absa_model

This model is a fine-tuned version of indobenchmark/indobert-base-p2 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.6475
• Accuracy: 0.8392
• F1: 0.8409
• Precision: 0.8444
• Recall: 0.8431

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: 16
• eval_batch_size: 16
• seed: 42
• 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	F1	Precision	Recall
No log	1.0	72	0.6475	0.8392	0.8409	0.8444	0.8431
No log	2.0	144	0.8639	0.8252	0.8156	0.8261	0.8177
No log	3.0	216	0.9170	0.7832	0.7800	0.8112	0.7676
No log	4.0	288	0.8206	0.8322	0.8359	0.8346	0.8405
No log	5.0	360	0.8318	0.8392	0.8417	0.8434	0.8404
No log	6.0	432	0.9578	0.8252	0.8255	0.8243	0.8325
0.0684	7.0	504	0.9713	0.8112	0.8027	0.8143	0.7967
0.0684	8.0	576	0.9850	0.8252	0.8137	0.8236	0.8089
0.0684	9.0	648	0.9955	0.8392	0.8258	0.8347	0.8203
0.0684	10.0	720	0.9964	0.8392	0.8258	0.8347	0.8203

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu121
• Datasets 2.15.0
• Tokenizers 0.15.0