top | item 45472820 Sinkhorn: Make LLMs even smaller through quantisation while maintaining accuracy 4 points| ilitirit | 4 months ago |github.com 1 comment order hn newest albertwang|4 months ago Media commentary: https://venturebeat.com/ai/huaweis-new-open-source-technique...
albertwang|4 months ago Media commentary: https://venturebeat.com/ai/huaweis-new-open-source-technique...
albertwang|4 months ago