31284d845c7fb5efc54b1449a79aef6e

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

• Loss: 2.5512
• Data Size: 1.0
• Epoch Runtime: 117.0608
• Bleu: 7.0560

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	10.3234	0	9.8881	0.2357
No log	1	419	7.0170	0.0078	11.1775	0.3332
No log	2	838	4.1419	0.0156	12.3041	1.0306
0.1424	3	1257	3.1324	0.0312	14.7464	2.6167
0.1424	4	1676	2.7495	0.0625	18.2611	3.6170
0.1873	5	2095	2.5542	0.125	25.1893	4.2706
0.3458	6	2514	2.4098	0.25	38.3494	4.9854
2.2767	7	2933	2.2968	0.5	63.2737	6.8040
1.9998	8.0	3352	2.2114	1.0	116.3627	6.3216
1.7057	9.0	3771	2.2050	1.0	115.1445	6.1795
1.448	10.0	4190	3.3394	1.0	115.3325	5.4415
1.153	11.0	4609	2.3391	1.0	116.4938	7.0565
0.9932	12.0	5028	2.4412	1.0	117.8925	6.5043
0.7631	13.0	5447	2.5512	1.0	117.0608	7.0560

Framework versions

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