c780d08fbf2b35fcec77e0687d19d66d

This model is a fine-tuned version of facebook/mbart-large-50 on the Helsinki-NLP/opus_books [es-no] dataset. It achieves the following results on the evaluation set:

• Loss: 4.0414
• Data Size: 1.0
• Epoch Runtime: 26.7309
• Bleu: 8.5570

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	11.5604	0	2.4056	0.0762
No log	1	89	9.9686	0.0078	3.0878	0.3271
No log	2	178	9.1052	0.0156	4.7822	0.2315
No log	3	267	8.5607	0.0312	6.9643	0.3833
No log	4	356	7.8710	0.0625	8.4867	0.8000
No log	5	445	6.7450	0.125	11.4297	1.0009
0.507	6	534	5.1633	0.25	12.6417	1.9903
1.7566	7	623	3.7198	0.5	16.4967	2.7706
13.7535	8.0	712	10.7252	1.0	27.8183	0.0000
3.9716	9.0	801	3.7041	1.0	27.2319	2.9232
2.6366	10.0	890	3.1651	1.0	26.2992	4.3056
1.8373	11.0	979	3.2687	1.0	26.4417	4.5739
1.2602	12.0	1068	3.4791	1.0	27.9064	4.4266
0.8923	13.0	1157	3.7760	1.0	28.0623	4.9271
0.5967	14.0	1246	4.0414	1.0	26.7309	8.5570

Framework versions

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