|
--- |
|
base_model: vinai/bartpho-syllable |
|
tags: |
|
- text2text-generation |
|
- generated_from_trainer |
|
metrics: |
|
- sacrebleu |
|
model-index: |
|
- name: nlp_vietnamese_spelling |
|
results: [] |
|
--- |
|
|
|
<!-- 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. --> |
|
|
|
# nlp_vietnamese_spelling |
|
|
|
This model is a fine-tuned version of [vinai/bartpho-syllable](https://huggingface.co/vinai/bartpho-syllable) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.3776 |
|
- Sacrebleu: 9.6073 |
|
|
|
## 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: 4 |
|
- eval_batch_size: 4 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 5 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Sacrebleu | |
|
|:-------------:|:------:|:----:|:---------------:|:---------:| |
|
| 1.6702 | 0.4803 | 500 | 0.8637 | 5.3242 | |
|
| 1.0413 | 0.9606 | 1000 | 0.6961 | 6.4143 | |
|
| 0.7764 | 1.4409 | 1500 | 0.5493 | 7.8371 | |
|
| 0.6854 | 1.9212 | 2000 | 0.4883 | 8.3411 | |
|
| 0.5523 | 2.4015 | 2500 | 0.4498 | 8.7689 | |
|
| 0.5078 | 2.8818 | 3000 | 0.4146 | 9.2720 | |
|
| 0.4226 | 3.3622 | 3500 | 0.4010 | 9.3349 | |
|
| 0.4222 | 3.8425 | 4000 | 0.3924 | 9.4525 | |
|
| 0.3832 | 4.3228 | 4500 | 0.3839 | 9.5898 | |
|
| 0.3634 | 4.8031 | 5000 | 0.3776 | 9.6073 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.40.2 |
|
- Pytorch 2.2.1+cu121 |
|
- Datasets 2.19.1 |
|
- Tokenizers 0.19.1 |
|
|
