Can't "stop" training - out of input

Bugs and Issues
#1

This error occurred, and I thought nothing of it and simply decided to restart the server etc. - but it occurred again immediately when the PL tab was opened, without clicking the Stop button.

I can’t stop this model in order to run it again :slight_smile: Ideas?

UPDATE restarted with -v=3 and have appended the log
UPDATE 2 There is no update, so why is there an update prompt in the top right? (I am on 0.13.1 and that is the latest according to pypi)
UPDATE 3 Now the error keeps repeating and PL is unusable because it keeps occurring… will try restarting again and will reopen in a private browsing tab :wink: Error still occurs - just not continuously… unless one tries stopping by also clicking the stop SQUARE, it seems

EOFError('Ran out of input')
Traceback (most recent call last):
  File "c:\users\julian\anaconda3\envs\pl_tf250_py3810\lib\site-packages\flask\app.py", line 1513, in full_dispatch_request
    rv = self.dispatch_request()
  File "c:\users\julian\anaconda3\envs\pl_tf250_py3810\lib\site-packages\flask\app.py", line 1499, in dispatch_request
    return self.ensure_sync(self.view_functions[rule.endpoint])(**req.view_args)
  File "perceptilabs\models\api.py", line 75, in perceptilabs.models.api.create_blueprint.training_status
  File "perceptilabs\models\interface.py", line 42, in perceptilabs.models.interface.ModelsInterface.get_training_status
  File "perceptilabs\resources\training_results.py", line 28, in perceptilabs.resources.training_results.TrainingResultsAccess.get_latest
  File "perceptilabs\resources\training_results.py", line 29, in perceptilabs.resources.training_results.TrainingResultsAccess.get_latest
  File "perceptilabs\resources\training_results.py", line 30, in perceptilabs.resources.training_results.TrainingResultsAccess.get_latest
EOFError: Ran out of input

Out of Input error log.txt (86.8 KB)

#2

This is gonna be a little hacky, but in your model.json you should have a field called “coreStatus” with an inner key called “Status”. Try changing the value from whatever it is at the moment (“Training” for example) to “Waiting” and see if that helps :slight_smile:
You also have a field called “error” where you can change the value to null (no “” around it) in case the first step doesn’t fix it.
(We have a fix coming in next release which will stop tying the status to the model.json that will make this a lot more stable)

Finally, are you getting this error constantly at the moment, or when doing key actions?

#3

I just updated with some more info… it’s the end of the week - I might wait for the new version rather than do the hacky thing :slight_smile:

#4

Yes, changing status to “Waiting” made the problem go away Call me persistent/obsessive - I just had to try.

1 Like