Yepes_0.0001_29_03

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: 0.1553
• Precision: 0.5294
• Recall: 0.4091
• F1: 0.4615
• Accuracy: 0.9756

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
0.3757	5.0	25	0.1967	0.0	0.0	0.0	0.9697
0.1852	10.0	50	0.1952	0.0	0.0	0.0	0.9697
0.1766	15.0	75	0.1572	0.0	0.0	0.0	0.9697
0.1231	20.0	100	0.1386	0.0	0.0	0.0	0.9697
0.0827	25.0	125	0.1293	0.3846	0.2557	0.3072	0.9729
0.0422	30.0	150	0.1563	0.4545	0.2841	0.3497	0.9744
0.0205	35.0	175	0.1520	0.4138	0.2727	0.3288	0.9744
0.0102	40.0	200	0.1496	0.5124	0.3523	0.4175	0.9758
0.0062	45.0	225	0.1514	0.5522	0.4205	0.4774	0.9763
0.0047	50.0	250	0.1509	0.5581	0.4091	0.4721	0.9764
0.0035	55.0	275	0.1553	0.5294	0.4091	0.4615	0.9756

Framework versions

• Transformers 4.27.4
• Pytorch 1.13.1+cu116
• Datasets 2.10.1
• Tokenizers 0.13.2