---
language: id
license: mit
tags:
  - image-classification
  - plant-identification
  - pytorch
  - cnn
  - vision-transformer
  - efficientnet
  - vit
datasets:
  - gumi-banten
metrics:
  - accuracy
model-index:
  - name: CNN EfficientNet-B4 — Gumi Banten
    results:
      - task:
            type: image-classification
        metrics:
          - type: accuracy
            value: 0.9851
---

# 🌿 Gumi Banten Plant Identification — CNN EfficientNet-B4

Model identifikasi tanaman Gumi Banten menggunakan arsitektur hybrid **CNN (EfficientNet-B4) + Vision Transformer (ViT)**.

## 📊 Performa Model
| Metrik | Nilai |
|--------|-------|
| **Test Accuracy** | 0.9851 (98.51%) |
| **Best Val Accuracy** | 0.9876 (98.76%) |
| **Jumlah Kelas** | 10 |

## 🏗️ Arsitektur
- **CNN Backbone**: EfficientNet-B4 (pretrained, global_pool=avg)
- **Classifier Head**: Dropout → Linear(1792→512) → GELU → Dropout → Linear(512→N)
- **Input Size**: 224×224 px
- **Framework**: PyTorch (weights disimpan sebagai HDF5/.h5)

## 📦 File dalam Repo
| File | Deskripsi |
|------|-----------|
| `gumi_banten_cnn_vit.h5` | Model weights dalam format HDF5 |
| `best_model.pth` | Checkpoint PyTorch lengkap |
| `config.json` | Konfigurasi model |
| `class_names.txt` | Daftar nama kelas |

## 🚀 Cara Menggunakan

### Load dari .pth (PyTorch — Direkomendasikan)
```python
import torch
from huggingface_hub import hf_hub_download

pth_path = hf_hub_download(repo_id="Wisnu1354/CNN-GumiBanten", filename="best_model.pth")
ckpt = torch.load(pth_path, map_location='cpu')
model.load_state_dict(ckpt['model_state'])
model.eval()
```

### Load dari .h5
```python
import h5py, torch, numpy as np
from huggingface_hub import hf_hub_download

def load_from_h5(h5_path, model_class, cfg):
    with h5py.File(h5_path, 'r') as hf:
        class_names = list(hf['metadata/class_names'][:])
        state_dict = {}
        def _load(name, obj):
            if isinstance(obj, h5py.Dataset):
                state_dict[name.replace('/', '.')] = torch.tensor(obj[()])
        hf['model_weights'].visititems(_load)
    model = model_class(cfg)
    model.load_state_dict(state_dict)
    model.eval()
    return model, class_names

h5_path = hf_hub_download(repo_id="Wisnu1354/CNN-GumiBanten", filename="gumi_banten_cnn_vit.h5")
model, class_names = load_from_h5(h5_path, EfficientNetB4, CFG)
```

## 📋 Kelas yang Didukung
- Daun Ancak
- Daun Base
- Daun Bila
- Daun Bingin
- Daun Dapdap
- Daun Intaran
- Daun Kayu Tulak
- Daun Kelor
- Daun Nagasari
- Daun Pucuk Rejuna

## 🧑‍💻 Training
Dilatih di Google Colab menggunakan GPU A100/V100/T4.

---
*Dibuat oleh: Gumi Banten Research | 2026-04-28*