Indonesian T5 Language Models
Collection
Indonesian T5 models pre-trained with nanoT5 and fine-tuned on IndoNLG tasks. GitHub: https://github.com/LazarusNLP/IndoT5/
•
5 items
•
Updated
LazarusNLP/IndoNanoT5-base-XPersona
This model is a fine-tuned version of LazarusNLP/IndoNanoT5-base on the indonlg dataset. It achieves the following results on the evaluation set:
- Loss: 1.8372
- Bleu: 4.0669
- Sacrebleu: 4.0669
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: 1e-05
- train_batch_size: 8
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 50
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Sacrebleu |
|---|---|---|---|---|---|
| 1.9872 | 1.0 | 15516 | 1.8482 | 3.7015 | 3.7015 |
| 1.888 | 2.0 | 31032 | 1.8434 | 4.0409 | 4.0409 |
| 1.8207 | 3.0 | 46548 | 1.8347 | 4.1239 | 4.1239 |
| 1.7716 | 4.0 | 62064 | 1.8340 | 4.3231 | 4.3231 |
| 1.6948 | 5.0 | 77580 | 1.8443 | 4.4283 | 4.4283 |
| 1.6442 | 6.0 | 93096 | 1.8563 | 4.5338 | 4.5338 |
| 1.5856 | 7.0 | 108612 | 1.8782 | 4.3033 | 4.3033 |
| 1.5451 | 8.0 | 124128 | 1.8930 | 4.3286 | 4.3286 |
| 1.5056 | 9.0 | 139644 | 1.9207 | 4.2773 | 4.2773 |
| 1.446 | 10.0 | 155160 | 1.9406 | 4.0629 | 4.0629 |
| 1.406 | 11.0 | 170676 | 1.9636 | 4.1382 | 4.1382 |
Framework versions
- Transformers 4.37.2
- Pytorch 2.2.0+cu118
- Datasets 2.16.1
- Tokenizers 0.15.1
- Downloads last month
- 20
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.
Model tree for LazarusNLP/IndoNanoT5-base-XPersona
Base model
LazarusNLP/IndoNanoT5-baseDataset used to train LazarusNLP/IndoNanoT5-base-XPersona
Collection including LazarusNLP/IndoNanoT5-base-XPersona
Evaluation results
- Bleu on indonlgtest set self-reported4.067
- Sacrebleu on indonlgtest set self-reported4.067