Modelpublic

oliverhoffmann/Certificate_Classification_with_20_epochs

sourceHugging Faceapache-2.0updated 1y agoView on Hugging Face

0likes4downloads

Model Card

CertificateClassificationwith20epochs

This model is a fine-tuned version of bert-base-uncased on the None dataset. It achieves the following results on the evaluation set:

—Loss: 0.2528
—Accuracy: 0.9471
—Precision: 0.9486
—Recall: 0.9471
—F1: 0.9471
—F1 Macro: 0.9615

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: 2e-05
—trainbatchsize: 16
—evalbatchsize: 16
—seed: 42
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1	F1 Macro
1.3788	1.0	90	0.6256	0.9182	0.9209	0.9182	0.9140	0.9124
0.3441	2.0	180	0.2432	0.9371	0.9366	0.9371	0.9363	0.9533
0.1447	3.0	270	0.1730	0.9497	0.9491	0.9497	0.9489	0.9623
0.099	4.0	360	0.1503	0.9560	0.9581	0.9560	0.9546	0.9664
0.0906	5.0	450	0.1989	0.9371	0.9366	0.9371	0.9363	0.9533
0.0826	6.0	540	0.1716	0.9560	0.9581	0.9560	0.9546	0.9664
0.066	7.0	630	0.1636	0.9497	0.9501	0.9497	0.9484	0.9619
0.0699	8.0	720	0.2132	0.9497	0.9535	0.9497	0.9479	0.9591
0.0647	9.0	810	0.2248	0.9434	0.9428	0.9434	0.9422	0.9575
0.0494	10.0	900	0.2294	0.9434	0.9428	0.9434	0.9422	0.9575
0.0564	11.0	990	0.2419	0.9497	0.9501	0.9497	0.9484	0.9619
0.0492	12.0	1080	0.2445	0.9434	0.9428	0.9434	0.9422	0.9575
0.0447	13.0	1170	0.2695	0.9497	0.9501	0.9497	0.9484	0.9619
0.0505	14.0	1260	0.2551	0.9497	0.9501	0.9497	0.9484	0.9619
0.0391	15.0	1350	0.2787	0.9497	0.9501	0.9497	0.9484	0.9619
0.0457	16.0	1440	0.2624	0.9434	0.9428	0.9434	0.9422	0.9575
0.0396	17.0	1530	0.2898	0.9497	0.9501	0.9497	0.9484	0.9619
0.043	18.0	1620	0.2830	0.9497	0.9501	0.9497	0.9484	0.9619
0.0356	19.0	1710	0.2893	0.9497	0.9501	0.9497	0.9484	0.9619
0.0403	20.0	1800	0.2888	0.9434	0.9428	0.9434	0.9422	0.9575

Framework versions

—Transformers 4.40.1
—Pytorch 2.6.0+cu124
—Datasets 2.14.4
—Tokenizers 0.19.1