mental-bert-base-uncased-finetuned-depression

This model is a fine-tuned version of mental/mental-bert-base-uncased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.5358
• Precision: 0.8986
• Recall: 0.8885
• F1: 0.8933
• Accuracy: 0.9158

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: 2e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	469	0.3929	0.8744	0.8346	0.8516	0.8849
0.4726	2.0	938	0.4405	0.9052	0.8359	0.8660	0.8955
0.2165	3.0	1407	0.4594	0.8627	0.8435	0.8515	0.8891
0.1263	4.0	1876	0.5213	0.9012	0.8781	0.8886	0.9094
0.0719	5.0	2345	0.4879	0.9036	0.8694	0.8851	0.9083
0.0471	6.0	2814	0.5628	0.9185	0.8639	0.8880	0.9104
0.0431	7.0	3283	0.5592	0.8980	0.8731	0.8846	0.9104
0.0402	8.0	3752	0.5948	0.9166	0.8591	0.8848	0.9094
0.0348	9.0	4221	0.5358	0.8986	0.8885	0.8933	0.9158
0.0276	10.0	4690	0.6361	0.9116	0.8619	0.8843	0.9094
0.0281	11.0	5159	0.6535	0.9095	0.8726	0.8897	0.9147
0.029	12.0	5628	0.6776	0.9098	0.8673	0.8868	0.9136
0.0188	13.0	6097	0.6940	0.9072	0.8629	0.8829	0.9072
0.0215	14.0	6566	0.7022	0.9168	0.8606	0.8856	0.9115
0.0184	15.0	7035	0.6996	0.9027	0.8687	0.8846	0.9126
0.0204	16.0	7504	0.6990	0.9063	0.8687	0.8861	0.9126
0.0204	17.0	7973	0.7268	0.9103	0.8677	0.8871	0.9115
0.0185	18.0	8442	0.7210	0.9066	0.8766	0.8907	0.9147
0.0181	19.0	8911	0.7346	0.9096	0.8732	0.8902	0.9147
0.0151	20.0	9380	0.7363	0.9090	0.8720	0.8892	0.9136

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1