nlp_vietnamese_spelling

This model is a fine-tuned version of vinai/bartpho-syllable on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.1084
• Sacrebleu: 19.1727

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: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 3
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Sacrebleu
1.0565	0.25	1000	0.4944	13.1438
0.5878	0.5	2000	0.2775	16.2433
0.4367	0.75	3000	0.2222	17.0610
0.3627	1.0	4000	0.1793	17.9412
0.2672	1.25	5000	0.1533	18.3610
0.2305	1.5	6000	0.1429	18.4508
0.2164	1.75	7000	0.1296	18.8490
0.1948	2.01	8000	0.1205	19.0083
0.1537	2.26	9000	0.1167	19.0106
0.1491	2.51	10000	0.1131	19.0764
0.1414	2.76	11000	0.1084	19.1727

Framework versions

• Transformers 4.39.3
• Pytorch 2.1.2
• Datasets 2.18.0
• Tokenizers 0.15.2