elidle/indobert-post-training-fin-sa-exp-2

indobert-post-training-fin-sa-4

This model is a fine-tuned version of elidle/indobert-fin_news-mlm-3 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2555
• Accuracy: 0.9615

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: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 8
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.0597	0.1961	10	0.6693	0.7527
0.5333	0.3922	20	0.4162	0.8187
0.3855	0.5882	30	0.2527	0.9176
0.2547	0.7843	40	0.2264	0.9341
0.1848	0.9804	50	0.2274	0.9231
0.1865	1.1765	60	0.1852	0.9451
0.1104	1.3725	70	0.1674	0.9615
0.0857	1.5686	80	0.1744	0.9505
0.123	1.7647	90	0.1966	0.9451
0.1361	1.9608	100	0.1715	0.9451
0.0433	2.1569	110	0.1485	0.9670
0.0244	2.3529	120	0.1727	0.9615
0.0542	2.5490	130	0.2046	0.9560
0.0673	2.7451	140	0.3079	0.9396
0.0993	2.9412	150	0.2305	0.9560
0.0052	3.1373	160	0.2376	0.9560
0.0678	3.3333	170	0.2565	0.9615
0.0095	3.5294	180	0.2565	0.9560
0.0233	3.7255	190	0.2555	0.9615

Framework versions

• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.1