The kernel appears to have died. It will restart automatically

Hi,
I have been working on PYNQ-Z1, and I am facing an issue of “The kernel appears to have died. It will restart automatically.” when running the following code:
model = load_model(model_backup_path)
model.predict(test_image)
where model_backup_path indicates the name of my " .h5" file/model.

Note: I have installed Tensorflow, Keras, and imported the package of load_model. It is okay to run “from keras.models import load_model”, however when calling the functions (of load_model and model.predict) the kernel dies.

Have you guys met the same issue? I would appreciate if you can help.

Regards
Chao

Looks like the model is too large to fit in memory. Some people came across this before.

I see, but I have no problem while evaluate on Z1 and Z2.
And basically Ultra96v2 is bigger DDR memory right?
So, any suggestion?

Hi, Radhen, Have you found the problem? I also encouter this problem now. My board is ZCU104 which is better than PYNQ-Z2

1 Like