How Minimax-01 Achieves 1M Token Context Length with Linear Attention (MIT)

2 points| research_pie | 11 months ago |yacinemahdid.com