Train results accuracy dont appears for 1.11.0

, ,
1

With the new prodigy version 1.11.0, when the text classification training is done don't appear the accuracy for each category

#17 245.1 ============================= Training pipeline =============================
#17 245.1 ℹ Pipeline: ['textcat']
#17 245.1 ℹ Initial learn rate: 0.001
#17 245.1 E    #       LOSS TEXTCAT  CATS_SCORE  SCORE
#17 245.1 ---  ------  ------------  ----------  ------
#17 245.1   0       0          0.00        0.00    0.00
#17 245.1  40     200          0.11        2.56    0.03
#17 245.1  90     400          0.05       10.71    0.11
#17 245.1 151     600          0.05       10.71    0.11
#17 245.1 217     800          0.03       10.71    0.11
#17 245.1 312    1000          0.02       10.71    0.11
#17 245.1 412    1200          0.01       10.71    0.11
#17 245.1 512    1400          0.02       10.71    0.11
#17 245.1 693    1600          0.01       16.27    0.16
#17 245.1 893    1800          0.00       16.27    0.16
#17 245.1 1093    2000          0.00       19.30    0.19
#17 245.1 1293    2200          0.00       20.41    0.20
#17 245.1 1493    2400          0.00       20.41    0.20
#17 245.1 1693    2600          0.00       18.51    0.19
#17 245.1 1893    2800          0.00       18.25    0.18
#17 245.1 2093    3000          0.00       18.25    0.18
#17 245.1 2293    3200          0.00       18.25    0.18
#17 245.1 2493    3400          0.00       18.25    0.18
#17 245.1 2693    3600          0.00       18.25    0.18
#17 245.1 2893    3800          0.00       18.25    0.18
#17 245.1 ✔ Saved pipeline to output directory
#17 245.1 temp_model/model-last
#17 DONE 245.3s

What is the new way to see model accuracy?

2

Hi! The "score" column shown here is the overall accuracy score of your model. If you want to see a per-label breakdown of the individual labels, you can set the --label-stats flag. Also see the docs of the new train command API: https://prodi.gy/docs/recipes#train

3

Thx works perfectly