<!-- 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. -->
mdeberta-v3-base_Greek_Offensive
This model is a fine-tuned version of studio-ousia/mluke-base on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.4114
- Macro F1: 0.8127
- Micro F1: 0.848
- Accuracy: 0.848
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: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Macro F1 | Micro F1 | Accuracy |
|---|---|---|---|---|---|---|
| 0.489 | 1.0 | 246 | 0.4237 | 0.7788 | 0.8194 | 0.8194 |
| 0.3623 | 2.0 | 492 | 0.3836 | 0.8090 | 0.8491 | 0.8491 |
| 0.3108 | 3.0 | 738 | 0.3845 | 0.8082 | 0.8469 | 0.8469 |
| 0.2615 | 4.0 | 984 | 0.4114 | 0.8127 | 0.848 | 0.848 |
Accuracy: 0.8639896373056994
F1 score: 0.7958536069388619
Precision: 0.7602529358626919
Recall : 0.8840309250866436
Matthews Correlation Coefficient: 0.6322821386560418
Precision of each class: [0.98148148 0.53902439]
Recall of each class: [0.85483871 0.91322314]
F1 score of each class: [0.9137931 0.67791411]
Framework versions
- Transformers 4.27.1
- Pytorch 2.0.1+cu117
- Datasets 2.9.0
- Tokenizers 0.13.3
