tmvar_5e-05

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: 0.0165
• Precision: 0.8814
• Recall: 0.9243
• F1: 0.9024
• Accuracy: 0.9977

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: 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
0.2905	1.47	25	0.0978	0.0	0.0	0.0	0.9843
0.0551	2.94	50	0.0382	0.3893	0.6270	0.4803	0.9887
0.0239	4.41	75	0.0192	0.5915	0.7514	0.6619	0.9947
0.0111	5.88	100	0.0153	0.8564	0.8703	0.8633	0.9964
0.0031	7.35	125	0.0126	0.8731	0.9297	0.9005	0.9975
0.002	8.82	150	0.0129	0.865	0.9351	0.8987	0.9978
0.0013	10.29	175	0.0163	0.8830	0.8973	0.8901	0.9968
0.0011	11.76	200	0.0171	0.9	0.9243	0.912	0.9970
0.001	13.24	225	0.0165	0.8808	0.9189	0.8995	0.9973
0.0008	14.71	250	0.0138	0.8923	0.9405	0.9158	0.9981
0.0007	16.18	275	0.0165	0.8763	0.9189	0.8971	0.9975
0.0005	17.65	300	0.0170	0.8854	0.9189	0.9019	0.9974
0.0005	19.12	325	0.0148	0.8731	0.9297	0.9005	0.9979
0.0005	20.59	350	0.0171	0.8848	0.9135	0.8989	0.9973
0.0005	22.06	375	0.0176	0.8848	0.9135	0.8989	0.9973
0.0005	23.53	400	0.0167	0.8860	0.9243	0.9048	0.9975
0.0004	25.0	425	0.0166	0.8860	0.9243	0.9048	0.9976
0.0004	26.47	450	0.0165	0.8814	0.9243	0.9024	0.9977
0.0004	27.94	475	0.0165	0.8814	0.9243	0.9024	0.9977

Framework versions

• Transformers 4.27.4
• Pytorch 2.0.0+cu118
• Datasets 2.11.0
• Tokenizers 0.13.2