GitHub - okuvshynov/slowllama: Finetune llama2-70b and codellama on MacBook Air without quantization
https://github.com/okuvshynov/slowllama
46800306