Skip to content
Shuai Yan edited this page Sep 24, 2024 · 2 revisions

How to train

Tips

  1. For training on Slurm-based HPCs, to enable multiple GPUs for training, the command in training_run.sh should be srun python perform_training.py --condensed_config_path config.toml. For more details, please refer to Training with Pytorch Lightning on Slurm.
Clone this wiki locally