46b679fe9bc03337d4ee03fb67085496

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

• Loss: 3.0865
• Data Size: 1.0
• Epoch Runtime: 5.2071
• Bleu: 0.8105

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	7.9239	0	1.0149	0.0465
No log	1	70	7.2510	0.0078	1.3865	0.0466
No log	2	140	6.6455	0.0156	1.1463	0.0335
No log	3	210	6.1437	0.0312	1.2868	0.0479
No log	4	280	5.7170	0.0625	1.4876	0.1236
No log	5	350	5.1893	0.125	1.8570	0.0544
No log	6	420	4.5745	0.25	2.5524	0.1313
0.7629	7	490	4.0746	0.5	3.8461	0.2245
3.8903	8.0	560	3.6585	1.0	5.9354	0.4664
3.5915	9.0	630	3.4200	1.0	5.7251	0.4608
3.2638	10.0	700	3.2757	1.0	5.2348	0.5928
3.0817	11.0	770	3.1627	1.0	5.2383	0.6393
2.9574	12.0	840	3.1059	1.0	5.5024	0.6496
2.7365	13.0	910	3.0377	1.0	5.6766	0.7629
2.5858	14.0	980	3.0371	1.0	5.5097	0.7874
2.4515	15.0	1050	3.0070	1.0	5.5140	0.8441
2.2979	16.0	1120	3.0067	1.0	5.3235	0.8295
2.1811	17.0	1190	3.0127	1.0	5.6762	0.7732
2.0082	18.0	1260	3.0292	1.0	5.3448	0.8707
1.9011	19.0	1330	3.0616	1.0	5.3447	0.7117
1.7533	20.0	1400	3.0865	1.0	5.2071	0.8105

Framework versions

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