420ea623dc59db96fe6ff3207a2ca9b8

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

• Loss: 2.6079
• Data Size: 1.0
• Epoch Runtime: 76.2638
• Bleu: 7.5478

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	8.7748	0	6.6082	0.1187
No log	1	1286	6.2774	0.0078	8.7866	0.2045
0.1261	2	2572	5.7243	0.0156	7.6610	0.4504
0.154	3	3858	5.2991	0.0312	10.2081	0.7864
0.2257	4	5144	4.8742	0.0625	11.1689	1.3082
4.583	5	6430	4.4024	0.125	15.3746	1.9953
4.0532	6	7716	3.9011	0.25	24.5200	3.0252
3.5102	7	9002	3.4312	0.5	40.5008	4.1634
3.1238	8.0	10288	3.0159	1.0	77.8465	5.4951
2.8234	9.0	11574	2.8110	1.0	74.8421	6.1374
2.5523	10.0	12860	2.7011	1.0	79.3606	6.5963
2.3361	11.0	14146	2.6453	1.0	77.3506	6.9008
2.2469	12.0	15432	2.5883	1.0	78.2020	7.0949
2.0549	13.0	16718	2.5708	1.0	79.4089	7.1978
1.9657	14.0	18004	2.5598	1.0	76.1451	7.3160
1.8865	15.0	19290	2.5527	1.0	77.4235	7.4165
1.7499	16.0	20576	2.5571	1.0	76.9110	7.4967
1.6983	17.0	21862	2.5678	1.0	77.8294	7.5138
1.5949	18.0	23148	2.5845	1.0	76.7946	7.5355
1.5304	19.0	24434	2.6079	1.0	76.2638	7.5478

Framework versions

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