Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,63 @@
|
|
| 1 |
---
|
| 2 |
license: apache-2.0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
license: apache-2.0
|
| 3 |
+
language:
|
| 4 |
+
- es
|
| 5 |
+
- en
|
| 6 |
+
base_model: vgaraujov/t5-base-spanish
|
| 7 |
+
tags:
|
| 8 |
+
- generated_from_trainer
|
| 9 |
+
datasets:
|
| 10 |
+
- vgaraujov/wmt13
|
| 11 |
+
metrics:
|
| 12 |
+
- bleu
|
| 13 |
+
model-index:
|
| 14 |
+
- name: t5-base-translation-es-en
|
| 15 |
+
results:
|
| 16 |
+
- task:
|
| 17 |
+
name: Translation
|
| 18 |
+
type: translation
|
| 19 |
+
dataset:
|
| 20 |
+
name: vgaraujov/wmt13 es-en
|
| 21 |
+
type: vgaraujov/wmt13
|
| 22 |
+
config: es-en
|
| 23 |
+
split: validation
|
| 24 |
+
args: es-en
|
| 25 |
+
metrics:
|
| 26 |
+
- name: Bleu
|
| 27 |
+
type: bleu
|
| 28 |
+
value: 29.7546
|
| 29 |
+
widget:
|
| 30 |
+
- text: Hola, soy T5S para traducción.
|
| 31 |
---
|
| 32 |
+
|
| 33 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 34 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 35 |
+
|
| 36 |
+
# T5S (base-sized model) for es-en translation
|
| 37 |
+
|
| 38 |
+
This model is a fine-tuned version of [T5S](https://huggingface.co/vgaraujov/t5-base-spanish) on a small portion of [WMT13](https://huggingface.co/datasets/vgaraujov/wmt13) es-en dataset.
|
| 39 |
+
It achieves the following results on the evaluation set:
|
| 40 |
+
- Loss: 1.2055
|
| 41 |
+
- Bleu: 29.7546
|
| 42 |
+
- Gen Len: 48.1618
|
| 43 |
+
|
| 44 |
+
### Training hyperparameters
|
| 45 |
+
|
| 46 |
+
The following hyperparameters were used during training:
|
| 47 |
+
- learning_rate: 0.005
|
| 48 |
+
- train_batch_size: 32
|
| 49 |
+
- eval_batch_size: 32
|
| 50 |
+
- seed: 42
|
| 51 |
+
- gradient_accumulation_steps: 12
|
| 52 |
+
- total_train_batch_size: 384
|
| 53 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 54 |
+
- lr_scheduler_type: linear
|
| 55 |
+
- lr_scheduler_warmup_steps: 40000
|
| 56 |
+
- training_steps: 10000
|
| 57 |
+
|
| 58 |
+
### Framework versions
|
| 59 |
+
|
| 60 |
+
- Transformers 4.33.0.dev0
|
| 61 |
+
- Pytorch 2.0.1+cu117
|
| 62 |
+
- Datasets 2.14.4
|
| 63 |
+
- Tokenizers 0.13.3
|