metadata
license: mit
base_model: gogamza/kobart-base-v2
tags:
- generated_from_trainer
metrics:
- rouge
model-index:
- name: kobart-base-v2-159-korean
results: []
kobart-base-v2-159-korean
This model is a fine-tuned version of gogamza/kobart-base-v2 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.2628
- Rouge1: 0.3634
- Rouge2: 0.1451
- Rougel: 0.3566
- Rougelsum: 0.3563
- Gen Len: 20.0
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: 5e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|---|---|---|---|---|---|---|---|---|
| 0.2804 | 1.44 | 500 | 0.2805 | 0.3321 | 0.1273 | 0.3272 | 0.3276 | 19.9992 |
| 0.2141 | 2.88 | 1000 | 0.2472 | 0.3577 | 0.1381 | 0.3526 | 0.3525 | 20.0 |
| 0.1407 | 4.33 | 1500 | 0.2495 | 0.3615 | 0.1457 | 0.3543 | 0.3543 | 20.0 |
| 0.1206 | 5.77 | 2000 | 0.2508 | 0.3592 | 0.1448 | 0.3533 | 0.3532 | 20.0 |
| 0.0853 | 7.21 | 2500 | 0.2603 | 0.3623 | 0.147 | 0.3561 | 0.3562 | 20.0 |
| 0.0777 | 8.65 | 3000 | 0.2628 | 0.3634 | 0.1451 | 0.3566 | 0.3563 | 20.0 |
Framework versions
- Transformers 4.36.2
- Pytorch 2.1.2+cu121
- Datasets 2.15.0
- Tokenizers 0.15.0