Rosemary Disease Classifier
A fine-tuned google/vit-base-patch16-224 that classifies the condition of rosemary (Rosmarinus officinalis) leaves. Built for the AgroSense rosemary-monitoring app.
Model source
- Kind
- Other
- File size
- 1.35 KB–327 MB
- Version
- vRevision eeb9855
- Updated
- Sep 17, 2026
- Downloads
- 2
- Source
- Hugging Face
Source description
Rosemary Disease Classifier
A fine-tuned google/vit-base-patch16-224 that classifies the condition of rosemary
(Rosmarinus officinalis) leaves. Built for the AgroSense rosemary-monitoring app.
Classes
Rosemary, Rosemary_Leaf_Spot, Rosemary_Pest_Damage, Rosemary_Powdery_Mildew
Training
- Base model: google/vit-base-patch16-224
- Epochs: 12
- Batch size: 16
- Learning rate: 2e-05
- Augmentation: random resized crop, horizontal flip, small rotation
- Splits: dataset-provided train / validation / test
Evaluation (held-out test split)
- Accuracy: 0.9014
- Macro-F1: 0.9121
Limitations
Trained on a custom rosemary image set; works best on clear, well-lit photos of a single leaf or plant. It only knows the four classes listed above.
Sources
1 source
Sources
1 sourceVerified Sep 2
Download files
Model artifacts
5 artifacts
Model artifacts
5 artifactscheckpoint-308/model.safetensors
safetensors · 327 MB · SHA-256 f9bbabb45a96…f715 · Hugging Face
Download filecheckpoint-336/model.safetensors
safetensors · 327 MB · SHA-256 65e67623df84…5cb2 · Hugging Face
Download fileSource excerpts
2 excerpts
Source excerpts
2 excerpts--- language: - en license: mit tags: - image-classification - vit - agriculture - rosemary - plant-disease metrics: - accuracy - f1 pipeline_tag: image-classification --- # Rosemary Disease Classifier A fine-tuned `google/vit-base-patch16-224` that classifies the condition of **rosemary** (Rosmarinus officinalis) leaves. Built for the AgroSense rosemary-monitoring app. ## Classes Rosemary, Rosemary_Leaf_Spot, Rosemary_Pest_Damage, Rosemary_Powdery_Mildew ## Training - **Base model:** google/vit-base-patch16-224 - **Epochs:** 12 - **Batch size:** 16 - **Learning rate:** 2e-05 - **Augmentation:** random resized crop, horizontal flip, small rotation - **Splits:** dataset-provided train / validation / test ## Evaluation (held-out test split) - **Accuracy:** 0.9014 - **Macro-F1:** 0.9121 ## Limitations Trained on a custom rosemary image set; works best on clear, well-lit photos of a single leaf or plant. It only knows the four classes listed above.
Source context: 2 downloads · 0 likes · Pipeline image-classification · Repo asafe51/rosemary-disease-classifier
