ntviet/mbart-hre-viet1.2

mBART Hre Vietnamese translation 1.2

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

• Loss: 0.0012
• Bleu: 83.4928

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: 5e-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: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu
No log	1.0	336	0.1086	4.3490
0.5946	2.0	672	0.0617	10.2703
0.0834	3.0	1008	0.0302	39.4194
0.0834	4.0	1344	0.0135	56.0138
0.0252	5.0	1680	0.0077	69.3363
0.0097	6.0	2016	0.0046	77.2427
0.0097	7.0	2352	0.0037	81.1891
0.004	8.0	2688	0.0021	85.4589
0.0025	9.0	3024	0.0014	85.5600
0.0025	10.0	3360	0.0012	83.4928

Framework versions

• Transformers 4.48.2
• Pytorch 2.5.1+cu124
• Tokenizers 0.21.0