xlm-roberta-base-finetuned-ner
This model is a fine-tuned version of xlm-roberta-base on the hi_ner-original dataset. It achieves the following results on the evaluation set:
- Loss: 0.2314
- Precision: 0.7366
- Recall: 0.6771
- F1: 0.7056
- Accuracy: 0.9359
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: 4
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 6
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.2025 | 0.74 | 7000 | 0.2146 | 0.7399 | 0.6197 | 0.6745 | 0.9316 |
| 0.1641 | 1.47 | 14000 | 0.2238 | 0.7618 | 0.6108 | 0.6780 | 0.9336 |
| 0.1404 | 2.21 | 21000 | 0.2302 | 0.7560 | 0.6327 | 0.6889 | 0.9350 |
| 0.1371 | 2.95 | 28000 | 0.2226 | 0.7395 | 0.6600 | 0.6975 | 0.9350 |
| 0.1248 | 3.68 | 35000 | 0.2314 | 0.7366 | 0.6771 | 0.7056 | 0.9359 |
| 0.1112 | 4.42 | 42000 | 0.2423 | 0.7089 | 0.7064 | 0.7077 | 0.9333 |
| 0.1048 | 5.16 | 49000 | 0.2599 | 0.7326 | 0.6793 | 0.7050 | 0.9349 |
| 0.1091 | 5.89 | 56000 | 0.2542 | 0.7244 | 0.6918 | 0.7077 | 0.9348 |
Framework versions
- Transformers 4.19.4
- Pytorch 1.11.0+cu102
- Datasets 2.3.2
- Tokenizers 0.12.1