Modelpublic

SMG0/Model3_Marabertv2_T2_WS_A100

sourceHugging Faceupdated 3y agoView on Hugging Face

0likes9downloads

Model Card

Model3Marabertv2T2WSA100

This model is a fine-tuned version of UBC-NLP/MARBERTv2 on the None dataset. It achieves the following results on the evaluation set:

—Loss: 0.0795
—F1: 0.8271
—Roc Auc: 0.9092
—Accuracy: 0.7430

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: 16
—evalbatchsize: 16
—seed: 42
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Roc Auc	Accuracy
No log	1.0	193	0.1589	0.5089	0.6835	0.3687
No log	2.0	386	0.1122	0.6879	0.7800	0.5680
0.1646	3.0	579	0.0932	0.7846	0.8623	0.6816
0.1646	4.0	772	0.0793	0.8138	0.8823	0.7318
0.1646	5.0	965	0.0916	0.7893	0.8787	0.6965
0.0598	6.0	1158	0.0739	0.8251	0.8935	0.7318
0.0598	7.0	1351	0.0752	0.8257	0.9002	0.7393
0.0266	8.0	1544	0.0800	0.8350	0.9128	0.7467
0.0266	9.0	1737	0.0810	0.8295	0.9123	0.7430
0.0266	10.0	1930	0.0795	0.8271	0.9092	0.7430

Framework versions

—Transformers 4.31.0
—Pytorch 2.0.1+cu118
—Datasets 2.14.4
—Tokenizers 0.13.3