Object Detection using Mojo and Yolov5s in Max engine

Intro to Machine Learning (2018)
1

Issues in loading and compiling the yolov5s.torchscript model in MAX Engine using Mojo. Referred the documents at https://docs.modular.com/max/mojo/get-started and https://docs.modular.com/max/model-formats, was not able to load the yolov5s model to perform object detection. Also tried loading the model by referring to the recent published article https://www.modular.com/blog/bring-your-own-pytorch-model leading to no success.

# Import necessary modules from MAX Engine
from max.engine import InferenceSession

# Define the main function (entry point)
def main():
    # Specify the path to your TorchScript model file
    var model_path = "/home/user/yolov5s.torchscript"

    # Create an instance of InferenceSession
    var session = InferenceSession()

    # Load the model
    var model = session.load(model_path)

    # Print model information (optional)
    print("Model loaded successfully.")

Tried the above code (worked without errors in the Mojo REPL) and got

mojo user :~$ mojo load_model.mojo Compiling model… Done! Unhandled exception caught during execution: Torch model is dynamic but does not have input specs set! Please set input specs before compiling the model. mojo: error: execution exited with a non-zero result: 1|