<!-- 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. -->
BERT_multilingual_Greek_Offensive
This model is a fine-tuned version of bert-base-multilingual-uncased on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.3697
- Macro F1: 0.8189
- Micro F1: 0.8549
- Accuracy: 0.8549
Results on test set:
Accuracy: 0.8762953367875648
F1 score: 0.810674379114759
Precision: 0.7737546507849132
Recall : 0.8930094831854363
Matthews Correlation Coefficient: 0.6560127249515464
Precision of each class: [0.98262381 0.5648855 ]
Recall of each class: [0.86866359 0.91735537]
F1 score of each class: [0.92213616 0.6992126 ]
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 |
---|---|---|---|---|---|---|
0.5032 | 1.0 | 246 | 0.4083 | 0.7954 | 0.8446 | 0.8446 |
0.3616 | 2.0 | 492 | 0.3697 | 0.8189 | 0.8549 | 0.8549 |
0.3013 | 3.0 | 738 | 0.3698 | 0.8073 | 0.8526 | 0.8526 |
0.2496 | 4.0 | 984 | 0.4053 | 0.8151 | 0.8526 | 0.8526 |
Framework versions
- Transformers 4.27.1
- Pytorch 2.0.1+cu118
- Datasets 2.9.0
- Tokenizers 0.13.3