da59cde9d17fdadf55fdd54c582a0d35

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

• Loss: 1.8312
• Data Size: 1.0
• Epoch Runtime: 5.5083
• Bleu: 9.4491

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	6.0916	0	1.0521	1.3080
No log	1	77	5.4782	0.0078	1.5819	1.4989
No log	2	154	4.9421	0.0156	1.2448	1.6449
No log	3	231	4.5210	0.0312	1.4423	1.5668
No log	4	308	4.1241	0.0625	1.6808	1.7150
No log	5	385	3.6333	0.125	1.8697	2.3368
0.3684	6	462	3.0974	0.25	2.5804	3.2482
1.2563	7	539	2.6046	0.5	3.4683	4.8448
2.3496	8.0	616	2.1792	1.0	5.8463	6.4885
2.026	9.0	693	1.9875	1.0	5.2534	7.4012
1.6246	10.0	770	1.8862	1.0	5.5555	8.0606
1.4995	11.0	847	1.8345	1.0	5.6784	8.7614
1.2373	12.0	924	1.7879	1.0	5.6302	8.8737
1.0809	13.0	1001	1.7736	1.0	5.3640	9.3247
0.9542	14.0	1078	1.7613	1.0	5.8307	9.5092
0.8418	15.0	1155	1.7636	1.0	5.7903	9.7305
0.7399	16.0	1232	1.7774	1.0	5.4496	9.6690
0.6453	17.0	1309	1.8026	1.0	5.4246	9.4353
0.5649	18.0	1386	1.8312	1.0	5.5083	9.4491

Framework versions

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