Modelpublic

alionder/acer_nitro_bert_turk

sourceHugging Facemitupdated 3y agoView on Hugging Face

0likes6downloads

Model Card

acernitrobert_turk

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

—Loss: 0.3791
—F1: 0.8532
—Roc Auc: 0.9236
—Accuracy: 0.7108

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: 2
—evalbatchsize: 2
—seed: 42
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Roc Auc	Accuracy
No log	1.0	166	0.3705	0.8426	0.9139	0.7229
No log	2.0	332	0.3450	0.8584	0.9380	0.7470
No log	3.0	498	0.3661	0.8491	0.9128	0.7108
0.0532	4.0	664	0.3745	0.8558	0.9213	0.7229
0.0532	5.0	830	0.3598	0.8571	0.9249	0.7349
0.0532	6.0	996	0.3747	0.8571	0.9249	0.7229
0.0219	7.0	1162	0.3540	0.8624	0.9297	0.7349
0.0219	8.0	1328	0.3761	0.8584	0.9285	0.7229
0.0219	9.0	1494	0.3836	0.8493	0.9223	0.7108
0.0108	10.0	1660	0.3791	0.8532	0.9236	0.7108

Framework versions

—Transformers 4.36.2
—Pytorch 2.1.0+cu121
—Datasets 2.16.1
—Tokenizers 0.15.0