|
--- |
|
language: |
|
- es |
|
- ngu |
|
license: apache-2.0 |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- bleu |
|
model-index: |
|
- name: byt5-base-es_ngu |
|
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. --> |
|
|
|
# byt5-base-es_ngu |
|
|
|
This model is a fine-tuned version of [google/byt5-base](https://huggingface.co/google/byt5-base) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.6839 |
|
- Bleu: 8.2072 |
|
- Gen Len: 97.5599 |
|
|
|
## 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: 16 |
|
- eval_batch_size: 16 |
|
- seed: 65 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 10.0 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len | |
|
|:-------------:|:-----:|:----:|:---------------:|:------:|:-------:| |
|
| No log | 1.0 | 397 | 1.0434 | 0.0481 | 19.0 | |
|
| 1.5127 | 2.0 | 794 | 0.9015 | 0.0772 | 19.0 | |
|
| 1.0539 | 3.0 | 1191 | 0.8207 | 0.0991 | 19.0 | |
|
| 0.932 | 4.0 | 1588 | 0.7716 | 0.0925 | 19.0 | |
|
| 0.932 | 5.0 | 1985 | 0.7394 | 0.1292 | 19.0 | |
|
| 0.8578 | 6.0 | 2382 | 0.7168 | 0.1101 | 19.0 | |
|
| 0.8102 | 7.0 | 2779 | 0.6981 | 0.1465 | 19.0 | |
|
| 0.7837 | 8.0 | 3176 | 0.6898 | 0.1574 | 19.0 | |
|
| 0.7597 | 9.0 | 3573 | 0.6846 | 0.1562 | 19.0 | |
|
| 0.7597 | 10.0 | 3970 | 0.6839 | 0.1592 | 19.0 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.27.4 |
|
- Pytorch 2.0.0+cu118 |
|
- Datasets 2.11.0 |
|
- Tokenizers 0.13.3 |
|
|
