amh-ner
This model is a fine-tuned version of mbeukman/xlm-roberta-base-finetuned-amharic-finetuned-ner-amharic on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.3026
- Precision: 0.8242
- Recall: 0.8595
- F1: 0.8415
- Accuracy: 0.9598
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
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 8
Training results
Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
---|---|---|---|---|---|---|---|
No log | 1.0 | 110 | 0.2826 | 0.7829 | 0.8156 | 0.7989 | 0.9541 |
No log | 2.0 | 220 | 0.2836 | 0.8144 | 0.8189 | 0.8166 | 0.9563 |
No log | 3.0 | 330 | 0.2879 | 0.8100 | 0.8375 | 0.8235 | 0.9576 |
No log | 4.0 | 440 | 0.3184 | 0.8091 | 0.8375 | 0.8231 | 0.9554 |
0.0667 | 5.0 | 550 | 0.3098 | 0.8208 | 0.8496 | 0.8350 | 0.9567 |
0.0667 | 6.0 | 660 | 0.3012 | 0.8319 | 0.8419 | 0.8369 | 0.9590 |
0.0667 | 7.0 | 770 | 0.3003 | 0.8246 | 0.8617 | 0.8427 | 0.9602 |
0.0667 | 8.0 | 880 | 0.3026 | 0.8242 | 0.8595 | 0.8415 | 0.9598 |
Framework versions
- Transformers 4.38.2
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2
- Downloads last month
- 8