const fs = require('fs');
const sharp = require('sharp');
const ort = require('onnxruntime-node');

(async () => {
  try {
    // Step 1: Load and preprocess the image
    const imageBuffer = await sharp('./training_images/shirt/00e745c9-97d9-429d-8c3f-d3db7a2d2991.jpg')
      .resize(128, 128) // Resize to 128x128
      .raw() // Get raw pixel data
      .toBuffer();

    // Convert to Float32 and normalize pixel values to [0, 1]
    const imgArray = Float32Array.from(imageBuffer).map(value => value / 255.0);

    // Add batch dimension [1, 128, 128, 3]
    const inputTensor = new ort.Tensor('float32', imgArray, [1, 128, 128, 3]);

    // Step 2: Load ONNX model
    const session = await ort.InferenceSession.create('./saved-model/model.onnx');

    // Step 3: Run inference
    const results = await session.run({ [session.inputNames[0]]: inputTensor });

    console.log('Inference outputs:', results[session.outputNames[0]]);
  } catch (err) {
    console.error('Error:', err);
  }
})();