byt5-small-finetune-thai-to-romanized

This model is a fine-tuned version of google/byt5-small on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0881
• Rouge1: 7.9223
• Rouge2: 0.5516
• Rougel: 7.9176
• Rougelsum: 7.9205
• Gen Len: 11.8775

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: 2e-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: 8

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
1.8962	1.0	1500	1.6419	0.0348	0.0	0.0351	0.0345	11.6767
1.1867	2.0	3000	0.6487	1.4739	0.0463	1.4793	1.4804	11.6179
0.6108	3.0	4500	0.2648	4.9434	0.2338	4.9373	4.9413	11.6379
0.363	4.0	6000	0.1561	6.8898	0.364	6.8827	6.8935	11.808
0.2671	5.0	7500	0.1132	7.5578	0.4855	7.5496	7.5604	11.8507
0.2198	6.0	9000	0.0976	7.7876	0.511	7.7819	7.7864	11.8772
0.1823	7.0	10500	0.0920	7.883	0.5422	7.8797	7.8834	11.8707
0.1854	8.0	12000	0.0881	7.9223	0.5516	7.9176	7.9205	11.8775

Framework versions

• Transformers 4.26.1
• Pytorch 1.13.1+cu116
• Datasets 2.9.0
• Tokenizers 0.13.2