d371d4d8be4be5eb4e90652fdaac9fee

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

• Loss: 1.6929
• Data Size: 1.0
• Epoch Runtime: 410.6452
• Bleu: 11.5005

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	5.7046	0	28.5595	1.1816
No log	1	806	4.0089	0.0078	32.0883	5.4944
No log	2	1612	3.1023	0.0156	42.0395	9.7847
No log	3	2418	2.6106	0.0312	52.0374	14.2410
0.1211	4	3224	2.2476	0.0625	64.4661	7.1959
2.5793	5	4030	1.9795	0.125	87.4268	7.8127
2.2415	6	4836	1.8527	0.25	131.6558	8.7641
2.0493	7	5642	1.7388	0.5	225.0196	9.4379
1.8549	8.0	6448	1.6425	1.0	412.5192	10.3146
1.6668	9.0	7254	1.6054	1.0	412.9333	10.7869
1.5133	10.0	8060	1.5867	1.0	412.3394	11.0634
1.4061	11.0	8866	1.5829	1.0	412.4480	11.2502
1.276	12.0	9672	1.6010	1.0	411.0898	11.3751
1.1394	13.0	10478	1.6242	1.0	411.5901	11.4016
1.0515	14.0	11284	1.6554	1.0	411.0146	11.4973
0.9416	15.0	12090	1.6929	1.0	410.6452	11.5005

Framework versions

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