bert-indo-base-stance-cls

This model is a fine-tuned version of indobenchmark/indobert-base-p1 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.0156
• Accuracy: 0.6892
• Precision: 0.6848
• Recall: 0.6892
• F1: 0.6859
• Against: {'precision': 0.6185567010309279, 'recall': 0.5555555555555556, 'f1-score': 0.5853658536585366, 'support': 216}
• For: {'precision': 0.7280453257790368, 'recall': 0.7764350453172205, 'f1-score': 0.7514619883040935, 'support': 331}

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: 16
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1	Against	For
No log	1.0	137	0.6423	0.6581	0.6894	0.6581	0.5917	{'precision': 0.7543859649122807, 'recall': 0.19907407407407407, 'f1-score': 0.31501831501831506, 'support': 216}	{'precision': 0.6469387755102041, 'recall': 0.9577039274924471, 'f1-score': 0.7722289890377587, 'support': 331}
No log	2.0	274	0.6146	0.6600	0.6691	0.6600	0.6628	{'precision': 0.5614754098360656, 'recall': 0.6342592592592593, 'f1-score': 0.5956521739130436, 'support': 216}	{'precision': 0.7392739273927392, 'recall': 0.676737160120846, 'f1-score': 0.7066246056782334, 'support': 331}
No log	3.0	411	0.7572	0.6545	0.6734	0.6545	0.6583	{'precision': 0.550561797752809, 'recall': 0.6805555555555556, 'f1-score': 0.608695652173913, 'support': 216}	{'precision': 0.7535714285714286, 'recall': 0.6374622356495468, 'f1-score': 0.6906710310965631, 'support': 331}
0.4855	4.0	548	0.7405	0.6892	0.6842	0.6892	0.6851	{'precision': 0.6210526315789474, 'recall': 0.5462962962962963, 'f1-score': 0.5812807881773399, 'support': 216}	{'precision': 0.7254901960784313, 'recall': 0.7824773413897281, 'f1-score': 0.7529069767441859, 'support': 331}
0.4855	5.0	685	1.1222	0.6856	0.6828	0.6856	0.6839	{'precision': 0.6078431372549019, 'recall': 0.5740740740740741, 'f1-score': 0.5904761904761905, 'support': 216}	{'precision': 0.7317784256559767, 'recall': 0.7583081570996979, 'f1-score': 0.7448071216617211, 'support': 331}
0.4855	6.0	822	1.4960	0.6892	0.6830	0.6892	0.6827	{'precision': 0.6292134831460674, 'recall': 0.5185185185185185, 'f1-score': 0.5685279187817258, 'support': 216}	{'precision': 0.7181571815718157, 'recall': 0.8006042296072508, 'f1-score': 0.7571428571428572, 'support': 331}
0.4855	7.0	959	1.6304	0.6801	0.6886	0.6801	0.6827	{'precision': 0.5843621399176955, 'recall': 0.6574074074074074, 'f1-score': 0.6187363834422658, 'support': 216}	{'precision': 0.756578947368421, 'recall': 0.6948640483383686, 'f1-score': 0.7244094488188976, 'support': 331}
0.1029	8.0	1096	1.8381	0.6673	0.6727	0.6673	0.6693	{'precision': 0.5726495726495726, 'recall': 0.6203703703703703, 'f1-score': 0.5955555555555555, 'support': 216}	{'precision': 0.7380191693290735, 'recall': 0.6978851963746223, 'f1-score': 0.717391304347826, 'support': 331}
0.1029	9.0	1233	1.9474	0.6929	0.6876	0.6929	0.6881	{'precision': 0.6290322580645161, 'recall': 0.5416666666666666, 'f1-score': 0.582089552238806, 'support': 216}	{'precision': 0.7257617728531855, 'recall': 0.7915407854984894, 'f1-score': 0.7572254335260115, 'support': 331}
0.1029	10.0	1370	2.0156	0.6892	0.6848	0.6892	0.6859	{'precision': 0.6185567010309279, 'recall': 0.5555555555555556, 'f1-score': 0.5853658536585366, 'support': 216}	{'precision': 0.7280453257790368, 'recall': 0.7764350453172205, 'f1-score': 0.7514619883040935, 'support': 331}

Framework versions

• Transformers 4.24.0
• Pytorch 1.12.1+cu113
• Datasets 2.7.0
• Tokenizers 0.13.2