bengali_pos_v1_300000
This model is a fine-tuned version of mHossain/bengali_pos_v1_400000 on the pos_tag_100k dataset. It achieves the following results on the evaluation set:
- Loss: 0.5865
- Precision: 0.7758
- Recall: 0.7797
- F1: 0.7778
- Accuracy: 0.8339
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: 3
Training results
Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
---|---|---|---|---|---|---|---|
0.587 | 1.0 | 22500 | 0.5821 | 0.7614 | 0.7643 | 0.7628 | 0.8232 |
0.4808 | 2.0 | 45000 | 0.5726 | 0.7733 | 0.7762 | 0.7747 | 0.8316 |
0.3909 | 3.0 | 67500 | 0.5865 | 0.7758 | 0.7797 | 0.7778 | 0.8339 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.0
- Downloads last month
- 8
Finetuned from
Evaluation results
- Precision on pos_tag_100kvalidation set self-reported0.776
- Recall on pos_tag_100kvalidation set self-reported0.780
- F1 on pos_tag_100kvalidation set self-reported0.778
- Accuracy on pos_tag_100kvalidation set self-reported0.834