0ab34be67e04e1847924ec391a83f215

This model is a fine-tuned version of google/mt5-xl on the Helsinki-NLP/opus_books [en-nl] dataset. It achieves the following results on the evaluation set:

• Loss: 1.6293
• Data Size: 1.0
• Epoch Runtime: 460.7426
• Bleu: 8.8811

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	8.2399	0	29.8236	0.0181
No log	1	966	3.7281	0.0078	33.2101	0.2949
No log	2	1932	2.8113	0.0156	44.5927	0.6551
0.0946	3	2898	2.2448	0.0312	55.6916	6.8671
2.6048	4	3864	2.0034	0.0625	69.1498	5.3386
2.3506	5	4830	1.8860	0.125	94.4882	5.8919
2.1551	6	5796	1.7800	0.25	144.9390	6.6040
1.9301	7	6762	1.6820	0.5	249.9024	7.2385
1.7261	8.0	7728	1.5819	1.0	461.0209	8.0034
1.5754	9.0	8694	1.5438	1.0	458.6111	8.3994
1.4337	10.0	9660	1.5312	1.0	458.6185	8.6294
1.2466	11.0	10626	1.5482	1.0	461.1795	8.7757
1.1534	12.0	11592	1.5559	1.0	462.2312	8.7856
1.0389	13.0	12558	1.5948	1.0	462.1568	8.8159
0.9576	14.0	13524	1.6293	1.0	460.7426	8.8811

Framework versions

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