Restart a fine-tuning or transfer learning job

[rajnichahal]

Hi, my question is how do we restart a fine-tuning or transfer learning job in deepmd-se if it can't be finished in one go?
e.g., a regular training from scratch can be restarted using "dp train --restart <input_script>"

Following similar for restarting a fine-tuning job seems to be incompatible:
dp train: error: argument -t/--finetune: not allowed with argument -r/--restart

Thanks!

[njzjz]

a regular training from scratch can be restarted using "dp train --restart <input_script>"

The same command for your case.

[rajnichahal]

Thanks. Got it!