Explain results

Rasa Open Source Feedback on Rasa Open Source
1

Hi everyone,

I am working on NLU customized pipeline for the Arabic language. I am trying to test different components and test the performance based on them.

rasa test nlu –config Tensorflow.yml mBERT.yml –nlu data/nlu.md –runs 3 –percentages 0 25 50 75 85 I got results as follows, So how did I explain it, is what

  • Accuracy: 0.8284313725490197,

  • F1-score: 0.7765795206971677,

  • Precision: 0.4318717230481936,

  • Recall: 0.32712418300653595

    “mBert”: [ [ 0.8284313725490197, 0.7765795206971677, 0.4318717230481936, 0.32712418300653595 ],

    [ 0.8503267973856208, 0.7988017429193901, 0.5104129530600119, 0.24694989106753815 ],

    [ 0.8457516339869281, 0.8441176470588235, 0.5795051353874883, 0.3290227201991908 ] ],

    “Tensorflow”: [ [ 0.7052287581699347, 0.7263460939931529, 0.307433579492403, 0.23715794009911653 ],

    [ 0.744226579520697, 0.5883442265795207, 0.39325396825396824, 0.2694819341878166 ],

    [ 0.6839869281045752, 0.6160130718954249, 0.3459694989106754, 0.20746187363834423 ] ] }