indonesian-distilbert-base-cased-finetuned-indonlu

This model is a fine-tuned version of Toshifumi/distilbert-base-multilingual-cased-finetuned-emotion on the indonlu dataset. It achieves the following results on the evaluation set:

• Loss: 1.1300
• Accuracy: 0.6114
• Precision: 0.6058
• Recall: 0.6114
• F1: 0.6069

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: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.01
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
No log	1.0	221	1.2623	0.475	0.4817	0.475	0.4458
No log	2.0	442	1.0937	0.55	0.5555	0.55	0.5444
1.2289	3.0	663	1.0749	0.5886	0.6003	0.5886	0.5898
1.2289	4.0	884	1.0836	0.5818	0.6019	0.5818	0.5800
0.7857	5.0	1105	1.1300	0.6114	0.6058	0.6114	0.6069
0.7857	6.0	1326	1.1595	0.6	0.5996	0.6	0.5984
0.5288	7.0	1547	1.1767	0.6	0.5986	0.6	0.5958
0.5288	8.0	1768	1.2195	0.6	0.5969	0.6	0.5952
0.5288	9.0	1989	1.2422	0.5932	0.5915	0.5932	0.5909
0.3685	10.0	2210	1.2406	0.5841	0.5842	0.5841	0.5830

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu121
• Datasets 2.17.1
• Tokenizers 0.15.2