Jagrit Digani

Thinking about possible methods, an option is to just bundle the metal source code and have the metal framework compile everything on the fly from the source - this might...

Thanks for flagging this! I'll get working on a fix right away

@awni I agree about raising exceptions from metal kernels My thinking was we let the underlying gather/take metal kernel remain the same, but instead do a check in the `__get_item__`...

I just ran the examples on a M2 Ultra, and can confirm that convolution isn't the major culprit - as @SarthakYadav said, switching it out with a reshape + linear...

Can you tell me what the results are when you run `xcode-select -p` ? Looking at the path `/Library/Developer/CommandLineTools`, it looks like you might not have Xcode installed or maybe...

Would you mind adding the following to the CMakeLists.txt in root (around line 79) and building again ? ``` message(STATUS "Building for processor ${CMAKE_SYSTEM_PROCESSOR}") add_compile_options(-mfloat-abi=softfp -mfpu=neon) ``` What's happening is...

> I tried this with my M2 chip and it still doesn't compile for me. Just to be sure, did the CMake build messages say “Building for arm” ? it...

> I run into this issue, just trying it out on an M3 Pro device. By prepending the processor setting, I get the build to succeed on my machine. The...

I'll admit I don't have any experience using MacOS VMs on Parallels - do you have access to AppleSilicon GPUs on your VM ? Running the command `system_profiler SPDisplaysDataType` might...

Oh, if llama.cpp works, then it must just be a missing ArgumentEncoder function for virtual devices down a Metal Framework level rather than something bigger like I was worried about...