How many bits we should use for quantization for finetuning this model? 4, 8 or 16?
· Sign up or log in to comment
