metadata
license: apache-2.0
base_model: Helsinki-NLP/opus-mt-ar-en
tags:
- generated_from_trainer
metrics:
- bleu
model-index:
- name: darija-latin-to-english-new
results: []
darija-latin-to-english-new
This model is a fine-tuned version of Helsinki-NLP/opus-mt-ar-en on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0635
- Bleu: 18.3432
- Gen Len: 9.0506
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: 2e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 2
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len |
|---|---|---|---|---|---|
| 0.046 | 0.24 | 500 | 0.0939 | 4.0335 | 11.1293 |
| 0.0347 | 0.48 | 1000 | 0.0788 | 9.8265 | 10.1861 |
| 0.0279 | 0.72 | 1500 | 0.0725 | 13.178 | 9.6648 |
| 0.0251 | 0.96 | 2000 | 0.0684 | 15.8154 | 9.536 |
| 0.0211 | 1.2 | 2500 | 0.0666 | 17.0063 | 9.3259 |
| 0.0197 | 1.44 | 3000 | 0.0650 | 17.4465 | 9.2086 |
| 0.0199 | 1.68 | 3500 | 0.0639 | 18.3809 | 9.0083 |
| 0.0191 | 1.92 | 4000 | 0.0635 | 18.3432 | 9.0506 |
Framework versions
- Transformers 4.38.1
- Pytorch 2.1.2
- Datasets 2.1.0
- Tokenizers 0.15.2