mbart-LenguajeClaro

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: 2.1078
• Bleu: 15.5051
• Gen Len: 119.5331

Intended uses & limitations

This model has been specially designed to improve the readability of Spanish texts, making them easier to understand and more engaging. It can be used for: Simplifying complex sentences Adjusting vocabulary for clarity. Enhacing readability for specific audiences.

Training and evaluation data

The model has been trained on a small database (easy-read). Therefore, the results can be very variable.

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5.6e-05
• train_batch_size: 4
• eval_batch_size: 4
• 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	241	2.1494	14.3395	108.657
No log	2.0	482	2.1078	15.5051	119.5331

Framework versions

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