###version of node to run the program Version of node : v20.19.1

###Testing on server site: node index.js on client site: curl -X POST -F "[image_path]" http://localhost:3000/predict

When run successfully, the terminal should be as example: Loading model from: /Users/admin/Desktop/University/DADN/tfjs-node-server/model/model.json Server running on http://localhost:3000 Model loaded successfully. POST /predict accessed Predictions: Float32Array(6) [ 1.2050749731429278e-8, 3.317679642123039e-8, 0.00047478932538069785, 0.000004734285539598204, 0.2472582757472992, 0.7522621750831604 ] 🧹 Uploaded file deleted. POST /predict accessed Predictions: Float32Array(6) [ 1.2050749731429278e-8, 3.317679642123039e-8, 0.00047478932538069785, 0.000004734285539598204, 0.2472582757472992, 0.7522621750831604 ] 🧹 Uploaded file deleted.

On the client site will be the class and the predictions for each class(trained with tensor flow): curl -X POST -F "image=@/Users/admin/Desktop/University/DADN/tfjs-node-server/img/download.jpeg" http://localhost:3000/predict

{"predictions":[0,0,0.05,0,24.73,75.23]}%