Size of the output tensor not properly define

[leettatt]

Let's create a dummy input tensor

dummy_input = torch.randn(1, input_size, requires_grad=True)  

ABOVE CAUSE ERROR AS input_size not properly define

Manually changed to below to get it work

dummy_input = torch.randn(1, 3, 32, 32, requires_grad=True)

---

Document Details

⚠ Do not edit this section. It is required for docs.microsoft.com ➟ GitHub issue linking.

• ID: b236fa85-23be-f5ec-5c5f-332937187f04
• Version Independent ID: 781552b8-a796-a963-05d3-ed98f509dbd4
• Content: Convert your PyTorch model to ONNX
• Content Source: docs/windows-ml/tutorials/pytorch-convert-model.md
• Product: machine-learning
• GitHub Login: @QuinnRadich
• Microsoft Alias: quradic