metadata
base_model: microsoft/deberta-v3-large
library_name: transformers
license: mit
metrics:
- accuracy
- precision
- recall
- f1
tags:
- generated_from_trainer
model-index:
- name: panclef_data_deberta_finetuned
results: []
panclef_data_deberta_finetuned
This model is a fine-tuned version of microsoft/deberta-v3-large on the PAN CLEF 2025 dataset. It achieves the following results on the evaluation set:
- Loss: 0.1185
- Accuracy: 0.9728
- Precision: 0.9735
- Recall: 0.9728
- F1: 0.9729
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: 3e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 |
|---|---|---|---|---|---|---|---|
| 0.3364 | 1.0 | 556 | 0.1755 | 0.9550 | 0.9572 | 0.9550 | 0.9552 |
| 0.2607 | 2.0 | 1112 | 0.2262 | 0.9347 | 0.9379 | 0.9347 | 0.9337 |
| 0.1895 | 3.0 | 1668 | 0.1562 | 0.9592 | 0.9619 | 0.9592 | 0.9595 |
| 0.1558 | 4.0 | 2224 | 0.1673 | 0.9627 | 0.9632 | 0.9627 | 0.9625 |
| 0.1617 | 5.0 | 2780 | 0.1505 | 0.9657 | 0.9659 | 0.9657 | 0.9658 |
| 0.1394 | 6.0 | 3336 | 0.1359 | 0.9632 | 0.9657 | 0.9632 | 0.9635 |
| 0.1354 | 7.0 | 3892 | 0.1310 | 0.9725 | 0.9725 | 0.9725 | 0.9725 |
| 0.1192 | 8.0 | 4448 | 0.1238 | 0.9683 | 0.9699 | 0.9683 | 0.9685 |
| 0.1096 | 9.0 | 5004 | 0.1222 | 0.9742 | 0.9744 | 0.9742 | 0.9742 |
| 0.1127 | 10.0 | 5560 | 0.1185 | 0.9728 | 0.9735 | 0.9728 | 0.9729 |
Framework versions
- Transformers 4.45.2
- Pytorch 2.4.1
- Datasets 3.0.0
- Tokenizers 0.20.1