Finetuning 3-Bit LLMs on Consumer GPUs by Integrating with Modular Quantizersarxiv.org2 points·volodia··0 commentsOpen articleSaveView on HN