e21e91e7789691f45e20639fa67acdb7

This model is a fine-tuned version of facebook/mbart-large-cc25 on the Helsinki-NLP/opus_books [es-nl] dataset. It achieves the following results on the evaluation set:

• Loss: 2.1591
• Data Size: 1.0
• Epoch Runtime: 211.7493
• Bleu: 8.9135

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	11.7714	0	18.3932	0.0852
No log	1	806	3.9850	0.0078	20.1238	2.1627
No log	2	1612	3.6031	0.0156	23.0132	2.8352
No log	3	2418	3.0832	0.0312	26.1991	3.6384
0.1182	4	3224	2.6769	0.0625	32.1997	4.3507
2.6709	5	4030	2.4286	0.125	44.0538	5.2008
2.4163	6	4836	2.2559	0.25	68.6989	5.7298
2.1599	7	5642	2.0984	0.5	116.7557	6.6180
1.8979	8.0	6448	1.9655	1.0	212.5366	8.1523
1.6867	9.0	7254	2.0046	1.0	213.5498	8.2413
1.4471	10.0	8060	2.0336	1.0	211.5121	8.3137
1.255	11.0	8866	2.0753	1.0	210.9009	16.0797
1.017	12.0	9672	2.1591	1.0	211.7493	8.9135

Framework versions

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