mbart-en-es-ddhh

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

mbart-en-es-ddhh

This model is a fine-tuned version of facebook/mbart-large-50 on a dataset based on translation EN-ES of Universal Declaration of Human Rights (https://huggingface.co/datasets/ntr2026/DDHH_en-es). It achieves the following results on the evaluation set:

• Loss: 4.7211
• Bleu: 11.5131
• Gen Len: 20.05

Model description

This model has been delivered as part of a class project.

Intended uses & limitations

The model is aim to translating from English to Spanish. As the low BLEU score shows, this model has very low accuracy, given the limitation of the dataset.

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 1e-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
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 5

Training results

Framework versions

• Transformers 4.51.2
• Pytorch 2.10.0+cu128
• Datasets 4.7.0
• Tokenizers 0.21.4