Edit model card

albertmartinez/bert-multilingual-sdg-classification

This model (BERT) is for classifying text with respect to the United Nations sustainable development goals (SDG).

Training results

epoch eval_loss eval_accuracy eval_precision eval_recall eval_f1
1 0.932697594165802 0.7336408412164803 0.7360917875267202 0.7336408412164803 0.7251914694909984
2 0.7605898976325989 0.777343444609491 0.7831176862515112 0.777343444609491 0.777992037469195
3 0.7255606651306152 0.7918849190837307 0.7922937492519824 0.7918849190837307 0.7897436793752113
4 0.7620322108268738 0.782581502619029 0.791463371337951 0.782581502619029 0.7836557631495363
5 0.7925569415092468 0.7989211164099758 0.7987793726148532 0.7989211164099758 0.7976319242111907
6 0.8901194930076599 0.8008756156672661 0.8026922096228485 0.8008756156672661 0.7999251604304751
7 0.9644309878349304 0.797044797122977 0.7998531076533889 0.797044797122977 0.7979133030034837
8 1.0754749774932861 0.8050973340630131 0.8031314135784893 0.8050973340630131 0.8026788633046699
9 1.1106163263320923 0.8066609334688453 0.806263013139723 0.8066609334688453 0.8061921653404509
10 1.1396784782409668 0.8066609334688453 0.8056981354865757 0.8066609334688453 0.8059943909432397
Downloads last month
67
Safetensors
Model size
167M params
Tensor type
F32
·

Dataset used to train albertmartinez/bert-multilingual-sdg-classification