Onnx to Hef conversion problem for consultation

General
1

Hello experts
There is the following code in my model, but we do not support grid_sample.

features = []
        for fm in feature_maps:
            print("111111111111",fm.shape,points_2d.shape)
            features.append(
                torch.nn.functional.grid_sample(
                    fm.flatten(end_dim=1), points_2d
                )
            )
            print("222222222222222",torch.nn.functional.grid_sample(
                    fm.flatten(end_dim=1), points_2d
                ).shape)

111111111111 torch.Size([1, 6, 256, 64, 176]) torch.Size([6, 900, 13, 2])
222222222222222 torch.Size([6, 256, 900, 13])

This question has been bothering me for a long time.Do you have any suggestions regarding the replacement of grid_sample?Looking forward to your reply.