d833d7de81038910f3292d9b1e33c5ef

This model is a fine-tuned version of google/umt5-xl on the Helsinki-NLP/opus_books [en-nl] dataset. It achieves the following results on the evaluation set:

• Loss: 1.7103
• Data Size: 1.0
• Epoch Runtime: 480.1061
• Bleu: 10.3820

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	5.7233	0	33.5297	1.4231
No log	1	966	3.8984	0.0078	37.3044	6.8565
No log	2	1932	2.9321	0.0156	48.6461	12.2803
0.0936	3	2898	2.4543	0.0312	60.2072	15.0861
2.8404	4	3864	2.1480	0.0625	72.3568	6.6557
2.4842	5	4830	1.9814	0.125	100.7990	7.3004
2.2809	6	5796	1.8787	0.25	153.3237	8.0975
2.0426	7	6762	1.7809	0.5	262.3911	8.7590
1.8457	8.0	7728	1.6847	1.0	486.4543	9.4688
1.6997	9.0	8694	1.6481	1.0	486.1958	9.8143
1.545	10.0	9660	1.6329	1.0	479.0897	10.0536
1.3679	11.0	10626	1.6401	1.0	480.3823	10.2404
1.2752	12.0	11592	1.6513	1.0	481.5326	10.2976
1.162	13.0	12558	1.6773	1.0	480.6777	10.3630
1.0743	14.0	13524	1.7103	1.0	480.1061	10.3820

Framework versions

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