sentiment-unipelt-1

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

• Loss: 0.3288
• Accuracy: 0.8872
• Precision: 0.8606
• Recall: 0.8727
• F1: 0.8662

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: 30
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20.0

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
0.5528	1.0	122	0.5051	0.7168	0.6521	0.6396	0.6444
0.4431	2.0	244	0.4333	0.7970	0.7711	0.8189	0.7789
0.3491	3.0	366	0.3269	0.8672	0.8579	0.8110	0.8293
0.2886	4.0	488	0.2995	0.8722	0.8498	0.8371	0.8430
0.2671	5.0	610	0.2967	0.8697	0.8428	0.8428	0.8428
0.2481	6.0	732	0.3346	0.8546	0.8230	0.8647	0.8365
0.2339	7.0	854	0.2965	0.8672	0.8372	0.8485	0.8425
0.217	8.0	976	0.2929	0.8772	0.8576	0.8406	0.8484
0.1984	9.0	1098	0.2778	0.8797	0.8572	0.8499	0.8534
0.1937	10.0	1220	0.2905	0.8772	0.8628	0.8331	0.8458
0.1734	11.0	1342	0.2919	0.8872	0.8624	0.8677	0.8650
0.1561	12.0	1464	0.3126	0.8797	0.8539	0.8574	0.8556
0.1579	13.0	1586	0.3165	0.8822	0.8610	0.8517	0.8561
0.1459	14.0	1708	0.3108	0.8972	0.8715	0.8873	0.8787
0.1399	15.0	1830	0.3192	0.8922	0.8657	0.8813	0.8728
0.1377	16.0	1952	0.3200	0.8847	0.8567	0.8734	0.8642
0.1259	17.0	2074	0.3303	0.8847	0.8589	0.8659	0.8623
0.1293	18.0	2196	0.3265	0.8922	0.8657	0.8813	0.8728
0.1187	19.0	2318	0.3305	0.8847	0.8589	0.8659	0.8623
0.1302	20.0	2440	0.3288	0.8872	0.8606	0.8727	0.8662

Framework versions

• Transformers 4.39.3
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.15.2