d2cf2a4cd7af58299b9a780ca707f40a

This model is a fine-tuned version of Helsinki-NLP/opus-mt-tc-bible-big-deu_eng_fra_por_spa-mul on the Helsinki-NLP/opus_books [fi-fr] dataset. It achieves the following results on the evaluation set:

• Loss: 3.7346
• Data Size: 1.0
• Epoch Runtime: 7.8480
• Bleu: 1.6243

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	8.8197	0	1.0293	0.0344
No log	1	88	6.7755	0.0078	1.4278	0.0920
No log	2	176	5.8588	0.0156	1.5437	0.1316
No log	3	264	5.4918	0.0312	2.1363	0.2228
No log	4	352	5.1599	0.0625	2.6050	0.3585
No log	5	440	4.7661	0.125	3.6915	0.5495
0.3907	6	528	4.3162	0.25	4.4763	0.5946
1.4424	7	616	3.9671	0.5	5.5326	0.9623
3.5809	8.0	704	3.6767	1.0	8.4031	1.1794
3.1766	9.0	792	3.5887	1.0	7.0507	1.5104
2.7967	10.0	880	3.5560	1.0	7.0607	1.5012
2.4757	11.0	968	3.5807	1.0	7.3833	1.5298
2.2014	12.0	1056	3.5927	1.0	7.3459	1.5771
1.9171	13.0	1144	3.6422	1.0	7.2252	1.5077
1.6564	14.0	1232	3.7346	1.0	7.8480	1.6243

Framework versions

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