search

RasaHQ / rasa

💬 Open source machine learning framework to automate text- and voice-based conversations: NLU, dialogue management, connect to Slack, Facebook, and more - Create chatbots and voice assistants
https://rasa.com/docs/rasa/
Apache License 2.0
18.91k stars 4.63k forks source link

Unable to finetune on rc2 #10247

Closed dakshvar22 closed 2 years ago

dakshvar22 commented 2 years ago

Rasa Open Source version

3.0.0rc2

Rasa SDK version

No response

Rasa X version

No response

Python version

3.8

What operating system are you using?

OSX

What happened?

Trained an initial model with rasa init, added a new training example to one of the existing intents and ran rasa train --finetune. Fails with:

InvalidConfigException: Cannot finetune because more than just the 'epoch' keys have been changed in the configuration. Please revert your configuration and only change the 'epoch' settings where needed.

Might be related to https://github.com/RasaHQ/rasa/issues/10246

Command / Request

No response

Relevant log output

No response

dakshvar22 commented 2 years ago

@ka-bu I have more evidence for you... This only happens when you use the config that comes with autoconfiguration of rasa init as it is. If you uncomment the lines of the config but still use the same pipeline of the config in the above setup, finetuning will proceed as expected. I observed the same thing for this issue as well. Strongly suspect that the bug might be common across the two tickets.