mbart-neutralization

This model is a fine-tuned version of facebook/mbart-large-50 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0132
• Bleu: 98.6021
• Gen Len: 18.5104

Intended uses & limitations

Translating Spanish sentences and texts into "neutral", "inclusive" language

Training and evaluation data

Training and evaluation dataset: Spanish Gender Neutralization dataset

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5.6e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
No log	1.0	440	0.0408	97.3967	18.7604
0.2255	2.0	880	0.0132	98.6021	18.5104

Framework versions

• Transformers 4.50.0
• Pytorch 2.6.0+cu124
• Datasets 3.4.1
• Tokenizers 0.21.1