metadata
license: mit
base_model: sagorsarker/bangla-bert-base
tags:
- generated_from_trainer
metrics:
- f1
- accuracy
model-index:
- name: bangla-bert-base-MLTC-BB2
results: []
bangla-bert-base-MLTC-BB2
This model is a fine-tuned version of sagorsarker/bangla-bert-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.4005
- F1: 0.8514
- F1 Weighted: 0.8504
- Roc Auc: 0.8470
- Accuracy: 0.5424
- Hamming Loss: 0.1530
- Jaccard Score: 0.7413
- Zero One Loss: 0.4576
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: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | F1 | F1 Weighted | Roc Auc | Accuracy | Hamming Loss | Jaccard Score | Zero One Loss |
|---|---|---|---|---|---|---|---|---|---|---|
| 0.4015 | 1.0 | 73 | 0.4027 | 0.8246 | 0.8201 | 0.8206 | 0.5039 | 0.1793 | 0.7016 | 0.4961 |
| 0.3579 | 2.0 | 146 | 0.3583 | 0.8541 | 0.8548 | 0.8522 | 0.5553 | 0.1478 | 0.7453 | 0.4447 |
| 0.2831 | 3.0 | 219 | 0.3768 | 0.8429 | 0.8402 | 0.8360 | 0.5398 | 0.1639 | 0.7284 | 0.4602 |
| 0.1799 | 4.0 | 292 | 0.3565 | 0.8534 | 0.8534 | 0.8489 | 0.5604 | 0.1510 | 0.7443 | 0.4396 |
| 0.1671 | 5.0 | 365 | 0.3724 | 0.8552 | 0.8547 | 0.8508 | 0.5681 | 0.1491 | 0.7470 | 0.4319 |
| 0.1451 | 6.0 | 438 | 0.3773 | 0.8493 | 0.8484 | 0.8477 | 0.5450 | 0.1523 | 0.7381 | 0.4550 |
| 0.0899 | 7.0 | 511 | 0.3986 | 0.8502 | 0.8494 | 0.8457 | 0.5398 | 0.1542 | 0.7394 | 0.4602 |
| 0.0744 | 8.0 | 584 | 0.4005 | 0.8514 | 0.8504 | 0.8470 | 0.5424 | 0.1530 | 0.7413 | 0.4576 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.1.2
- Datasets 2.19.2
- Tokenizers 0.19.1