Open davidw0311 opened 1 year ago
This is related to #242 as cpuAndNeuralEngine
is meant to be used with a SPLIT_EINSUM
(or SPLIT_EINSUM_V2
) model whereas the model you linked to is an ORIGINAL
model which is meant for cpuAndGPU
. We will share some instructions on this soon.
I am trying to deploy a core ml model to my iPhone 13 Pro Max (iOS 17 beta)
Using the checkpoints from https://huggingface.co/apple/coreml-stable-diffusion-mixed-bit-palettization/tree/main/coreml-stable-diffusion-xl-base_mbp_4_50_palettized.
My swift initialization of the model is as follows:
The model loading runs for around 9 mins before crashing with the error:
Doing monitoring with xcode, it appears that memory spikes right before the crash and the text_encoder_2 is not finished loading:
Appreciate any suggestions on how to deply the sd-xl model successfully on iPhone.