Top Tweets for #LongLLaMA
#NeurIPS2023 is over. 🔚
"Our papers gained a lot of attention, opening possibilities for cooperation with other research labs," says @CStanKonrad, PhD student at IDEAS NCBR and @UniWarszawski, co-author of #LongLLaMA paper
💪 @OlkoMateusz @danie1marczak @pszwnzl @NeurIPSConf
🦙Let's gather for a #LongLLaMA session!
🗓️On Dec. 21, @CStanKonrad, co-author of the "Focused Transfomer" paper, will give online talk about LongLLaMA. The event is a collaboration of @gathers_pc and IDEAS NCBR. Learn how to join 👉
https://t.co/swBgCnJ68E
🎇Introducing LongLLaMA-Code 7B Instruct 🦙!🎇
A step towards an open-source alternative for Claude 2.
Run in 🆓 Colab (8bit).
🗨 Answers questions about 📑 papers and >_ code.
SOTA 7B reasoning :
🎓 GSM8K: 65% 🐍 PoT 0-shot, 42% std CoT 8-shot setting.
>_ 37%: HumanEval
#LongLLaMA users can now chat with the model after two clicks.
🦙Follow the link➡️ https://t.co/a1OhRPjDcZ
Thanks to @PiotrRMilos and team 🆙 #LLMs
🦙Polscy badacze z @UniWarszawski, @PAN_akademia i @IDEAS_NCBR opracowali duży model językowy #LongLLaMA, który potencjalnie pozwoli obsługiwać 64 razy więcej tekstu niż #ChatGPT.
Czytaj więcej➡️https://t.co/G3li15ewVD
i przetestuj możliwości #OpenLLaMA➡ https://t.co/b5fZBJfG3E
#LongLLaMA w przeciwieństwie do #ChatGPT nie posiada interfejsu w internecie - każdy może pobrać model ze strony HuggingFace. Model pozwoli obsługiwać 64 razy więcej tekstu niż ChatGPT - stwierdzają jego twórcy @IDEAS_NCBR @UniWarszawski @PAN_akademia
https://t.co/S3ObLbDPXU
Polscy naukowcy z @UniWarszawski, @PAN_akademia i @IDEAS_NCBR opracowali duży model językowy #LongLLaMA, który potencjalnie pozwoli obsługiwać 64 razy więcej tekstu niż #ChatGPT. Każdy może go pobrać i uruchomić na własnym komputerze.
https://t.co/Eix2ZTmJlw
🦙 And the Focused Transformer paper, which allowed to produce #LongLLaMA, has just been accepted to #NeurIPS2023!
Great news @s_tworkowski @CStanKonrad @MikolajPacek @Yuhu_ai_ @hmichalewski @PiotrRMilos @NeurIPSConf
✨Announcing LongLLaMA-Code 7B!✨
Have you wondered how GPT3.5 obtained its capability?
Are base models of code better reasoners? 🤔
We continue pre-training CodeLLaMA on text & code to improve reasoning 🧠
Bonus: 3x faster inference @ 16K context, using Focused Transformer 🎯
🦙 Behold the majestic #LongLLaMA, gracefully grazing across the lush meadows of #MachineLearning
👏 @s_tworkowski @PiotrRMilos @MikolajPacek @CStanKonrad @Yuhu_ai_ @hmichalewski
Top ML Papers of the Week (July 10 - July 16):
- CM3Leon
- LongLLaMA
- AnimateDiff
- Patch n’ Pack: NaViT
- Secrets of RLHF in LLMs
- LLMs as General Pattern Machines
...
📢 #LONGLLAMA 🔥 Focused Transformer (FOT): a technique that employs a training process inspired by #contrastive learning.
💡 This novel approach enhances the structure of the (key, value) space, enabling an extension of the context length.
https://t.co/rfS49sIY4k
#ai #llms
4/6 For a quick read on this breakthrough, check out this article: https://t.co/S4bczHSmzU It provides a concise overview of LongLLaMA and its capabilities. #AI #LongLLaMA
1/6 🚀💡 Exciting breakthrough in #AI: Meet LongLLaMA, a Large Language Model that can handle an astounding 256k tokens! This is a game-changer in the field of AI, opening up new possibilities for complex tasks. #LongLLaMA #ArtificialIntelligence @s_tworkowski
🔔Great news! Research team of @PiotrRMilos, @s_tworkowski, @CStanKonrad, @MikolajPacek (IDEAS NCBR) and @Yuhu_ai_, @hmichalewski from @GoogleAI released research preview of #LongLLaMA, a special unlimited-context version of OpenLLaMA!
👏Congratulations!
More in Szymon's thread⤵️
Introducing LongLLaMA 🦙, an unlimited-context version of OpenLLaMA fine-tuned at 8k & capable of extrapolating to 256k tokens!
We train it using our new Focused Transformer 🎯 technique (FoT). No degradation on short context, drop-in compatibility & Apache 2.0 license 🔥🔥
🧵
Last Seen Hashtags on Sotwe
Feylom
Seen from United States
ayeshagilmana
Seen from United States
ノガレ120F
Seen from United States
เอวีญี่ปุ่น
Seen from Thailand
EightyYearsStrong
Seen from United States
PulangPisau
Seen from United States
ميقا_حصريات
Seen from Saudi Arabia
premierepro
Seen from Argentina
ハナコ
Seen from United States
YGIGHeartstruck
Seen from United States
Most Popular Users
Elon Musk 
@elonmusk
241.5M followers
Barack Obama
@barackobama
119.1M followers
Cristiano Ronaldo
@cristiano
113.5M followers
Donald J. Trump
@realdonaldtrump
111.8M followers
Narendra Modi
@narendramodi
107.2M followers
Rihanna
@rihanna
98.5M followers
NASA
@nasa
92.3M followers
Justin Bieber
@justinbieber
91.7M followers
KATY PERRY
@katyperry
89.5M followers
Taylor Swift
@taylorswift13
83.4M followers
Lady Gaga
@ladygaga
74.9M followers
Virat Kohli
@imvkohli
72.5M followers
Kim Kardashian
@kimkardashian
70.7M followers
YouTube
@youtube
68.8M followers
Neymar Jr
@neymarjr
65.5M followers
Bill Gates
@billgates
64.8M followers
Selena Gomez
@selenagomez
62.5M followers
The Ellen Show
@theellenshow
62.3M followers
CNN
@cnn
61.8M followers
X
@x
60.8M followers