7234fbc9fb2680183d8c48a81f82d4ac

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

• Loss: 2.6016
• Data Size: 1.0
• Epoch Runtime: 5.4966
• Bleu: 2.5651

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	6.8778	0	1.0118	0.0915
No log	1	74	6.1856	0.0078	1.3727	0.1241
No log	2	148	5.7128	0.0156	1.2249	0.1024
0.2052	3	222	5.3496	0.0312	1.3044	0.0671
0.2052	4	296	4.9954	0.0625	1.3931	0.1853
0.3506	5	370	4.5174	0.125	1.7363	0.2739
0.3506	6	444	4.0586	0.25	2.2478	0.3311
0.9088	7	518	3.6293	0.5	3.1960	0.6039
2.3431	8.0	592	3.1987	1.0	5.7161	0.9474
3.0586	9.0	666	2.9847	1.0	5.0990	1.2268
2.8323	10.0	740	2.8373	1.0	5.2420	1.5073
2.5609	11.0	814	2.7508	1.0	5.1466	1.6652
2.4155	12.0	888	2.6689	1.0	5.3888	1.8865
2.1954	13.0	962	2.6220	1.0	5.5423	2.0717
2.0836	14.0	1036	2.5937	1.0	5.2130	2.1357
1.9069	15.0	1110	2.5827	1.0	5.5706	2.2860
1.7963	16.0	1184	2.5671	1.0	5.3078	2.3413
1.6678	17.0	1258	2.5653	1.0	5.3269	2.3327
1.5764	18.0	1332	2.5713	1.0	5.2825	2.4068
1.453	19.0	1406	2.5680	1.0	5.2886	2.4990
1.3687	20.0	1480	2.5974	1.0	5.8346	2.5674
1.2576	21.0	1554	2.6016	1.0	5.4966	2.5651

Framework versions

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