Modelpublic

npark95/finetuned_ClinicalLongformer_negsampling

sourceHugging Faceupdated 2y agoView on Hugging Face

0likes16downloads

Model Card

finetunedClinicalLongformernegsampling

This model is a fine-tuned version of yikuan8/Clinical-Longformer on the None dataset. It achieves the following results on the evaluation set:

—Loss: 0.4650
—F1: 0.8741

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: 1e-05
—trainbatchsize: 8
—evalbatchsize: 8
—seed: 42
—optimizer: Use adamwtorch with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
—lrschedulertype: linear
—num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	F1
0.6855	1.0	38	0.5065	0.7731
0.4446	2.0	76	0.3917	0.8676
0.3021	3.0	114	0.4088	0.8676
0.2602	4.0	152	0.3813	0.8361
0.2525	5.0	190	0.5171	0.7788
0.2292	6.0	228	0.4650	0.8741
0.1429	7.0	266	0.5181	0.8571
0.1025	8.0	304	0.5355	0.8507
0.1185	9.0	342	0.6183	0.8333
0.0565	10.0	380	0.7438	0.8281
0.0886	11.0	418	0.8736	0.8209
0.0143	12.0	456	0.8986	0.8254
0.051	13.0	494	0.9389	0.8130
0.0254	14.0	532	0.8966	0.8333
0.0121	15.0	570	0.9377	0.8333

Framework versions

—Transformers 4.48.0
—Pytorch 2.5.1+cu124
—Datasets 3.2.0
—Tokenizers 0.21.0