SlowLlama: Finetune llama2-70B and codellama on MacBook Air without quantizationgithub.com/okuvshynov156 pointsbehnamoh3 years ago