3 ms·LoftQ = Quantization optimized for LoRA + Better LoRA Adaptor initialization + LoRA fine-tuning.by tourzhao 3y agoLoftQ = Quantization optimized for LoRA + Better LoRA Adaptor initialization + LoRA fine-tuning.