--label-stats for spaCy train

following the steps described in training with transformer model, I wanted to train my custom ner model in spaCy V3 (involving transformer models)
All steps from data conversion, training work fine:

python -m prodigy data-to-spacy ./corpus --ner train_data --eval-split 0.3
python -m spacy train config.cfg --output ./output --paths.train ./train --paths.dev ./dev

However, in addition I wanted to get the metrics (precision, recall and F-measuer) also per entity type.
I tried the following argument we use with prodigy train recipe:


But it doesn't work. Could you tell me how to get in addition the metrics per entity type using the command line?

I found the solution in this pot: NER Trained Model Analysis

1 Like

Glad to hear you found the answer and thanks for sharing the link :slight_smile: