<!-- 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. -->
multibert_1310seed7
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: 0.4338
- Precisions: 0.8841
- Recall: 0.8144
- F-measure: 0.8437
- Accuracy: 0.9402
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: 7.5e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 7
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 14
Training results
| Training Loss | Epoch | Step | Validation Loss | Precisions | Recall | F-measure | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.4441 | 1.0 | 236 | 0.2809 | 0.8700 | 0.7020 | 0.7222 | 0.9118 |
| 0.2161 | 2.0 | 472 | 0.2575 | 0.8741 | 0.7653 | 0.7818 | 0.9250 |
| 0.1277 | 3.0 | 708 | 0.2644 | 0.8331 | 0.8115 | 0.8175 | 0.9299 |
| 0.0891 | 4.0 | 944 | 0.2614 | 0.8671 | 0.8120 | 0.8341 | 0.9390 |
| 0.0559 | 5.0 | 1180 | 0.3259 | 0.8806 | 0.7923 | 0.8279 | 0.9332 |
| 0.0322 | 6.0 | 1416 | 0.3770 | 0.8807 | 0.8064 | 0.8333 | 0.9373 |
| 0.0241 | 7.0 | 1652 | 0.4548 | 0.8430 | 0.8213 | 0.8223 | 0.9323 |
| 0.0162 | 8.0 | 1888 | 0.3705 | 0.8493 | 0.8239 | 0.8343 | 0.9405 |
| 0.0099 | 9.0 | 2124 | 0.4498 | 0.8463 | 0.8094 | 0.8245 | 0.9369 |
| 0.0069 | 10.0 | 2360 | 0.4445 | 0.8606 | 0.8141 | 0.8328 | 0.9381 |
| 0.0062 | 11.0 | 2596 | 0.4429 | 0.8880 | 0.8075 | 0.8405 | 0.9383 |
| 0.0045 | 12.0 | 2832 | 0.4496 | 0.8794 | 0.8017 | 0.8322 | 0.9393 |
| 0.0041 | 13.0 | 3068 | 0.4338 | 0.8841 | 0.8144 | 0.8437 | 0.9402 |
| 0.0029 | 14.0 | 3304 | 0.4401 | 0.8850 | 0.8135 | 0.8437 | 0.9400 |
Framework versions
- Transformers 4.34.0
- Pytorch 2.0.1+cu118
- Datasets 2.14.5
- Tokenizers 0.14.1
