| Property | Value |
|---|---|
| License | MIT |
| Base Model | cahya/bert-base-indonesian-1.5G |
| Task | Sentiment Analysis |
| Accuracy | 93.73% |
| Downloads | 125,074 |
What is bert-base-indonesian-1.5G-sentiment-analysis-smsa?
This is a specialized sentiment analysis model fine-tuned on the IndoNLU dataset using the BERT base Indonesian architecture. The model represents a significant achievement in Indonesian language processing, demonstrating exceptional accuracy in sentiment classification tasks.
Implementation Details
The model was trained using carefully optimized hyperparameters, including a learning rate of 2e-05, batch size of 16, and Adam optimizer. The training process spanned 10 epochs, showing consistent improvement in performance with the best validation loss of 0.3390.
- Linear learning rate scheduler
- Adam optimizer with betas=(0.9,0.999)
- Trained on PyTorch framework
- Implements transformer architecture
Core Capabilities
- High-accuracy sentiment analysis for Indonesian text
- Achieves 93.73% accuracy on evaluation set
- Optimized for production deployment
- Supports inference endpoints
Frequently Asked Questions
Q: What makes this model unique?
This model stands out for its exceptional accuracy in Indonesian sentiment analysis, achieved through careful fine-tuning of the BERT base architecture. The training process shows stable improvement without overfitting, making it reliable for production use.
Q: What are the recommended use cases?
The model is ideal for Indonesian text sentiment analysis tasks, including social media monitoring, customer feedback analysis, and automated content moderation. It's particularly suitable for applications requiring high-accuracy sentiment classification in Indonesian language contexts.