tst-gun-gub-pt
This model is a fine-tuned version of /content/opus-mt-en-mul on the tiagoblima/translation-pt-indigenouns dataset.
It achieves the following results on the evaluation set:
- Loss: 1.8158
- Bleu: 8.5368
- Gen Len: 59.24
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: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3.0
Training results
Training Loss |
Epoch |
Step |
Validation Loss |
Bleu |
Gen Len |
2.4421 |
0.29 |
4000 |
2.5285 |
3.3785 |
68.48 |
2.1667 |
0.59 |
8000 |
2.3018 |
4.5883 |
58.6 |
2.0255 |
0.88 |
12000 |
2.1290 |
5.1052 |
67.3 |
1.8995 |
1.18 |
16000 |
2.0535 |
7.8429 |
55.48 |
1.8322 |
1.47 |
20000 |
1.9960 |
7.2663 |
58.24 |
1.7868 |
1.77 |
24000 |
1.9224 |
7.0981 |
66.34 |
1.7012 |
2.06 |
28000 |
1.8869 |
7.5657 |
60.3 |
1.6773 |
2.36 |
32000 |
1.8613 |
7.9888 |
61.18 |
1.6631 |
2.65 |
36000 |
1.8354 |
8.0862 |
60.5 |
1.6379 |
2.94 |
40000 |
1.8158 |
8.4077 |
60.18 |
Framework versions
- Transformers 4.34.0
- Pytorch 2.0.1+cu118
- Datasets 2.14.5
- Tokenizers 0.14.1