callbacks

A callback refers to a set of functions that are automatically invoked after the completion of each training epoch. At a minimum, you need at least one callback for saving the model. If you have any special requirements, or if you want to use any built-in features provided by Lightning, you can add them here as callbacks.

Example

callbacks:
early_stopping:
    _target_: lightning.pytorch.callbacks.EarlyStopping
    verbose: true
    log_rank_zero_only: true
    monitor: ${synth_metric.monitor_metric_name}
    # min_delta: 1e-5
    patience: 50

ema: # ema is always recommended
    _target_: tace.utils.callbacks.EMACallback
    decay: 0.999 # 0.99 - 0.999
    use_num_updates: true

checkpoint_epoch: # at leas one checkpoint is required
    _target_: lightning.pytorch.callbacks.ModelCheckpoint
    dirpath: checkpoints_epoch
    filename: TACE-{${misc.project_name}-{epoch}-{step}-{${synth_metric.monitor_metric_name}:.4f}
    # monitor: ${synth_metric.monitor_metric_name}
    save_top_k: -1
    save_last: true
    every_n_epochs: 10
    # mode: min
    save_weights_only: false
    auto_insert_metric_name: false
    verbose: false

# # If you also need to save checkpoints based on training steps, you can uncomment this section.
# checkpoint_step:
#   _target_: lightning.pytorch.callbacks.ModelCheckpoint
#   dirpath: checkpoints_step
#   filename: TACE-{${misc.project_name}-{epoch}-{step}
#   save_top_k: -1
#   save_last: false
#   every_n_train_steps: 10000
#   save_weights_only: false
#   auto_insert_metric_name: false
#   verbose: false