5f1361050d64792ce520deecccdcac6b

This model is a fine-tuned version of Helsinki-NLP/opus-mt-tc-bible-big-deu_eng_fra_por_spa-mul on the Helsinki-NLP/opus_books [en-pt] dataset. It achieves the following results on the evaluation set:

• Loss: 1.9442
• Data Size: 1.0
• Epoch Runtime: 4.5685
• Bleu: 12.4822

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: 8
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• total_train_batch_size: 32
• total_eval_batch_size: 32
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: constant
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Data Size	Epoch Runtime	Bleu
No log	0	0	7.0935	0	0.7808	0.0973
No log	1	35	5.1097	0.0078	1.1464	1.8174
No log	2	70	4.5185	0.0156	1.0575	3.5008
No log	3	105	4.1224	0.0312	1.4811	4.5561
No log	4	140	3.8304	0.0625	1.5562	5.3584
No log	5	175	3.4689	0.125	1.6133	6.0616
No log	6	210	3.0196	0.25	1.9661	6.7411
No log	7	245	2.4245	0.5	2.4968	8.6789
0.5757	8.0	280	2.0572	1.0	3.9901	10.3071
1.7736	9.0	315	1.9358	1.0	3.8998	11.1828
1.1775	10.0	350	1.8893	1.0	3.3234	11.5380
1.1775	11.0	385	1.8634	1.0	4.6761	12.0152
0.7677	12.0	420	1.8759	1.0	3.9816	12.4841
0.5658	13.0	455	1.9078	1.0	4.3978	12.1625
0.5658	14.0	490	1.9384	1.0	4.4825	12.4980
0.4015	15.0	525	1.9442	1.0	4.5685	12.4822

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.2.0
• Tokenizers 0.22.1