trainer

For a complete list of all parameters and their detailed descriptions, please refer to the official PyTorch Lightning Trainer documentation: https://lightning.ai/docs/pytorch/stable/common/trainer.html

Example

trainer:
  # As long as your system has NVIDIA GPUs, the following configuration will automatically detect and use all available GPUs for training.
  # If you wish to control which GPUs are visible, you can do so by setting the environment variable like export CUDA_VISIBLE_DEVICES=0,2.
  _target_: lightning.Trainer
  num_nodes: 1
  accelerator: auto
  devices: auto        # auto or list
  max_time: '90:00:00:00' # 90 days
  max_epochs: 20000
  min_epochs: 1        # null or int > 1
  precision: 32        # only 32 or 64 are allowed, not allow bf16, fp16 ....
  strategy: auto # For single-GPU/multi-node training, use auto
  # strategy: # Recommended for single-node training.
  #   _target_: tace.utils.strategy.SimpleDDPStrategy
  gradient_clip_val: 10.0
  enable_progress_bar: true # always set this to true to show progress_bar
  log_every_n_steps: 1000

  # Generally, no modification is required
  enable_model_summary: true
  enable_checkpointing: true
  check_val_every_n_epoch: 1
  detect_anomaly: false
  inference_mode: false
  deterministic: false
  # benchmark: true