Virtual Diana (VNA) format conversion of llama2.c TinyStories models.
This version uses INT8 quantization, but will be expanded with other quantization flavors soon.
It's intended to be used with EMMA
Virtual Diana (VNA) format conversion of llama2.c TinyStories models.
This version uses INT8 quantization, but will be expanded with other quantization flavors soon.
It's intended to be used with EMMA