INT4 LoRA fantastic-tuning vs QLoRA: A user inquired about the differences between INT4 LoRA good-tuning and QLoRA in terms of accuracy and speed. A different member explained that QLoRA with HQQ requires frozen quantized weights, will not use tinnygemm, and utilizes dequantizing alongside torch.matmulTweet from Robert Graham (@ErrataRob): nVidia i