Teven Le Scao (@fluke_ellington) 's Twitter Profile
Teven Le Scao

@fluke_ellington

Researcher @MistralAI, producer @ my bedroom, no BLOOM slander authorized on this account

ID: 1225131780531019778

calendar_today05-02-2020 18:59:45

396 Tweet

2,2K Followers

605 Following

lmarena.ai (formerly lmsys.org) (@lmarena_ai) 's Twitter Profile Photo

[Arena] Exciting update! Mistral Medium has gathered 6000+ votes and is showing remarkable performance, reaching the level of Claude. Congrats Mistral AI! We have also revamped our leaderboard with more Arena stats (votes, CI). Let us know any thoughts :) Leaderboard

[Arena] Exciting update! Mistral Medium has gathered 6000+ votes and is showing remarkable performance, reaching the level of Claude. Congrats <a href="/MistralAI/">Mistral AI</a>!

We have also revamped our leaderboard with more Arena stats (votes, CI). Let us know any thoughts :)

Leaderboard
Leo Tronchon (@leotronchon) 's Twitter Profile Photo

Today we release Idefics2 our newest 8B Vison-Language Model! 💪 With only 8B parameters Idefics is one of the strongest open models out there 📋 We used multiple OCR datasets, including PDFA and IDL from Ross Wightman and Pablo Montalvo, and increased resolution up to 980x980 to improve

Today we release Idefics2 our newest 8B Vison-Language Model!
💪 With only 8B parameters Idefics is one of the strongest open models out there
📋 We used multiple OCR datasets, including PDFA and IDL from <a href="/wightmanr/">Ross Wightman</a> and <a href="/m_olbap/">Pablo Montalvo</a>, and increased resolution up to 980x980 to improve
Albert Jiang (@albertqjiang) 's Twitter Profile Photo

Nice paper! Some surprising highlights: 1. Mixtral 8x22B is ~GPT4-turbo level on GSM8K and GSM1K. Mistral large is better on both. 2. On GSM1K, Mixtral-8x22B-Instruct (84.3%) > claude-2 (83.6%) >> claude-3-haiku (79.1%) >> claude-3-sonnet (72.4%) 🤔 Also worth highlighting how

Nice paper! Some surprising highlights:
1. Mixtral 8x22B is ~GPT4-turbo level on GSM8K and GSM1K. Mistral large is better on both.
2. On GSM1K, Mixtral-8x22B-Instruct (84.3%) &gt; claude-2 (83.6%) &gt;&gt; claude-3-haiku (79.1%) &gt;&gt; claude-3-sonnet (72.4%) 🤔
Also worth highlighting how
Teven Le Scao (@fluke_ellington) 's Twitter Profile Photo

long overdue getting into shaders, project this on the wall at your next party shadertoy.com/view/4flcDN it's so cool that llms know every language ever, you can just ask mistral-large to hold your hand for the first day you're learning them and go in medias res

long overdue getting into shaders, project this on the wall at your next party shadertoy.com/view/4flcDN

it's so cool that llms know every language ever, you can just ask mistral-large to hold your hand for the first day you're learning them and go in medias res
Mistral AI (@mistralai) 's Twitter Profile Photo

magnet:?xt=urn:btih:7278e625de2b1da598b23954c13933047126238a&dn=pixtral-12b-240910&tr=udp%3A%2F%2Ftracker.opentrackr.org%3A1337%2Fannounce&tr=udp%3A%2F%https://t.co/2UepcMHjvL%3A1337%2Fannounce&tr=http%3A%2F%https://t.co/NsTRgy7h8S%3A80%2Fannounce

Teven Le Scao (@fluke_ellington) 's Twitter Profile Photo

Super happy to release what I've been working on for the last few months :D you wouldn't believe how much intelligence you can fit in your dad's USB key

Teven Le Scao (@fluke_ellington) 's Twitter Profile Photo

mistral.ai/news/magistral super happy to release what I've been working on for the last few months. The model is on Le Chat, we find it super fun to use and we hope you will too. This is just the start, come build the next chapters with us!