Prgrg/ja-en-JESC-full-v1.0

This model is a fine-tuned version of Helsinki-NLP/opus-mt-jap-en on an unknown dataset. It achieves the following results on the evaluation set:

• Train Loss: 3.0489
• Validation Loss: 3.2533
• Epoch: 5

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:

• optimizer: {'name': 'AdamWeightDecay', 'learning_rate': {'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 0.0005, 'decay_steps': 419604, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}}, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False, 'weight_decay_rate': 0.001}
• training_precision: mixed_float16

Training results

Train Loss	Validation Loss	Epoch
3.4694	3.4733	0
3.5768	3.5449	1
3.3064	3.3898	2
3.1931	3.3273	3
3.1131	3.2747	4
3.0489	3.2533	5

Framework versions

• Transformers 4.26.1
• TensorFlow 2.11.0
• Datasets 2.10.1
• Tokenizers 0.13.2