Background execution

During training using auto train I switched to VPN and my space auto paused. Is there way to continue where training was stopped or enable background execution until an error or done even with a closed url of the space??

i dont think auto-pausing of space is related to vpn. a training space will auto-pause when training is finished (or encountered an error). you can see the logs by clicking the logs button in the paused space. the training runs even when you have closed the tab/window.

@abhishek I’ve looked through logs, there was “network error”. Is there a way I can see that run logs somewhere to analyse them again?

refreshing clears the message network error, else please provide space id and we can look internally.

@abhishek https://huggingface.co/spaces/FoodIntake/portion-to-quantity-unit-task, I don’t know what do you mean under refreshing. Its ok, I have restarted, just wanted to clarify, better use your time on something more important.

no. i just meant ctrl+r, but anyways, lets see if it works now. btw, if it worked previously, you could also see the models in your account.

🚀 INFO   | 2024-02-02 14:32:04 | __main__:train:218 - Pushing model to hub...
❌ ERROR  | 2024-02-02 14:32:04 | autotrain.trainers.common:wrapper:91 - train has failed due to an exception: Traceback (most recent call last):
  File "/app/env/lib/python3.10/site-packages/autotrain/trainers/common.py", line 88, in wrapper
    return func(*args, **kwargs)
  File "/app/env/lib/python3.10/site-packages/autotrain/trainers/seq2seq/__main__.py", line 220, in train
    api.create_repo(repo_id=config.repo_id, repo_type="model", private=True)
  File "/app/env/lib/python3.10/site-packages/huggingface_hub/utils/_validators.py", line 110, in _inner_fn
    validate_repo_id(arg_value)
  File "/app/env/lib/python3.10/site-packages/huggingface_hub/utils/_validators.py", line 158, in validate_repo_id
    raise HFValidationError(
huggingface_hub.utils._validators.HFValidationError: Repo id must be in the form 'repo_name' or 'namespace/repo_name': 'FoodIntake/FoodIntake/portion-to-quantity-unit-task'. Use `repo_type` argument if needed.

❌ ERROR  | 2024-02-02 14:32:04 | autotrain.trainers.common:wrapper:92 - Repo id must be in the form 'repo_name' or 'namespace/repo_name': 'FoodIntake/FoodIntake/portion-to-quantity-unit-task'. Use `repo_type` argument if needed.
🚀 INFO   | 2024-02-02 14:32:04 | autotrain.trainers.common:pause_space:49 - Pausing space...

Oh that’s bad, there’s no verification at the start. So I have spent processing time to see that my model can’t be saved

did your project name consist of a “/” ?

Yes seems like I have accidentally copy-pasted full path to the space instead of just space name. But this validation should be at the beginning I think not after an hour of training at saving.

indeed. ill add it. thanks for reporting.