flair.trainers.plugins.LossFilePlugin#
- class flair.trainers.plugins.LossFilePlugin(base_path, epoch, metrics_to_collect=None)View on GitHub#
Bases:
TrainerPlugin
Plugin that manages the loss.tsv file output.
- __init__(base_path, epoch, metrics_to_collect=None)View on GitHub#
Initialize the base plugin.
Methods
__init__
(base_path, epoch[, metrics_to_collect])Initialize the base plugin.
after_evaluation
(epoch, **kw)This prints all relevant metrics.
attach_to
(pluggable)Attach this plugin to a Pluggable.
before_training_epoch
(epoch, **kw)Get the current epoch for loss file logging.
detach
()Detach a plugin from the Pluggable it is attached to.
hook
([first_arg])Convience function for BasePlugin.mark_func_as_hook).
mark_func_as_hook
(func, *events)Mark method as a hook triggered by the Pluggable.
metric_recorded
(record)Add the metric of a record to the current row.
Attributes
If set, the plugin will be attached to all processes when distributed, not just the main process.
corpus
model
pluggable
trainer
- get_state()View on GitHub#
- Return type:
dict
[str
,Any
]
- before_training_epoch(epoch, **kw)View on GitHub#
Get the current epoch for loss file logging.
- metric_recorded(record)View on GitHub#
Add the metric of a record to the current row.
- after_evaluation(epoch, **kw)View on GitHub#
This prints all relevant metrics.
- property attach_to_all_processes: bool#
If set, the plugin will be attached to all processes when distributed, not just the main process.