Absolutely love this project for keeping my OnePlus 6T alive.
Such sad state of affairs for Android. They dropped the ball on making any working edge deep learning inference framework. iOS is way better at this of all things. For being an OSS platform the amount of rigidity in not letting users customize to the fullest without rooting is just tragic.
But there arnt any official new builds for OnePlus 6t same with my OnePlus 6 due to the Strict eBPF & Kernel Requirements
Are you running unofficial builds right now ?
Doesn't OnePlus 6/6T have good "close to mainline" kernel support already? Why wouldn't it work within LineageOS eBPF/version requirements?
>Doesn't OnePlus 6/6T have good "close to mainline" kernel support already?
Source? Lineageos lists kernel version as 4.9, which definitely isn't "mainline".
https://wiki.lineageos.org/devices/enchilada/
SDM845 has increasing support in mainline Linux, but the current official LineageOS builds still rely on the vendor-provided kernel, which is why that is still listed on the wiki.
Untill 22(Android 15) yeah but compared to the manufacturer this is insane.
What even is there in Android 17 to talk about, same old UI, no non-google AI features to run on-device without root
AICore? Not sure what you mean if not this.
https://developers.google.com/ml-kit/custom-models
Just see out of touch with reality this section is compared to the insanse community work on Apple Silicon across whisper.cpp/mflux/llama.cpp/MLX/Exo & way more
What about it? Google's investment in Tensorflow Lite has paid off a lot better than MLX or CoreML did.
Android phones didn't need "insane community work" because (unlike Apple hardware) they support the same Vulkan compute shaders that PC GPUs use.