Hi, I have converted a pytorch(1.5.0) pth model to onnx. When I put the output model into asset folder, I got this error. ArgumentException: Cannot reshape array of size 37632 into shape with multiple of 1024 elements Barracuda.TensorExtensions.Reshape (Barracuda.TensorShape shape, System.Int32 size) (at Library/PackageCachefirstname.lastname@example.org/Barracuda/Core/TensorExtensions.cs:310) Model : Yolact Resnet50 Onnx Version : 9 Batch size : 1 Input Shape(1,550,550,3) When I got this error, The Reshaping size : (256,1,4,-1) I saw a forum page dealing with simillar issues, the comment was to make batch size larger than sequence length. But in my case, input batch is 1(realtime image). So have anyone encountered simillar case like me? Or solve it?