<!-- 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. -->
indobart-indonlg-nusax-jv-id
This model is a fine-tuned version of indobenchmark/indobart-v2 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.5583
- Bleu: 20.5014
- Gen Len: 19.2932
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: 64
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 20
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len |
|---|---|---|---|---|---|
| No log | 1.0 | 108 | 0.5734 | 17.987 | 19.3211 |
| No log | 2.0 | 216 | 0.5692 | 18.4406 | 19.3032 |
| No log | 3.0 | 324 | 0.5651 | 18.7223 | 19.33 |
| No log | 4.0 | 432 | 0.5627 | 19.1538 | 19.3367 |
| 0.3863 | 5.0 | 540 | 0.5617 | 19.4173 | 19.3055 |
| 0.3863 | 6.0 | 648 | 0.5571 | 19.438 | 19.2977 |
| 0.3863 | 7.0 | 756 | 0.5582 | 19.6099 | 19.3021 |
| 0.3863 | 8.0 | 864 | 0.5554 | 19.7952 | 19.3211 |
| 0.3863 | 9.0 | 972 | 0.5563 | 19.9403 | 19.3043 |
| 0.2923 | 10.0 | 1080 | 0.5536 | 20.2294 | 19.2943 |
| 0.2923 | 11.0 | 1188 | 0.5527 | 20.2212 | 19.3043 |
| 0.2923 | 12.0 | 1296 | 0.5526 | 20.24 | 19.2988 |
| 0.2923 | 13.0 | 1404 | 0.5543 | 20.0935 | 19.2776 |
| 0.245 | 14.0 | 1512 | 0.5551 | 20.2451 | 19.2765 |
| 0.245 | 15.0 | 1620 | 0.5576 | 20.3303 | 19.3088 |
| 0.245 | 16.0 | 1728 | 0.5579 | 20.4869 | 19.2977 |
| 0.245 | 17.0 | 1836 | 0.5581 | 20.4655 | 19.2921 |
| 0.245 | 18.0 | 1944 | 0.5577 | 20.4589 | 19.2899 |
| 0.2109 | 19.0 | 2052 | 0.5584 | 20.4604 | 19.2921 |
| 0.2109 | 20.0 | 2160 | 0.5583 | 20.5014 | 19.2932 |
Framework versions
- Transformers 4.33.1
- Pytorch 2.0.1+cu118
- Datasets 2.14.5
- Tokenizers 0.13.3
