๊ฐ€์ž… ํ›„ ์ดˆ๋Œ€ ๋งํฌ๋ฅผ ๊ณต์œ ํ•˜๋ฉด ๋™์˜์ƒ ์žฌ์ƒ ๋ฐ ์ดˆ๋Œ€ ๋ณด์ƒ์„ ๋ฐ›์„ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

Adina Yakup
@AdinaYakup
@huggingface ๐Ÿค— | opinions are my own
๊ฐ€์ž… April 2023
905 ํŒ”๋กœ์ž‰ ์ค‘    14.3K ํŒฌ
Ant group just dropped Ring-2.6-1T ๐Ÿ”ฅ 1T reasoning model, built for real world agent workflows. โœจ MIT license โœจ 128K >> 256K context (YaRN) โœจ Async RL + IcePop training architecture โœจ Dual reasoning effort: "high" for fast agent loops, "xhigh" for deep reasoning = Better cost/performance tradeoff ๐Ÿ‘€
๋” ๋ณด๊ธฐ