6f03e3ccf540d38445168caa2136e4d9

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-it] dataset. It achieves the following results on the evaluation set:

• Loss: 1.8446
• Data Size: 1.0
• Epoch Runtime: 52.9441
• Bleu: 5.1567

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	7.2017	0	4.2182	0.0631
No log	1	808	4.0840	0.0078	4.8122	1.2194
No log	2	1616	3.4222	0.0156	5.5382	1.7416
No log	3	2424	2.9804	0.0312	6.5497	2.3377
0.1023	4	3232	2.6271	0.0625	7.9265	2.7971
2.6121	5	4040	2.3497	0.125	10.8154	3.3428
2.2643	6	4848	2.1416	0.25	16.7074	3.8203
1.9605	7	5656	1.9658	0.5	28.3719	4.2605
1.8202	8.0	6464	1.8226	1.0	52.4091	4.7310
1.5705	9.0	7272	1.7613	1.0	53.1537	5.0511
1.4066	10.0	8080	1.7451	1.0	54.3745	5.1034
1.2842	11.0	8888	1.7499	1.0	53.6888	5.1784
1.1317	12.0	9696	1.7716	1.0	53.7651	5.1852
1.0433	13.0	10504	1.8133	1.0	53.3684	5.2086
0.9277	14.0	11312	1.8446	1.0	52.9441	5.1567

Framework versions

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