xlmr-large-ar-CLS-B / test_eval.txt
HHansi's picture
Upload folder using huggingface_hub
a463de7 verified
Default classification report:
precision recall f1-score support
F 0.6811 0.6620 0.6714 500
T 0.6712 0.6900 0.6805 500
accuracy 0.6760 1000
macro avg 0.6761 0.6760 0.6759 1000
weighted avg 0.6761 0.6760 0.6759 1000
ADJ
Accuracy = 0.6632653061224489
Weighted Recall = 0.6632653061224489
Weighted Precision = 0.6779640427599611
Weighted F1 = 0.6621071774202123
Macro Recall = 0.6702306079664571
Macro Precision = 0.6726190476190477
Macro F1 = 0.6629494528400208
ADV
Accuracy = 0.9
Weighted Recall = 0.9
Weighted Precision = 0.9111111111111111
Weighted F1 = 0.8862745098039216
Macro Recall = 0.75
Macro Precision = 0.9444444444444444
Macro F1 = 0.803921568627451
NOUN
Accuracy = 0.6720647773279352
Weighted Recall = 0.6720647773279352
Weighted Precision = 0.6773406973555232
Weighted F1 = 0.6702285129058585
Macro Recall = 0.673
Macro Precision = 0.6769450033534541
Macro F1 = 0.6705039525691701
VERB
Accuracy = 0.678391959798995
Weighted Recall = 0.678391959798995
Weighted Precision = 0.6873837887630304
Weighted F1 = 0.6751217863015382
Macro Recall = 0.6794201580927848
Macro Precision = 0.6869556064314097
Macro F1 = 0.6754332313965341