indonesia-election-topic-classification-undersampling-double-ovr

This model is a fine-tuned version of indolem/indobertweet-base-uncased on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.5974
• Balanced Accuracy: 0.7667

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 14

Training results

Training Loss	Epoch	Step	Validation Loss	Balanced Accuracy
No log	1.0	201	1.0434	0.6784
No log	2.0	402	0.8096	0.7479
1.0106	3.0	603	1.0421	0.7634
1.0106	4.0	804	1.1982	0.7483
0.2636	5.0	1005	1.3720	0.7504
0.2636	6.0	1206	1.4238	0.7398
0.2636	7.0	1407	1.5469	0.7491
0.0802	8.0	1608	1.5200	0.7548
0.0802	9.0	1809	1.5341	0.7620
0.0464	10.0	2010	1.6248	0.7560
0.0464	11.0	2211	1.6143	0.7632
0.0464	12.0	2412	1.5974	0.7667
0.0285	13.0	2613	1.6216	0.7644
0.0285	14.0	2814	1.6438	0.7623

Framework versions

• Transformers 4.41.2
• Pytorch 2.1.2
• Datasets 2.19.2
• Tokenizers 0.19.1