johannawawi/model-for-sosmed-analysis

results_final

This model is a fine-tuned version of w11wo/indonesian-roberta-base-sentiment-classifier on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.4153
• Accuracy: 0.8881
• F1 Macro: 0.8881
• F1 Weighted: 0.8881
• Precision Macro: 0.8885
• Recall Macro: 0.8881
• Precision Weighted: 0.8885
• Recall Weighted: 0.8881

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: 0.001
• train_batch_size: 8
• eval_batch_size: 8
• 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: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 8

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1 Macro	F1 Weighted	Precision Macro	Recall Macro	Precision Weighted	Recall Weighted
0.4974	1.7182	500	0.5158	0.8399	0.8392	0.8393	0.8496	0.8397	0.8495	0.8399
0.4026	3.4364	1000	0.5560	0.8692	0.8683	0.8683	0.8760	0.8692	0.8761	0.8692
0.4068	5.1546	1500	0.3820	0.8864	0.8863	0.8863	0.8872	0.8864	0.8872	0.8864
0.2641	6.8729	2000	0.4153	0.8881	0.8881	0.8881	0.8885	0.8881	0.8885	0.8881

Framework versions

• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 2.14.4
• Tokenizers 0.21.1