Shane John Paul

So, I went through the source for both ConvNeXt and RegNet. There were two separate checks for the `input_tensor` from this [line](https://github.com/keras-team/keras/blob/b80dd12da9c0bc3f569eca3455e77762cf2ee8ef/keras/applications/convnext.py#L441-L452). ``` if input_tensor is None: img_input = layers.Input(shape=input_shape)...

Okay.

> https://dl.sipeed.com/MAIX/MaixPy/release/master/maixpy_v0.5.0_124_g92d9c04 > >  This actually worked!!! (the one with model v4 support) But now I can only load my model. When I try to set the frame size...

Thanks. I'll try them.

Does the online compiler work? I chose the kmodel v4 option in it and it wasn't able to load the model.

I compiled the firmware as mentioned in [here](https://github.com/sipeed/MaixPy/blob/master/build.md). I went with the maixpy_k210_minimum project. I set all the options as mentioned above. But still, it's running out of memory when...

I'm facing the same issue. Any updates on this?

There's a typo in get_nncase.sh Right now get_nncase.sh is this, ``` #!/bin/bash mkdir -p ncc mkdir -p workspace mkdir -p images mkdir -p log cd ncc wget https://github.com/kendryte/nncase/releases/download/v0.1.0-rc5/ncc-linux-x86_64.tar.xz tar -Jxf...