minimal.py

Go to the documentation of this file.

from onert import infer
import sys
 
 
def main(nnpackage_path, backends="cpu"):
    # Create session and load nnpackage
    # The default value of backends is "cpu".
    session = infer.session(nnpackage_path, backends)
 
    # Prepare input. Here we just allocate dummy input arrays.
    input_size = session.input_size()
    session.set_inputs(input_size)
 
    outputs = session.inference()
 
    print(f"nnpackage {nnpackage_path.split('/')[-1]} runs successfully.")
    return
 
 
def main(nnpackage_path, backends="cpu"): …
if __name__ == "__main__":
    argv = sys.argv[1:]
    main(*argv)