Is it possible to implement a precision recall curve in rasa? Thanks
@revet Hi, I guess rasa have inbuilt one, did you visited this link: Testing Your Assistant Or you want something else, please elaborate more.
I suggest you use Tensorboard to make comparisons and choose an optimal configuration. This is doable on DIET, ResponseSelector, and TED like so for example:
- name: DIETClassifier
// other parameters
evaluate_on_number_of_examples: 200
evaluate_every_number_of_epochs: 5
tensorboard_log_directory: ./tensorboard/DIET
tensorboard_log_level: epoch
Try to set evaluate_on_number_of_examples
to about 20% of your total number of examples (of course, this means these examples will not be used for training and you will have to give a bit more examples). You can use this script I wrote to count the number of examples you have.
I would also recommend setting a random_seed
: If you want to accurately compare two Pipeline Components or Policies across multiple trainings, you could set a Seed for DIET, ResponseSelector, and TED like so for example:
- name: DIETClassifier
random_seed: 1
// other parameters