Inference Modes

Running inference with trained models.

Command Line

rtnn --inference True --load_model True --load_checkpoint_name model.pth

Python API

from rtnn import load_model, DataPreprocessor

# Load trained model
model = load_model(args)
checkpoint = torch.load('model.pth')
model.load_state_dict(checkpoint['state_dict'])

# Run inference
model.eval()
with torch.no_grad():
    predictions = model(input_data)