How can I use --resume_from_checkpoint?

[LJX010204]

The model breaks during training, I want the model to restart training at the checkpoints which are already saved. How can i use --resume_from_checkpoint? What necessary codes should I add to the train.xl?

[suyash-tyloones]

did you find a solution to this?

[Nadia269]

did you find a solution to this?

add this line parser.add_argument("--resume_from_checkpoint", type=str, default=None, help="Path to a checkpoint to resume training from") and in default parameter add your checkpoint path .