3b2334c90e65fdafe46bc0ed4ac9c720

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

• Loss: 2.8134
• Data Size: 1.0
• Epoch Runtime: 3.0816
• Bleu: 4.0930

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	7.0548	0	0.8425	0.1199
No log	1	33	6.4366	0.0078	1.3227	0.1463
No log	2	66	6.1314	0.0156	0.9727	0.1556
No log	3	99	5.8117	0.0312	1.2482	0.1539
0.3201	4	132	5.4563	0.0625	1.2946	0.2079
0.3201	5	165	5.0594	0.125	1.4617	0.5365
0.3201	6	198	4.5586	0.25	1.5666	1.1983
0.9251	7	231	4.0749	0.5	2.1381	1.5639
2.7707	8.0	264	3.5701	1.0	3.3482	2.3374
2.7707	9.0	297	3.2877	1.0	3.0498	2.7173
3.3967	10.0	330	3.1018	1.0	3.0993	3.2394
2.8804	11.0	363	2.9988	1.0	2.9140	3.5012
2.8804	12.0	396	2.9066	1.0	3.0558	3.6486
2.5354	13.0	429	2.8641	1.0	3.1840	3.7053
2.2723	14.0	462	2.7971	1.0	3.1932	3.9408
2.2723	15.0	495	2.7822	1.0	3.2175	4.0416
2.0626	16.0	528	2.7652	1.0	3.2975	4.1085
1.8581	17.0	561	2.7932	1.0	3.3129	4.1016
1.8581	18.0	594	2.7642	1.0	3.3412	4.1780
1.7025	19.0	627	2.7937	1.0	3.1036	4.2254
1.5083	20.0	660	2.7810	1.0	3.2457	4.1295
1.5083	21.0	693	2.7840	1.0	3.2604	4.2025
1.3861	22.0	726	2.8134	1.0	3.0816	4.0930

Framework versions

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