Modelpublic

lupobricco/relation_classification_correlations

sourceHugging Faceupdated 2y agoView on Hugging Face

0likes8downloads

Model Card

relationclassificationcorrelations

This model is a fine-tuned version of Musixmatch/umberto-commoncrawl-cased-v1 on the None dataset. It achieves the following results on the evaluation set:

—Loss: 0.4935
—F1: 0.6391
—Roc Auc: 0.6959
—Accuracy: 0.5426

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: 10

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Roc Auc	Accuracy
No log	1.0	33	0.5801	0.2301	0.5	0.5271
No log	2.0	66	0.5229	0.4619	0.6167	0.5969
No log	3.0	99	0.5122	0.4834	0.6295	0.6202
No log	4.0	132	0.5035	0.5805	0.6626	0.5814
No log	5.0	165	0.5133	0.5608	0.6503	0.5736
No log	6.0	198	0.4935	0.6391	0.6959	0.5426
No log	7.0	231	0.5092	0.6170	0.6784	0.5504
No log	8.0	264	0.5135	0.6194	0.6780	0.5814
No log	9.0	297	0.5189	0.6159	0.6790	0.5426
No log	10.0	330	0.5217	0.6066	0.6693	0.5659

Framework versions

—Transformers 4.40.1
—Pytorch 2.3.0+cu118
—Datasets 2.19.0
—Tokenizers 0.19.1