README.md

metadata

license: mit
tags:
  - image-classification
  - computer-vision
  - custom-model
  - cameroon-food
  - pytorch
library_name: transformers
model_name: cameroon-meals
language: en
datasets: []

🇨🇲 Cameroon Meals Classifier

...

A PyTorch image classification model trained to identify 37 traditional Cameroonian meals using a ResNet34 backbone.

---

🍽️ Classes (37 meals)

• Pepper Soup
• Kati Kati and Njama Njama
• Koki Beans
• Mbongo Tchobi
• Dodo
• Soya
• Chin-Chin
• Groundnut Soup
• Sese Plantains
• Okra Soup
• Puff Puff
• Beignet Haricot
• Egusi Soup
• Yassa Chicken
• Meat Pie
• Kwacoco Bible
• Kondre
• Roasted Plantain and Plum
• Banga Soup
• Ekwang
• Bobolo
• Black Soup
• Cornchaff
• Accra Banana
• Egusi Pudding
• Poulet DG
• Sangah
• Banane Malaxée
• Hot Pot Potatoes
• Groundnut Sweet
• Fish Roll
• Garri with Groundnuts
• Eru
• Ndolé
• Achu
• Jollof Rice

---

🧠 Model Details

• Backbone: ResNet50
• Framework: PyTorch
• Format: model_weights.pth
• Custom class: CameroonMealsModel (see modeling.py)

---

🚀 Usage

from transformers import AutoModel
import torch

model = AutoModel.from_pretrained("paulinusjua/cameroon-meals", trust_remote_code=True)
model.eval()

# Example forward (x: a batch of image tensors)
logits = model(x)