bert-base-thai-intent-booking

This model is a fine-tuned version of monsoon-nlp/bert-base-thai on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.1511
• Accuracy: 0.1937
• F1: 0.1641
• Precision: 0.2236
• Recall: 0.1937

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: 16
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 64
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
2.368	1.0	65	2.3624	0.0901	0.0149	0.0081	0.0901
2.3586	2.0	130	2.2374	0.1577	0.0963	0.1107	0.1577
2.2987	3.0	195	2.2589	0.1216	0.0626	0.0890	0.1216
2.279	4.0	260	2.1771	0.1757	0.1329	0.2163	0.1757
2.2326	5.0	325	2.2099	0.2027	0.1548	0.1497	0.2027
2.2273	6.0	390	2.1809	0.1712	0.1245	0.1127	0.1712
2.2303	7.0	455	2.2168	0.1486	0.1030	0.1190	0.1486
2.196	8.0	520	2.1862	0.1937	0.1478	0.1615	0.1937
2.1848	9.0	585	2.1320	0.2162	0.1773	0.2192	0.2162
2.183	10.0	650	2.1771	0.1712	0.1240	0.1703	0.1712
2.1669	11.0	715	2.1672	0.2117	0.1849	0.2453	0.2117
2.1586	12.0	780	2.1237	0.2162	0.1939	0.3552	0.2162
2.1465	13.0	845	2.1269	0.2117	0.1834	0.2440	0.2117
2.1454	14.0	910	2.1160	0.2162	0.1939	0.3552	0.2162
2.1404	15.0	975	2.1089	0.2162	0.1936	0.3561	0.2162
2.1293	16.0	1040	2.1272	0.2162	0.1947	0.3584	0.2162
2.1193	17.0	1105	2.1043	0.2117	0.1836	0.2431	0.2117
2.1094	18.0	1170	2.1053	0.2117	0.1895	0.2977	0.2117
2.1063	19.0	1235	2.1055	0.2117	0.1901	0.2989	0.2117
2.0888	20.0	1300	2.1067	0.2117	0.1895	0.2977	0.2117

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 3.0.1
• Tokenizers 0.19.1