Howest animal classifier (cats / dogs / panda)
Small Keras CNN trained in an Azure Machine Learning pipeline (Howest MCT MLOps, assignment 1.2).
- Input: RGB image, 64x64, pixel values scaled to [0, 1]
- Output: softmax over
["cats", "dogs", "panda"] - Architecture: 3x (Conv2D + MaxPooling2D) -> Dense(64) -> Dropout(0.5) -> Dense(3)
- Training: 2400 images, 5 epochs, test accuracy 68.8% on 600 images
import keras
model = keras.saving.load_model("hf://jonyyd/howest-hf-ai")
- Downloads last month
- 3,240