout'] = layer.rate
if isinstance(layer, Dense):
info['neurons'] = layer.units
info['activaton'] = layer.activation.__name__
if isinstance(layer, Convolution2D):
info['receptiveField'] = layer.kernel_size
info['features'] = layer.filters
if isinstance(layer, MaxPooling2D):
info['poolingArea'] = [layer.pool_size[0], layer.pool_size[1]]
if isinstance(layer, Embedding):
info['inputDim'] = layer.input_dim
info['outputDim'] = layer.output_dim
if isinstance(layer, Activation):
info['activaton'] = layer.activation.__name__
if isinstance(layer, Merge):
info['mode'] = layer.mode
if isinstance(layer, RepeatVector):
info['n'] = layer.n
if isinstance(layer, InputLayer):
info['inputShape'] = layer.input_shape
if hasattr(layer, 'output_shape'):
info['outputShape'] = layer.output_shape
return {
'name': layer.name,
'class': type(layer).__name__,
'width': 60,
'height': 40,
'info': info
}
def add_layer(layer):
graph['nodes'].append(layer_to_dict(layer))
map['flatten'].append(layer)
map['idx'][layer.name] = len(graph['nodes']) - 1
# if map['group_pointer'] >= 0:
# graph['groups'][map['group_pointer']].append(len(graph['nodes'])-1)
def get_idx(layer):
return map['idx'][layer.name]
def extract_layers(layers):
for layer in layers:
if layer not in map['flatten']:
add_layer(layer)
if hasattr(layer, 'layers') and isinstance(layer.layers, list):
# graph['groups'].append([])
# map['group_pointer'] += 1
extract_layers(layer.layers)
# map['group_pointer'] -= 1
elif hasattr(layer, 'inbound_nodes'):
for inbound_node in layer.inbound_nodes:
extract_layers(inbound_node.inbound_layers)
extract_layers(model.layers)
# build edges
for layer in map['flatten']:
if hasattr(layer, 'inbound_nodes'):
for inbound_node in layer.inbound_nodes:
for inbound_layer in inbound_node.inbound_layers:
graph['links'].append({
'source': get_idx(inbound_layer),
'target': get_idx(layer),
})
if hasattr(layer, 'layers') and isinstance(layer.layers, list):
graph['links'].append({
'source': get_idx(layer.layers[-1]),
'target': get_idx(layer),
})
return graph
