mbart50-es-zapv2

This model is a fine-tuned version of facebook/mbart-large-50 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.2902
• Bleu: 18.7613
• Ter: 69.5611
• Meteor: 0.4206

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: 3e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 50
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Ter	Meteor
5.9172	1.0	179	0.4674	3.6424	91.5076	0.2051
0.6639	2.0	358	0.7822	0.0	100.0	0.0
0.4960	3.0	537	0.2733	8.1370	84.6374	0.2895
0.1756	4.0	716	0.2302	16.6049	77.2901	0.4079
0.1186	5.0	895	0.2245	18.6545	69.3702	0.4340
0.0656	6.0	1074	0.2474	20.0563	68.9885	0.4478
0.0496	7.0	1253	0.3485	14.4966	76.7176	0.3331
0.0469	8.0	1432	0.2902	18.7613	69.5611	0.4206

Framework versions

• Transformers 5.4.0
• Pytorch 2.10.0+cu128
• Datasets 4.8.4
• Tokenizers 0.22.2