Model Analysis

Purpose: Verify that the loaded .h5 model has the correct architecture, input shape, and layer flow required for NDP conversion.

How to Use

• Load the .h5 model
• Click “Visualize Model”
• The architecture diagram will be displayed automatically

What This Diagram Shows The diagram displays the quantized model structure layer by layer:

• Input shape of the model (e.g., 40 × 40 × 1)
• Order of layers (Conv → Conv → Conv → Flatten → Dense → Softmax)
• Output shape after each layer
• Presence of Quantize wrappers (QuantizeWrapperV2)
• Final number of output classes (e.g., 5)