Quarterturn

Confirmed on OSX Safari. Entering an SSID and password then configure button leads to a connection reset on /configure. Refreshing the page returns "Parameters not found" Checking the ESP8266 filesystem...

I tried adjusting the GPU memory split to 15/16, but it threw an OOM exception even faster. ``` torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 228.00 MiB (GPU 0;...

Works on linux with CUDA 12.1: NVIDIA-SMI 530.30.02 Driver Version: 530.30.02 CUDA Version: 12.1

It compiles, finds my cuda devices, but is it using them? ``` ggml_init_cublas: found 2 CUDA devices: 1. Tesla P40 2. Tesla P40 llama.cpp: loading model from ./models/65B/ggml-model-q4_0.bin llama_model_load_internal: format...

> You need to set the --n-gpu-layers CLI argument to utilize the GPUs. Is there a way to know how many layers the model needs at a certain bit depth?...

That's coming from the Adafruit rgb-led-matrix library framebuffer.cc (https://github.com/adafruit/rpi-rgb-led-matrix/blob/master/lib/framebuffer.cc). Make sure you can run the demo python code for the library first.