Doesn't support quantized model via bitsandbytes (8bit, 4bit)

[Dianagle2]

Bug

ValueError: `.to` is not supported for `4-bit` or `8-bit` bitsandbytes models. Please use the model as it is, since the model has already been set to the correct devices and casted to the correct `dtype`.

To Reproduce
Quantize a HuggingFace model with bitsandbytes (for example 8-bit).
Draw_graph has:

ValueError                                Traceback (most recent call last)
File /usr/local/lib/python3.8/dist-packages/torchview/torchview.py:256, in forward_prop(model, x, device, model_graph, mode, **kwargs)
    255 if isinstance(x, (list, tuple)):
--> 256     _ = model.to(device)(*x, **kwargs)
    257     # _ = model

.to is not supported for this kind of model.

Expected behavior
To be able to draw the quantized graph.
Another point: I can successfully generate the graph using torchviz's make_dot.
Both repositories produce Graphviz representation of PyTorch autograd graph, so right now I use make_dot for quantized models.