moudook commited on
Commit
e0da999
·
verified ·
1 Parent(s): a0551fa

Delete README.md

Browse files
Files changed (1) hide show
  1. README.md +0 -24
README.md DELETED
@@ -1,24 +0,0 @@
1
- # Plant Disease Detection ViT
2
-
3
- This model is a fine-tuned Vision Transformer (ViT) for classifying plant diseases using the PlantVillage dataset.
4
-
5
- ## Model Description
6
- - **Base Model:** google/vit-base-patch16-224 12 layers
7
- - **Task:** Image Classification (38 classes)
8
- - **Framework:** PyTorch & Hugging Face Transformers
9
-
10
- ## Training Configuration
11
- - **Image Size:** 224x224
12
- - **Batch Size:** 32
13
- - **Epochs:** 12
14
- - **Learning Rate:** 2e-05
15
- - **Weight Decay:** 0.0001
16
- - **Device:** cuda
17
- - **Train/Val/Test Split:** 75% / 15% / 10%
18
-
19
- ## Performance
20
- - **Best Validation Accuracy:** 99.36%
21
- - **Final Test Accuracy:** 99.32%
22
-
23
- ## Dataset
24
- Automated download of PlantVillage dataset containing 38 classes of healthy and diseased plant leaves.