pytorch_lightning.trainer.logging module

class pytorch_lightning.trainer.logging.TrainerLoggingMixin

Bases: abc.ABC

add_progress_bar_metrics(metrics)

configure_logger(logger)

log_metrics(metrics, grad_norm_dic, step=None)

Logs the metric dict passed in. If step parameter is None and step key is presented is metrics, uses metrics[“step”] as a step

Parameters

• metrics (dict) – Metric values

• grad_norm_dic (dict) – Gradient norms

• step (int) – Step for which metrics should be logged. Default value corresponds to self.global_step

metrics_to_scalars(metrics)

process_output(output, train=False)

Reduces output according to the training mode.

Separates loss from logging and progress bar metrics

reduce_distributed_output(output, num_gpus)

current_epoch: int = None

default_root_dir: str = None

global_rank: int = None

global_step: int = None

log_gpu_memory: ... = None

logged_metrics: ... = None

logger: Union[LightningLoggerBase, bool] = None

num_gpus: int = None

on_gpu: bool = None

progress_bar_metrics: ... = None

slurm_job_id: int = None

use_ddp2: bool = None

use_dp: bool = None