def setup_logging(parameters=None,
coloredlogs=False,
logging_file=None,
default_setup_file='logging.yaml',
default_level=logging.INFO,
environmental_variable='LOG_CFG'):
"""Setup logging configuration
Parameters
----------
parameters : dict
Parameters in dict
Default value None
coloredlogs : bool
Use coloredlogs
Default value False
logging_file : str
Log filename for file based logging, if none given no file logging is used.
Default value None
environmental_variable : str
Environmental variable to get the logging setup filename, if set will override default_setup_file
Default value 'LOG_CFG'
default_setup_file : str
Default logging parameter file, used if one is not set in given ParameterContainer
Default value 'logging.yaml'
default_level : logging.level
Default logging level, used if one is not set in given ParameterContainer
Default value 'logging.INFO'
Returns
-------
nothing
"""
class LoggerFilter(object):
def __init__(self, level):
self.__level = level
def filter(self, log_record):
return log_record.levelno <= self.__level
formatters = {
'simple': "[%(levelname).1s] %(message)s",
'normal': "%(asctime)s\t[%(name)-20s]\t[%(levelname)-8s]\t%(message)s",
'extended': "[%(asctime)s] [%(name)s]\t [%(levelname)-8s]\t %(message)s \t(%(filename)s:%(lineno)s)",
'extended2': "[%(levelname).1s] %(message)s \t(%(filename)s:%(lineno)s)",
'file_extended': "[%(levelname).1s] [%(asctime)s] %(message)s",
}
if not parameters:
logging_parameter_file = default_setup_file
value = os.getenv(environmental_variable, None)
if value:
# If environmental variable set
logging_parameter_file = value
if os.path.exists(logging_parameter_file):
with open(logging_parameter_file, 'rt') as f:
config = yaml.safe_load(f.read())
logging.config.dictConfig(config)
try:
# Check if coloredlogs is available
import coloredlogs
coloredlogs.install(
level=config['handlers']['console']['level'],
fmt=config['formatters'][config['handlers']['console']['formatter']]['format']
)
except ImportError:
pass
else:
if coloredlogs:
try:
# Check if coloredlogs is available
import coloredlogs
coloredlogs.install(
level=logging.INFO,
fmt=formatters['simple'],
reconfigure=True
)
except ImportError:
logger = logging.getLogger()
logger.setLevel(default_level)
console_info = logging.StreamHandler()
console_info.setLevel(logging.INFO)
console_info.setFormatter(logging.Formatter(formatters['simple']))
console_info.addFilter(LoggerFilter(logging.INFO))
logger.addHandler(console_info)
console_debug = logging.StreamHandler()
console_debug.setLevel(logging.DEBUG)
console_debug.setFormatter(logging.Formatter(formatters['simple']))
console_debug.addFilter(LoggerFilter(logging.DEBUG))
logger.addHandler(console_debug)
console_warning = logging.StreamHandler()
console_warning.setLevel(logging.WARNING)
console_warning.setFormatter(logging.Formatter(formatters['simple']))
console_warning.addFilter(LoggerFilter(logging.WARNING))
logger.addHandler(console_warning)
console_critical = logging.StreamHandler()
console_critical.setLevel(logging.CRITICAL)
console_critical.setFormatter(logging.Formatter(formatters['extended2']))
console_critical.addFilter(LoggerFilter(logging.CRITICAL))
logger.addHandler(console_critical)
console_error = logging.StreamHandler()
console_error.setLevel(logging.ERROR)
console_error.setFormatter(logging.Formatter(formatters['extended2']))
console_error.addFilter(LoggerFilter(logging.ERROR))
logger.addHandler(console_error)
if logging_file:
file_info = logging.handlers.RotatingFileHandler(
filename=logging_file,
maxBytes=10485760,
backupCount=20,
encoding='utf8'
)
file_info.setLevel(logging.INFO)
file_info.setFormatter(logging.Formatter(formatters['file_extended']))
logger.addHandler(file_info)
else:
logger = logging.getLogger()
logger.setLevel(default_level)
console_info = logging.StreamHandler()
console_info.setLevel(logging.INFO)
console_info.setFormatter(logging.Formatter(formatters['simple']))
console_info.addFilter(LoggerFilter(logging.INFO))
logger.addHandler(console_info)
console_debug = logging.StreamHandler()
console_debug.setLevel(logging.DEBUG)
console_debug.setFormatter(logging.Formatter(formatters['simple']))
console_debug.addFilter(LoggerFilter(logging.DEBUG))
logger.addHandler(console_debug)
console_warning = logging.StreamHandler()
console_warning.setLevel(logging.WARNING)
console_warning.setFormatter(logging.Formatter(formatters['simple']))
console_warning.addFilter(LoggerFilter(logging.WARNING))
logger.addHandler(console_warning)
console_critical = logging.StreamHandler()
console_critical.setLevel(logging.CRITICAL)
console_critical.setFormatter(logging.Formatter(formatters['extended2']))
console_critical.addFilter(LoggerFilter(logging.CRITICAL))
logger.addHandler(console_critical)
console_error = logging.StreamHandler()
console_error.setLevel(logging.ERROR)
console_error.setFormatter(logging.Formatter(formatters['extended2']))
console_error.addFilter(LoggerFilter(logging.ERROR))
logger.addHandler(console_error)
if logging_file:
file_info = logging.handlers.RotatingFileHandler(
filename=logging_file,
maxBytes=10485760,
backupCount=20,
encoding='utf8'
)
file_info.setLevel(logging.INFO)
file_info.setFormatter(logging.Formatter(formatters['file_extended']))
logger.addHandler(file_info)
else:
from dcase_util.containers import DictContainer
parameters = DictContainer(parameters)
logging.config.dictConfig(parameters.get('parameters'))
if (parameters.get('colored', False) and
'console' in parameters.get_path('parameters.handlers')):
try:
# Check if coloredlogs is available
import coloredlogs
coloredlogs.install(
level=parameters.get_path('parameters.handlers.console.level'),
fmt=parameters.get_path('parameters.formatters')[
parameters.get_path('parameters.handlers.console.formatter')
].get('format')
)
except ImportError:
pass
# Function to handle uncaught expections
def handle_exception(exc_type, exc_value, exc_traceback):
if issubclass(exc_type, KeyboardInterrupt):
sys.__excepthook__(exc_type, exc_value, exc_traceback)
return
logger.error('Uncaught exception', exc_info=(exc_type, exc_value, exc_traceback))
sys.excepthook = handle_exception
def system_meta(self,
results,
task=None,
check_development_dataset=True,
check_evaluation_dataset=True):
"""Check system result scores given in the meta data
Parameters
----------
results : dict
Result meta data
task : str, optional
Temporal override for the task parameter given to class constructor.
check_development_dataset : bool
Check development dataset results
check_evaluation_dataset : bool
Check evaluation dataset results
Returns
-------
self
"""
if task is None:
task = self.task
if results is None:
self.error_log.append(u'No results section')
else:
results = DictContainer(results)
if check_development_dataset:
# Check development dataset results
if results.get('development_dataset') is None:
self.error_log.append(u'No development results given')
else:
if task == 'ASC':
if results.get_path(
'development_dataset.overall.accuracy'
) is None:
self.error_log.append(
u'No development overall result given ')
elif task == 'SED_event':
if results.get_path(
'development_dataset.event_based.overall.er'
) is None:
self.error_log.append(
u'No development overall result given [event_based.overall.er]'
)
if results.get_path(
'development_dataset.event_based.overall.f1'
) is None:
self.error_log.append(
u'No development overall result given [event_based.overall.f1]'
)
elif task == 'SED_segment':
if results.get_path(
'development_dataset.segment_based.overall.er'
) is None:
self.error_log.append(
u'No development overall result given [segment_based.overall.er]'
)
if results.get_path(
'development_dataset.segment_based.overall.f1'
) is None:
self.error_log.append(
u'No development overall result given [segment_based.overall.f1]'
)
elif task == 'task4':
pass
# Check development dataset / class wise results
if task == 'ASC':
if results.get_path(
'development_dataset.class_wise') is None:
self.error_log.append(
u'No class_wise development results given')
else:
if len(
results.get_path(
'development_dataset.class_wise')
) != len(self.class_labels):
self.error_log.append(
u'Incorrect number class-wise development results given [{class_wise:d}/{target:d}]'
.format(class_wise=len(
results.get_path(
'development_dataset.class_wise')),
target=len(self.class_labels)))
for class_label, class_data in iteritems(
results.get_path(
'development_dataset.class_wise')):
if 'accuracy' not in class_data or class_data[
'accuracy'] is None:
self.error_log.append(
u'Incorrect class-wise development results given for [{class_label:s}]'
.format(class_label=class_label))
elif task == 'SED_event':
if results.get_path(
'development_dataset.event_based.class_wise'
) is not None:
if len(
results.get_path(
'development_dataset.event_based.class_wise'
)) != len(self.class_labels):
self.error_log.append(
u'Incorrect number class-wise development results given [{class_wise:d}/{target:d}]'
.format(class_wise=len(
results.get_path(
'development_dataset.event_based.class_wise'
)),
target=len(self.class_labels)))
for class_label, class_data in iteritems(
results.get_path(
'development_dataset.event_based.class_wise'
)):
if class_data.get('er') is None:
self.error_log.append(
u'Incorrect class-wise development results given for [{class_label:s} / er]'
.format(class_label=class_label))
if class_data.get('f1') is None:
self.error_log.append(
u'Incorrect class-wise development results given for [{class_label:s} / f1]'
.format(class_label=class_label))
else:
self.error_log.append(
u'No class_wise development results given')
elif task == 'SED_segment':
if results.get_path(
'development_dataset.segment_based.class_wise'
) is not None:
if len(
results.get_path(
'development_dataset.segment_based.class_wise'
)) != len(self.class_labels):
self.error_log.append(
u'Incorrect number class-wise development results given [{class_wise:d}/{target:d}]'
.format(class_wise=len(
results.get_path(
'development_dataset.segment_based.class_wise'
)),
target=len(self.class_labels)))
for class_label, class_data in iteritems(
results.get_path(
'development_dataset.segment_based.class_wise'
)):
if class_data.get('er') is None:
self.error_log.append(
u'Incorrect class-wise development results given for [{class_label:s} / er]'
.format(class_label=class_label))
if class_data.get('f1') is None:
self.error_log.append(
u'Incorrect class-wise development results given for [{class_label:s} / f1]'
.format(class_label=class_label))
else:
self.error_log.append(
u'No class_wise development results given')
elif task == 'task4':
pass
if check_evaluation_dataset:
# Check evaluation dataset results
if 'evaluation_dataset' not in results:
self.error_log.append(u'No evaluation results given')
else:
if task == 'ASC':
if results.get_path(
'evaluation_dataset.overall') is None:
self.error_log.append(
u'No evaluation results given')
if results.get_path(
'evaluation_dataset.class_wise') is not None:
if len(
results.get_path(
'evaluation_dataset.class_wise')
) != len(self.class_labels):
self.error_log.append(
u'Incorrect number class-wise evaluation results given [{class_wise:d}/{target:d}]'
.format(class_wise=len(
results.get_path(
'evaluation_dataset.class_wise')),
target=len(self.class_labels)))
for class_label, class_data in iteritems(
results.get_path(
'evaluation_dataset.class_wise')):
if class_data.get('accuracy') is None:
self.error_log.append(
u'Incorrect class-wise evaluation results given for [{class_label:s}]'
.format(class_label=class_label))
else:
self.error_log.append(
u'No class_wise development results given')
elif task == 'SED_event':
if results.get_path(
'evaluation_dataset.event_based.overall.er'
) is None:
self.error_log.append(
u'No evaluation results given [event_based.overall.er]'
)
if results.get_path(
'evaluation_dataset.event_based.overall.f1'
) is None:
self.error_log.append(
u'No evaluation results given [event_based.overall.f1]'
)
if results.get_path(
'evaluation_dataset.event_based.class_wise'
) is not None:
if len(
results.get_path(
'evaluation_dataset.event_based.class_wise'
)) != len(self.class_labels):
self.error_log.append(
u'Incorrect number class-wise evaluation results given [{class_wise:d}/{target:d}]'
.format(class_wise=len(
results.get_path(
'evaluation_dataset.event_based.class_wise'
)),
target=len(self.class_labels)))
for class_label, class_data in iteritems(
results.get_path(
'evaluation_dataset.event_based.class_wise'
)):
if class_data.get('er') is None:
self.error_log.append(
u'Incorrect class-wise evaluation results given for [{class_label:s} / er]'
.format(class_label=class_label))
if class_data.get('f1') is None:
self.error_log.append(
u'Incorrect class-wise evaluation results given for [{class_label:s} / f1]'
.format(class_label=class_label))
else:
self.error_log.append(
u'No class_wise evaluation results given')
elif task == 'SED_segment':
if results.get_path(
'evaluation_dataset.segment_based.overall.er'
) is None:
self.error_log.append(
u'No evaluation results given [segment_based.overall.er]'
)
if results.get_path(
'evaluation_dataset.segment_based.overall.f1'
) is None:
self.error_log.append(
u'No evaluation results given [segment_based.overall.f1]'
)
if results.get_path(
'evaluation_dataset.segment_based.class_wise'
) is not None:
if len(
results.get_path(
'evaluation_dataset.segment_based.class_wise'
)) != len(self.class_labels):
self.error_log.append(
u'Incorrect number class-wise evaluation results given [{class_wise:d}/{target:d}]'
.format(class_wise=len(
results.get_path(
'evaluation_dataset.segment_based.class_wise'
)),
target=len(self.class_labels)))
for class_label, class_data in iteritems(
results.get_path(
'evaluation_dataset.segment_based.class_wise'
)):
if class_data.get('er') is None:
self.error_log.append(
u'Incorrect class-wise evaluation results given for [{class_label:s} / er]'
.format(class_label=class_label))
if class_data.get('f1') is None:
self.error_log.append(
u'Incorrect class-wise evaluation results given for [{class_label:s} / f1]'
.format(class_label=class_label))
else:
self.error_log.append(
u'No class_wise evaluation results given')
elif task == 'task4':
pass
return self
def submission_authors(self,
authors,
check_email=True,
check_affiliation=True,
check_affiliation_abbreviation=True,
check_affiliation_department=True):
"""Check submission authors
Parameters
----------
authors : list of dict
List of authors dicts.
check_email : bool
Check that author email exists.
check_affiliation : bool
Check author affiliation.
check_affiliation_abbreviation : bool
Check that affiliation abbreviation exists.
check_affiliation_department : bool
Check that affiliation has department defined.
Returns
-------
self
"""
if not isinstance(authors, list):
self.error_log.append(
u'Authors not given in list format for the submission')
for author in authors:
author = DictContainer(author)
if author.get('lastname') is None:
self.error_log.append(
u'No lastname given for author ({last_name:s}, {first_name:s})'
.format(last_name=author['lastname'],
first_name=author['firstname']))
if author.get('firstname') is None:
self.error_log.append(
u'No firstname given for author ({last_name:s}, {first_name:s})'
.format(last_name=author['lastname'],
first_name=author['firstname']))
if check_email:
if author.get('email') is None:
self.error_log.append(
u'No email given for author ({last_name:s}, {first_name:s})'
.format(last_name=author['lastname'],
first_name=author['firstname']))
if check_affiliation:
if author.get('affiliation') is None:
self.error_log.append(
u'No affiliation given for author ({last_name:s}, {first_name:s})'
.format(last_name=author['lastname'],
first_name=author['firstname']))
else:
if isinstance(author.get('affiliation'), list):
for a in author.get('affiliation'):
affiliation = ', '.join(
filter(None, list(a.values())))
if check_affiliation_abbreviation:
if a.get('abbreviation') is None:
self.error_log.append(
u'No abbreviation given ({last_name:s}, {first_name:s}, {affiliation:s})'
.format(last_name=author['lastname'],
first_name=author['firstname'],
affiliation=affiliation))
if check_affiliation_department:
if a.get('department') is None:
self.error_log.append(
u'No department given ({last_name:s}, {first_name:s}, {affiliation:s})'
.format(last_name=author['lastname'],
first_name=author['firstname'],
affiliation=affiliation))
if a.get('institute') is None:
self.error_log.append(
u'No institute given ({last_name:s}, {first_name:s}, {affiliation:s})'
.format(last_name=author['lastname'],
first_name=author['firstname'],
affiliation=affiliation))
if a.get('location') is None:
self.error_log.append(
u'No location given ({last_name:s}, {first_name:s}, {affiliation:s})'
.format(last_name=author['lastname'],
first_name=author['firstname'],
affiliation=affiliation))
else:
affiliation = ', '.join(
filter(None, list(author['affiliation'].values())))
if check_affiliation_abbreviation:
if author.get_path(
'affiliation.abbreviation') is None:
self.error_log.append(
u'No abbreviation given ({last_name:s}, {first_name:s}, {affiliation:s})'
.format(last_name=author['lastname'],
first_name=author['firstname'],
affiliation=affiliation))
if check_affiliation_department:
if author.get_path(
'affiliation.department') is None:
self.error_log.append(
u'No department given ({last_name:s}, {first_name:s}, {affiliation:s})'
.format(last_name=author['lastname'],
first_name=author['firstname'],
affiliation=affiliation))
if author.get_path('affiliation.institute') is None:
self.error_log.append(
u'No institute given ({last_name:s}, {first_name:s}, {affiliation:s})'
.format(last_name=author['lastname'],
first_name=author['firstname'],
affiliation=affiliation))
if author.get_path('affiliation.location') is None:
self.error_log.append(
u'No location given ({last_name:s}, {first_name:s})'
.format(last_name=author['lastname'],
first_name=author['firstname'],
affiliation=affiliation))
return self
def model_summary_string(keras_model, mode='keras', show_parameters=True, display=False):
"""Model summary in a formatted string, similar to Keras model summary function.
Parameters
----------
keras_model : keras model
Keras model
mode : str
Summary mode ['extended', 'keras']. In case 'keras', standard Keras summary is returned.
Default value keras
show_parameters : bool
Show model parameter count and input / output shapes
Default value True
display : bool
Display summary immediately, otherwise return string
Default value False
Returns
-------
str
Model summary
"""
if is_jupyter():
ui = FancyHTMLStringifier()
html_mode = True
else:
ui = FancyStringifier()
html_mode = False
output = ''
output += ui.line('Model summary') + '\n'
if mode == 'extended' or mode == 'extended_wide':
layer_name_map = {
'BatchNormalization': 'BatchNorm',
}
layer_type_html_tags = {
'InputLayer': '<span class="label label-default">{0:s}</span>',
'Dense': '<span class="label label-primary">{0:s}</span>',
'TimeDistributed': '<span class="label label-primary">{0:s}</span>',
'BatchNorm': '<span class="label label-default">{0:s}</span>',
'Activation': '<span class="label label-default">{0:s}</span>',
'Dropout': '<span class="label label-default">{0:s}</span>',
'Flatten': '<span class="label label-success">{0:s}</span>',
'Reshape': '<span class="label label-success">{0:s}</span>',
'Permute': '<span class="label label-success">{0:s}</span>',
'Conv1D': '<span class="label label-warning">{0:s}</span>',
'Conv2D': '<span class="label label-warning">{0:s}</span>',
'MaxPooling1D': '<span class="label label-success">{0:s}</span>',
'MaxPooling2D': '<span class="label label-success">{0:s}</span>',
'MaxPooling3D': '<span class="label label-success">{0:s}</span>',
'AveragePooling1D': '<span class="label label-success">{0:s}</span>',
'AveragePooling2D': '<span class="label label-success">{0:s}</span>',
'AveragePooling3D': '<span class="label label-success">{0:s}</span>',
'GlobalMaxPooling1D': '<span class="label label-success">{0:s}</span>',
'GlobalMaxPooling2D': '<span class="label label-success">{0:s}</span>',
'GlobalMaxPooling3D': '<span class="label label-success">{0:s}</span>',
'GlobalAveragePooling1D': '<span class="label label-success">{0:s}</span>',
'GlobalAveragePooling2D': '<span class="label label-success">{0:s}</span>',
'GlobalAveragePooling3D': '<span class="label label-success">{0:s}</span>',
'RNN': '<span class="label label-danger">{0:s}</span>',
'SimpleRNN': '<span class="label label-danger">{0:s}</span>',
'GRU': '<span class="label label-danger">{0:s}</span>',
'CuDNNGRU': '<span class="label label-danger">{0:s}</span>',
'LSTM': '<span class="label label-danger">{0:s}</span>',
'CuDNNLSTM': '<span class="label label-danger">{0:s}</span>',
'Bidirectional': '<span class="label label-danger">{0:s}</span>'
}
from tensorflow import keras
from distutils.version import LooseVersion
import tensorflow.keras.backend as keras_backend
table_data = {
'layer_type': [],
'output': [],
'parameter_count': [],
'name': [],
'connected_to': [],
'activation': [],
'initialization': []
}
row_separators = []
prev_name = None
for layer_id, layer in enumerate(keras_model.layers):
connections = []
if LooseVersion(keras.__version__) >= LooseVersion('2.1.3'):
for node_index, node in enumerate(layer._inbound_nodes):
for i in range(len(node.inbound_layers)):
inbound_layer = node.inbound_layers[i].name
inbound_node_index = node.node_indices[i]
inbound_tensor_index = node.tensor_indices[i]
connections.append(
inbound_layer + '[' + str(inbound_node_index) + '][' + str(inbound_tensor_index) + ']'
)
else:
for node_index, node in enumerate(layer.inbound_nodes):
for i in range(len(node.inbound_layers)):
inbound_layer = node.inbound_layers[i].name
inbound_node_index = node.node_indices[i]
inbound_tensor_index = node.tensor_indices[i]
connections.append(
inbound_layer + '[' + str(inbound_node_index) + '][' + str(inbound_tensor_index) + ']'
)
config = DictContainer(layer.get_config())
layer_name = layer.__class__.__name__
if layer_name in layer_name_map:
layer_name = layer_name_map[layer_name]
if html_mode and layer_name in layer_type_html_tags:
layer_name = layer_type_html_tags[layer_name].format(layer_name)
if config.get_path('kernel_initializer.class_name') == 'VarianceScaling':
init = str(config.get_path('kernel_initializer.config.distribution', '---'))
elif config.get_path('kernel_initializer.class_name') == 'RandomUniform':
init = 'uniform'
else:
init = '-'
name_parts = layer.name.split('_')
if prev_name != name_parts[0]:
row_separators.append(layer_id)
prev_name = name_parts[0]
table_data['layer_type'].append(layer_name)
table_data['output'].append(str(layer.output_shape))
table_data['parameter_count'].append(str(layer.count_params()))
table_data['name'].append(layer.name)
table_data['connected_to'].append(str(connections[0]) if len(connections) > 0 else '-')
table_data['activation'].append(str(config.get('activation', '-')))
table_data['initialization'].append(init)
trainable_count = int(
numpy.sum([keras_backend.count_params(p) for p in set(keras_model.trainable_weights)])
)
non_trainable_count = int(
numpy.sum([keras_backend.count_params(p) for p in set(keras_model.non_trainable_weights)])
)
# Show row separators only if they are useful
if len(row_separators) == len(keras_model.layers):
row_separators = None
if mode == 'extended':
output += ui.table(
cell_data=[table_data['name'], table_data['layer_type'], table_data['output'], table_data['parameter_count']],
column_headers=['Layer name', 'Layer type', 'Output shape', 'Parameters'],
column_types=['str30', 'str20', 'str25', 'str20'],
column_separators=[1, 2],
row_separators=row_separators,
indent=4
)
elif mode == 'extended_wide':
output += ui.table(
cell_data=[table_data['name'], table_data['layer_type'], table_data['output'], table_data['parameter_count'],
table_data['activation'], table_data['initialization']],
column_headers=['Layer name', 'Layer type', 'Output shape', 'Parameters', 'Act.', 'Init.'],
column_types=['str30', 'str20', 'str25', 'str20', 'str15', 'str15'],
column_separators=[1, 2, 3],
row_separators=row_separators,
indent=4
)
if show_parameters:
output += ui.line('') + '\n'
output += ui.line('Parameters', indent=4) + '\n'
output += ui.data(indent=6, field='Total', value=trainable_count + non_trainable_count) + '\n'
output += ui.data(indent=6, field='Trainable', value=trainable_count) + '\n'
output += ui.data(indent=6, field='Non-Trainable', value=non_trainable_count) + '\n'
else:
output_buffer = []
keras_model.summary(print_fn=output_buffer.append)
for line in output_buffer:
if is_jupyter():
output += ui.line('<code>'+line+'</code>', indent=4) + '\n'
else:
output += ui.line(line, indent=4) + '\n'
model_config = keras_model.get_config()
if show_parameters:
output += ui.line('') + '\n'
output += ui.line('Input', indent=4) + '\n'
output += ui.data(indent=6, field='Shape', value=keras_model.input_shape) + '\n'
output += ui.line('Output', indent=4) + '\n'
output += ui.data(indent=6, field='Shape', value=keras_model.output_shape) + '\n'
if isinstance(model_config, dict) and 'layers' in model_config:
output += ui.data(
indent=6,
field='Activation',
value=model_config['layers'][-1]['config'].get('activation')
) + '\n'
elif isinstance(model_config, list):
output += ui.data(
indent=6,
field='Activation',
value=model_config[-1].get('config', {}).get('activation')
) + '\n'
if display:
if is_jupyter():
from IPython.core.display import display, HTML
display(HTML(output))
else:
print(output)
else:
return output
def create_sequential_model(model_parameter_list, input_shape=None, output_shape=None, constants=None, return_functional=False):
"""Create sequential Keras model
Example parameters::
model_parameter_list = [
{
'class_name': 'Dense',
'config': {
'units': 'CONSTANT_B',
'kernel_initializer': 'uniform',
'activation': 'relu'
}
},
{
'class_name': 'Dropout',
'config': {
'rate': 0.2
}
},
{
'class_name': 'Dense',
'config': {
'units': 'CONSTANT_A' * 2,
'kernel_initializer': 'uniform',
'activation': 'relu'
}
},
{
'class_name': 'Dropout',
'config': {
'rate': 0.2
}
},
{
'class_name': 'Dense',
'config': {
'units': 'CLASS_COUNT',
'kernel_initializer': 'uniform',
'activation': 'softmax'
}
}
]
constants = {
'CONSTANT_A': 50,
'CONSTANT_B': 100
}
Parameters
----------
model_parameter_list : dict or DictContainer
Model parameters
input_shape : int
Size of the input layer
Default value None
output_shape : int
Size of the output layer
Default value None
constants : dict or DictContainer
Constants used in the model_parameter definitions.
Default value None
return_functional : bool
Convert sequential model into function model.
Default value False
Returns
-------
Keras model
"""
from tensorflow.keras.models import Sequential
keras_model = Sequential()
tuple_fields = [
'input_shape',
'kernel_size',
'pool_size',
'dims',
'target_shape',
'strides'
]
# Get constants for model
if constants is None:
constants = {}
if 'INPUT_SHAPE' not in constants and input_shape is not None:
constants['INPUT_SHAPE'] = input_shape
if 'OUTPUT_SHAPE' not in constants and output_shape is not None:
constants['OUTPUT_SHAPE'] = output_shape
if 'CLASS_COUNT' not in constants:
constants['CLASS_COUNT'] = output_shape
if 'FEATURE_VECTOR_LENGTH' not in constants:
constants['FEATURE_VECTOR_LENGTH'] = input_shape
def logger():
logger_instance = logging.getLogger(__name__)
if not logger_instance.handlers:
setup_logging()
return logger_instance
def process_field(value, constants_dict):
math_eval = SimpleMathStringEvaluator()
if isinstance(value, str):
sub_fields = value.split()
if len(sub_fields) > 1:
# Inject constants to math formula
for subfield_id, subfield in enumerate(sub_fields):
if subfield in constants_dict:
sub_fields[subfield_id] = str(constants_dict[subfield])
value = ''.join(sub_fields)
else:
# Inject constants
if value in constants_dict:
value = str(constants_dict[value])
return math_eval.eval(value)
elif isinstance(value, list):
processed_value_list = []
for item_id, item in enumerate(value):
processed_value_list.append(
process_field(
value=item,
constants_dict=constants_dict
)
)
return processed_value_list
else:
return value
# Inject constant into constants with equations
for field in list(constants.keys()):
constants[field] = process_field(
value=constants[field],
constants_dict=constants
)
# Setup layers
for layer_id, layer_setup in enumerate(model_parameter_list):
# Get layer parameters
layer_setup = DictContainer(layer_setup)
if 'config' not in layer_setup:
layer_setup['config'] = {}
# Get layer class
try:
layer_class = getattr(
importlib.import_module('tensorflow.keras.layers'),
layer_setup['class_name']
)
except AttributeError:
message = 'Invalid Keras layer type [{type}].'.format(
type=layer_setup['class_name']
)
logger().exception(message)
raise AttributeError(message)
# Inject constants
for config_field in list(layer_setup['config'].keys()):
layer_setup['config'][config_field] = process_field(
value=layer_setup['config'][config_field],
constants_dict=constants
)
# Convert lists into tuples
for field in tuple_fields:
if field in layer_setup['config'] and isinstance(layer_setup['config'][field], list):
layer_setup['config'][field] = tuple(layer_setup['config'][field])
# Inject input shape for Input layer if not given
if layer_id == 0 and layer_setup.get_path('config.input_shape') is None and input_shape is not None:
# Set input layer dimension for the first layer if not set
layer_setup['config']['input_shape'] = (input_shape,)
if 'wrapper' in layer_setup:
# Get layer wrapper class
try:
wrapper_class = getattr(
importlib.import_module("tensorflow.keras.layers"),
layer_setup['wrapper']
)
except AttributeError:
message = 'Invalid Keras layer wrapper type [{type}].'.format(
type=layer_setup['wrapper']
)
logger().exception(message)
raise AttributeError(message)
wrapper_parameters = layer_setup.get('config_wrapper', {})
if layer_setup.get('config'):
keras_model.add(
wrapper_class(layer_class(**dict(layer_setup.get('config'))), **dict(wrapper_parameters)))
else:
keras_model.add(wrapper_class(layer_class(), **dict(wrapper_parameters)))
else:
if layer_setup.get('config'):
keras_model.add(layer_class(**dict(layer_setup.get('config'))))
else:
keras_model.add(layer_class())
if return_functional:
from tensorflow.keras.layers import Input
from tensorflow.keras.models import Model
input_layer = Input(batch_shape=keras_model.layers[0].input_shape)
prev_layer = input_layer
for layer in keras_model.layers:
prev_layer = layer(prev_layer)
keras_model = Model(
inputs=[input_layer],
outputs=[prev_layer]
)
return keras_model
def model_summary_string(keras_model, mode='keras'):
"""Model summary in a formatted string, similar to Keras model summary function.
Parameters
----------
keras_model : keras model
Keras model
mode : str
Summary mode ['extended', 'keras']. In case 'keras', standard Keras summary is returned.
Default value keras
Returns
-------
str
Model summary
"""
ui = FancyStringifier()
output = ''
output += ui.line('Model summary') + '\n'
if mode == 'extended':
layer_name_map = {
'BatchNormalization': 'BatchNorm',
}
import keras
from distutils.version import LooseVersion
import keras.backend as keras_backend
output += ui.row('Layer type',
'Output',
'Param',
'Name',
'Connected to',
'Activ.',
'Init',
widths=[15, 25, 10, 20, 25, 10, 10],
indent=4) + '\n'
output += ui.row('-', '-', '-', '-', '-', '-', '-') + '\n'
for layer in keras_model.layers:
connections = []
if LooseVersion(keras.__version__) >= LooseVersion('2.1.3'):
for node_index, node in enumerate(layer._inbound_nodes):
for i in range(len(node.inbound_layers)):
inbound_layer = node.inbound_layers[i].name
inbound_node_index = node.node_indices[i]
inbound_tensor_index = node.tensor_indices[i]
connections.append(inbound_layer + '[' +
str(inbound_node_index) + '][' +
str(inbound_tensor_index) + ']')
else:
for node_index, node in enumerate(layer.inbound_nodes):
for i in range(len(node.inbound_layers)):
inbound_layer = node.inbound_layers[i].name
inbound_node_index = node.node_indices[i]
inbound_tensor_index = node.tensor_indices[i]
connections.append(inbound_layer + '[' +
str(inbound_node_index) + '][' +
str(inbound_tensor_index) + ']')
config = DictContainer(layer.get_config())
layer_name = layer.__class__.__name__
if layer_name in layer_name_map:
layer_name = layer_name_map[layer_name]
if config.get_path(
'kernel_initializer.class_name') == 'VarianceScaling':
init = str(
config.get_path('kernel_initializer.config.distribution',
'---'))
elif config.get_path(
'kernel_initializer.class_name') == 'RandomUniform':
init = 'uniform'
else:
init = '---'
output += ui.row(
layer_name, str(layer.output_shape), str(layer.count_params()),
str(layer.name),
str(connections[0]) if len(connections) > 0 else '---',
str(config.get('activation', '---')), init) + '\n'
trainable_count = int(
numpy.sum([
keras_backend.count_params(p)
for p in set(keras_model.trainable_weights)
]))
non_trainable_count = int(
numpy.sum([
keras_backend.count_params(p)
for p in set(keras_model.non_trainable_weights)
]))
output += ui.line('') + '\n'
output += ui.line(
'Parameters',
indent=4,
) + '\n'
output += ui.data(indent=6,
field='Total',
value=trainable_count + non_trainable_count) + '\n'
output += ui.data(indent=6, field='Trainable',
value=trainable_count) + '\n'
output += ui.data(
indent=6, field='Non-Trainable', value=non_trainable_count) + '\n'
else:
output_buffer = []
keras_model.summary(print_fn=output_buffer.append)
for line in output_buffer:
output += ui.line(line, indent=4) + '\n'
output += ui.line('') + '\n'
output += ui.data(
indent=4, field='Input shape', value=keras_model.input_shape) + '\n'
output += ui.data(
indent=4, field='Output shape', value=keras_model.output_shape) + '\n'
return output
