llama.cpp now supports various small OCR models that can run on low-end devices. These models are small enough to run on GPU with 4GB VRAM, and some of them can even run on CPU with decent performance.
In this post, I will show you how to use these OCR models with llama.cpp 👇
もっと見る
