mbart-large-50-en-es-translation-lr-1e-05-weight-decay-0.0

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: 0.9549
• Bleu: 45.0307
• Rouge: {'rouge1': 0.7049318825090395, 'rouge2': 0.5238048751750992, 'rougeL': 0.684187379601513, 'rougeLsum': 0.6843574853855577}

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: 1e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Rouge
1.4627	1.0	4500	1.0255	42.1880	{'rouge1': 0.6725633216905762, 'rouge2': 0.48605402524493657, 'rougeL': 0.6498853764470456, 'rougeLsum': 0.6501981166312041}
0.8878	2.0	9000	0.9572	44.1734	{'rouge1': 0.6912686406245903, 'rouge2': 0.5093695171345348, 'rougeL': 0.6701896043455414, 'rougeLsum': 0.6703473419504804}
0.7125	3.0	13500	0.9414	44.8709	{'rouge1': 0.7051197958532004, 'rouge2': 0.5210482863677958, 'rougeL': 0.6843075431636916, 'rougeLsum': 0.6846265298079588}
0.6092	4.0	18000	0.9549	45.0821	{'rouge1': 0.7047932899349161, 'rouge2': 0.523739339466653, 'rougeL': 0.6840127607742443, 'rougeLsum': 0.684202100852132}

Framework versions

• Transformers 4.33.0.dev0
• Pytorch 2.0.1+cu118
• Datasets 2.14.4.dev0
• Tokenizers 0.13.3