Modelpublic

ProrabVasili/roberta_ukr-psyop-6_3

sourceHugging Faceupdated 3y agoView on Hugging Face

0likes9downloads

Model Card

robertaukr-psyop-63

This model is a fine-tuned version of youscan/ukr-roberta-base on an unknown dataset. It achieves the following results on the evaluation set:

—Loss: 0.0326
—Accuracy: 0.9929
—Precision: 0.9951
—Recall: 0.9906
—F1-score: 0.9928
—Matthews Corrcoef: 0.9857

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

—learning_rate: 5e-05
—trainbatchsize: 14
—evalbatchsize: 14
—seed: 42
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—lrschedulerwarmup_steps: 500
—num_epochs: 6

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1-score	Matthews Corrcoef
0.0545	1.0	1787	0.0350	0.9904	0.9849	0.9961	0.9905	0.9809
0.0001	2.0	3574	0.0512	0.9925	0.9925	0.9925	0.9925	0.9851
0.0069	3.0	5361	0.0326	0.9929	0.9951	0.9906	0.9928	0.9857
0.0	4.0	7148	0.0714	0.9912	0.9874	0.9951	0.9913	0.9825
0.0	5.0	8935	0.0709	0.9925	0.9961	0.9890	0.9925	0.9851
0.0	6.0	10722	0.0629	0.9933	0.9932	0.9935	0.9933	0.9867

Framework versions

—Transformers 4.38.2
—Pytorch 2.2.1+cu121
—Datasets 2.18.0
—Tokenizers 0.15.2