AgroMind Plant Disease Classifier (NFNet-F1)

Model Description

NFNet-F1 image classifier trained to detect 88 plant disease classes across multiple crop types including Apple, Cassava, Cherry, Chili, Coffee, Corn, Cucumber, Grape, Mango, Peach, Pepper, Pomegranate, Potato, Rice, Soybean, Strawberry, Sugarcane, Tea, Tomato, and Wheat.

Framework

• Architecture: NFNet-F1 (via timm)
• Format: SafeTensors
• Input size: 512×512 RGB
• Normalization: mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5]

Usage

from huggingface_hub import hf_hub_download
import timm
from safetensors.torch import load_file
import json, torch
from torchvision import transforms
from PIL import Image

repo_id = "Arko007/agromind-plant-disease-nfnet"
model_path = hf_hub_download(repo_id, "model.safetensors")
config_path = hf_hub_download(repo_id, "config.json")

with open(config_path) as f:
    config = json.load(f)

model = timm.create_model(config["architecture"], pretrained=False, num_classes=config["num_classes"])
state_dict = load_file(model_path)
model.load_state_dict(state_dict)
model.eval()

transform = transforms.Compose([
    transforms.Resize((512, 512)),
    transforms.ToTensor(),
    transforms.Normalize(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5]),
])

img = Image.open("leaf.jpg").convert("RGB")
with torch.no_grad():
    logits = model(transform(img).unsqueeze(0))
    pred = logits.argmax(dim=1).item()
    print(config["class_names"][pred])

Output

Returns logits for 88 plant disease/healthy classes. See config.json for the full class list.