Yepes_0.0001_250

This model is a fine-tuned version of microsoft/BiomedNLP-PubMedBERT-base-uncased-abstract-fulltext on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1555
• Precision: 0.5922
• Recall: 0.4552
• F1: 0.5148
• Accuracy: 0.9768

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: 0.0001
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• training_steps: 500

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.4065	1.39	25	0.2115	0.0	0.0	0.0	0.9672
0.1995	2.78	50	0.2120	0.0	0.0	0.0	0.9672
0.1995	4.17	75	0.2108	0.0	0.0	0.0	0.9672
0.1694	5.56	100	0.1646	0.0	0.0	0.0	0.9672
0.1493	6.94	125	0.1513	0.0	0.0	0.0	0.9672
0.1266	8.33	150	0.1446	0.0	0.0	0.0	0.9672
0.106	9.72	175	0.1396	0.4019	0.2139	0.2792	0.9704
0.086	11.11	200	0.1162	0.5037	0.3408	0.4065	0.9740
0.0613	12.5	225	0.1230	0.5015	0.4104	0.4514	0.9740
0.047	13.89	250	0.1306	0.5333	0.4378	0.4809	0.9753
0.0351	15.28	275	0.1351	0.5629	0.4453	0.4972	0.9757
0.0266	16.67	300	0.1453	0.5617	0.4303	0.4873	0.9765
0.02	18.06	325	0.1441	0.5573	0.4478	0.4966	0.9757
0.0153	19.44	350	0.1555	0.5922	0.4552	0.5148	0.9768

Framework versions

• Transformers 4.27.4
• Pytorch 1.13.1+cu116
• Datasets 2.11.0
• Tokenizers 0.13.2