tmVar_5e-05_30_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: 0.0230
Precision: 0.8677
Recall: 0.8865
F1: 0.8770
Accuracy: 0.9964

Model description

Trained on Token set with max_length=475

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: 5e-05
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
0.3602	1.39	25	0.0547	0.4823	0.3676	0.4172	0.9851
0.0498	2.78	50	0.0305	0.4518	0.5568	0.4988	0.9912
0.0237	4.17	75	0.0198	0.6338	0.7297	0.6784	0.9942
0.0089	5.56	100	0.0164	0.7895	0.8919	0.8376	0.9960
0.0036	6.94	125	0.0138	0.7826	0.8757	0.8265	0.9967
0.0023	8.33	150	0.0148	0.8462	0.8919	0.8684	0.9969
0.0012	9.72	175	0.0159	0.7890	0.9297	0.8536	0.9966
0.0012	11.11	200	0.0163	0.845	0.9135	0.8779	0.9970
0.001	12.5	225	0.0165	0.8534	0.8811	0.8670	0.9967
0.0012	13.89	250	0.0215	0.8020	0.8757	0.8372	0.9961
0.0008	15.28	275	0.0192	0.875	0.9081	0.8912	0.9970
0.0007	16.67	300	0.0192	0.875	0.9081	0.8912	0.9970
0.0005	18.06	325	0.0192	0.875	0.9081	0.8912	0.9970
0.0009	19.44	350	0.0230	0.8677	0.8865	0.8770	0.9964

Framework versions

Transformers 4.27.4
Pytorch 1.13.1+cu116
Datasets 2.11.0
Tokenizers 0.13.2