Modelpublic

Wariano/bsc-bio-ehr-es-vih-10k

sourceHugging Faceapache-2.0updated 3y agoView on Hugging Face

0likes7downloads

Model Card

bsc-bio-ehr-es-vih-10k

This model is a fine-tuned version of PlanTL-GOB-ES/bsc-bio-ehr-es on the None dataset. It achieves the following results on the evaluation set:

—Loss: 0.9958
—Positives Preds: 598
—Negative Preds: 402
—Positives Refs: 500
—Negative Refs: 500
—Tp: 411
—Fn: 89
—Fp: 187
—Tn: 313
—Accuracy: 0.724
—Precision: 0.6873
—Recall: 0.822
—F1: 0.7486

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: 3e-05
—trainbatchsize: 32
—evalbatchsize: 32
—seed: 42
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Positives Preds	Negative Preds	Positives Refs	Negative Refs	Tp	Fn	Fp	Tn	Accuracy	Precision	Recall	F1
0.4309	1.0	250	0.4999	384	616	500	500	316	184	68	432	0.748	0.8229	0.632	0.7149
0.2849	2.0	500	0.6391	546	454	500	500	396	104	150	350	0.746	0.7253	0.792	0.7572
0.1931	3.0	750	0.7333	610	390	500	500	414	86	196	304	0.718	0.6787	0.828	0.7459
0.1255	4.0	1000	0.8917	604	396	500	500	417	83	187	313	0.73	0.6904	0.834	0.7554
0.0918	5.0	1250	0.9958	598	402	500	500	411	89	187	313	0.724	0.6873	0.822	0.7486

Framework versions

—Transformers 4.30.2
—Pytorch 2.0.1+cu117
—Datasets 2.13.1
—Tokenizers 0.13.3