import onnxruntime print(onnxruntime.__version__) print(onnxruntime.get_device() ) # 如果得到的输出结果是GPU，所以按理说是找到了GPU的 ort_session = onnxruntime.InferenceSession("fly.onnx", providers=['CUDAExecutionProvider']) print(ort_session.get_providers())
