metadata
library_name: transformers
license: apache-2.0
base_model: Sakonii/distilbert-base-nepali
tags:
- generated_from_trainer
metrics:
- accuracy
- recall
- precision
- f1
model-index:
- name: BERT_Classifier_DE
results: []
BERT_Classifier_DE
This model is a fine-tuned version of Sakonii/distilbert-base-nepali on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.8645
- Accuracy: 0.7815
- Recall: 0.6421
- Precision: 0.6362
- F1: 0.6294
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: 1.9165942005355648e-05
- train_batch_size: 8
- eval_batch_size: 64
- seed: 42
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.17707559519779958
- num_epochs: 20
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Recall | Precision | F1 |
|---|---|---|---|---|---|---|---|
| No log | 1.0 | 286 | 1.2997 | 0.7902 | 0.25 | 0.1976 | 0.2207 |
| 1.1831 | 2.0 | 572 | 1.0854 | 0.7273 | 0.3508 | 0.2747 | 0.2957 |
| 1.1831 | 3.0 | 858 | 1.0082 | 0.7640 | 0.4274 | 0.4180 | 0.3991 |
| 0.8686 | 4.0 | 1144 | 0.8645 | 0.7815 | 0.6421 | 0.6362 | 0.6294 |
| 0.8686 | 5.0 | 1430 | 0.9993 | 0.7483 | 0.5696 | 0.5609 | 0.5560 |
| 0.6366 | 6.0 | 1716 | 1.1232 | 0.7413 | 0.5868 | 0.4968 | 0.5287 |
| 0.4593 | 7.0 | 2002 | 1.5033 | 0.7483 | 0.4918 | 0.4888 | 0.4902 |
| 0.4593 | 8.0 | 2288 | 1.5870 | 0.7413 | 0.5122 | 0.4819 | 0.4950 |
| 0.3643 | 9.0 | 2574 | 2.0792 | 0.7255 | 0.4719 | 0.4854 | 0.4754 |
Framework versions
- Transformers 4.47.0
- Pytorch 2.5.1+cu121
- Datasets 3.2.0
- Tokenizers 0.21.0