This model is fine-tuned on agricultural datasets for tasks like crop recommendation, disease detection, and yield prediction. It is based on the unsloth/Phi-4 model and uses reinforcement learning for improved performance.
Evaluation Results
Task
Metric
Score
Crop Recommendation
Accuracy
92%
Disease Detection
F1 Score
0.91
Yield Prediction
RMSE
0.15
Developed by: DARJYO
Fine-tuned: persadian_14B-GRPO
Base Architecture: Transformer-based/Phi-4
Model type: Fine-tuned language model for agricultural tasks.