dcase_framework.keras_utils.ProgressLoggerCallback

class dcase_framework.keras_utils.ProgressLoggerCallback(*args, **kwargs)[source]

Keras callback to store metrics with tqdm progress bar or logging interface. Implements Keras Callback API.

This callback is very similar to standard ProgbarLogger Keras callback, however it adds support for logging interface and tqdm based progress bars, and external metrics (metrics calculated outside Keras training process).

Constructor

Parameters:

epochs : int

Total amount of epochs

metric : str

Metric name

manual_update : bool

Manually update callback, use this to when injecting external metrics Default value True

manual_update_interval : int

Epoch interval for manual update, used anticipate updates Default value 1

disable_progress_bar : bool

Disable tqdm based progress bar Default value False

close_progress_bar : bool

Close tqdm progress bar on training end Default value True

log_progress : bool

Print progress into logging interface Default value False

external_metric_labels : dict or OrderedDict

Dictionary with

__init__(*args, **kwargs)[source]

Constructor

Parameters:

epochs : int

Total amount of epochs

metric : str

Metric name

manual_update : bool

Manually update callback, use this to when injecting external metrics Default value True

manual_update_interval : int

Epoch interval for manual update, used anticipate updates Default value 1

disable_progress_bar : bool

Disable tqdm based progress bar Default value False

close_progress_bar : bool

Close tqdm progress bar on training end Default value True

log_progress : bool

Print progress into logging interface Default value False

external_metric_labels : dict or OrderedDict

Dictionary with

Methods

__init__(\*args, \*\*kwargs) Constructor
add_external_metric(metric_id) Add external metric to be monitored
close() Manually close progress logging
get_operator(metric)
on_batch_begin(batch[, logs])
on_batch_end(batch[, logs])
on_epoch_begin(epoch[, logs])
on_epoch_end(epoch[, logs])
on_train_begin([logs])
on_train_end([logs])
set_external_metric_value(metric_label, ...) Add external metric value
set_model(model)
set_params(params)
update() Update
update_progress_bar([increase]) Update progress to tqdm progress bar
update_progress_log() Update progress to logging interface