<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
moderate_severe_depression_longformerlarge_model
This model is a fine-tuned version of allenai/longformer-scico on the 'christinacdl/balanced_depression_dataset' dataset. It achieves the following results on the evaluation set:
- Loss: 0.4331
- Macro F1: 0.8927
- Micro F1: 0.8929
- Accuracy: 0.8929
Model performance on test set:
Accuracy: 0.882091212458287
F1 score: 0.8819016546683596
Precision: 0.8844185528069173
Recall : 0.882091212458287
Matthews Correlation Coefficient: 0.8244878843339586
Precision of each class: [0.97601371 0.85618932 0.82105263]
Recall of each class: [0.95050056 0.78476085 0.91101224]
F1 score of each class: [0.96308819 0.81892049 0.86369628]
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: 6
- eval_batch_size: 6
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 12
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 6
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Macro F1 | Micro F1 | Accuracy |
|---|---|---|---|---|---|---|
| 0.3517 | 1.0 | 1798 | 0.3561 | 0.8741 | 0.8745 | 0.8745 |
| 0.2664 | 2.0 | 3596 | 0.3241 | 0.8737 | 0.8732 | 0.8732 |
| 0.2337 | 3.0 | 5394 | 0.3824 | 0.8887 | 0.8892 | 0.8892 |
| 0.2019 | 4.0 | 7192 | 0.4330 | 0.8927 | 0.8929 | 0.8929 |
| 0.1489 | 5.0 | 8990 | 0.5230 | 0.8917 | 0.8919 | 0.8919 |
| 0.1187 | 6.0 | 10788 | 0.5839 | 0.8882 | 0.8882 | 0.8882 |
Framework versions
- Transformers 4.27.1
- Pytorch 2.0.1+cu118
- Datasets 2.9.0
- Tokenizers 0.13.3
