onnx version of this model

[MadeByKit]

hello @1adrianb is there an onnx runtime or openvino version for this pytorch model?
https://www.adrianbulat.com/downloads/python-fan/s3fd-619a316812.pth

your response will be appreciated

[Chappie74]

face_alignment_net = torch.jit.load('2DFAN4-cd938726ad.zip')
face_alignment_net.eval()
img_tensor = torch.randn(1, 3, 256, 256)
if img_tensor.dtype != torch.float32:
    img_tensor = img_tensor.to(torch.float32)
torch.onnx.export(face_alignment_net, img_tensor,
                  'landmark.onnx',
                  opset_version=18,  # The ONNX version to export the model tohe model's input names
                  input_names=['input0'],  # The model's input names
                  output_names=['output0']  #
                  )

# Load the ONNX model
onnx_model = onnx.load('landmark.onnx')

# Check that the model is well-formed
onnx.checker.check_model(onnx_model)

try using this