Can models be exported as code? [ANSWERED]

[TheSuperiorTwin]

Can models be exported as lines of code?

[glenn-jocher]

@TheSuperiorTwin not sure what you mean. Every model is composed of code. All supported export formats are indicated here:

Formats

YOLOv5 inference is officially supported in 11 formats:

Format	export.py --include	Model
PyTorch	-	yolov5s.pt
TorchScript	torchscript	yolov5s.torchscript
ONNX	onnx	yolov5s.onnx
OpenVINO	openvino	yolov5s_openvino_model/
TensorRT	engine	yolov5s.engine
CoreML	coreml	yolov5s.mlmodel
TensorFlow SavedModel	saved_model	yolov5s_saved_model/
TensorFlow GraphDef	pb	yolov5s.pb
TensorFlow Lite	tflite	yolov5s.tflite
TensorFlow Edge TPU	edgetpu	yolov5s_edgetpu.tflite
TensorFlow.js	tfjs	yolov5s_web_model/

You can visualize all exports with Netron, load them in their respective environments for introspection etc. For example if you download your trained PyTorch model:

model = torch.load('downloaded_model.pt')
print(model['model'])

[TheSuperiorTwin]

This is what I meant, sorry for the vague wording. Thank you!

[glenn-jocher]

Great! Let us know if you have any other questions or ideas for improving HUB :)