Exciting news! Today we are releasing Cerebras-GPT, a family of 7 GPT models from 111M to 13B parameters trained using the Chinchilla formula. These are the highest accuracy models for a compute budget and are available today open-source! (1/5) Press:
LLMS
-
CEO discusses Dolly open-source LLM on Bloomberg
By
–
Check out our CEO @alighodsi on @BloombergTV chatting all things Dolly! Read the story behind establishing our own #opensource LLM — and how models like Dolly will make #LLMs more accessible to all companies https://
bloom.bg/40oxXgS -
Expertise in AI chatbots and coding proficiency discussion
By
–
oui il parle de sujet qu'il maîtrise pas, moi ça fait 2 ans que j'utilises GPT3, et y'a beaucoup plus expert que moi. Et j'ai commencé à m'intéresser aux chatbot y'a plus de 15 ans. Pareil niveau code, je suis nul, et je l'assume, hier il parlait de 90% d'informaticiens, bah si
-
GPT-4 productivity gap between expert and mediocre programmers
By
–
Les réactions à ce tweet sont instructives Les excellents informaticiens par exemple les Polytechniciens se servent tres bien de GPT4 qui explose leur productivité Les médiocres expliquent que GPT4 est nul et ne sert à rien On voit la fracture entre ceux qui sont tres
-
AI Models Now Outperforming Average Human Performance
By
–
We are at the point now that model outputs are better than the average human being huh.
-
Microsoft Identifies Sparks of AGI in GPT-4 Deep Dive
By
–
Microsoft Does a Deep Dive on GPT-4, Finds “Sparks of AGI” https://
syncedreview.com/2023/03/27/mic
rosoft-does-a-deep-dive-on-gpt-4-finds-sparks-of-agi/
… -
Open-Source Decentralized LLMs vs Closed-Source Competition
By
–
The biggest near-term question in AI: Will open-source LLMs with decentralized training be competitive with closed-source and centralized LLMs?
-
The Learnability of In-Context Learning: Theoretical Framework
By
–
The Learnability of In-Context Learning In context-learning is a term that refers to the emerging ability of very large models to perform tasks with few or zero demonstrations. This paper attempts to give a theory behind in-context learning. https://
arxiv.org/abs/2303.07895 -
Multilingual Task-Oriented Dialogue Dataset Released
By
–
Introducing a multilingual dataset for parsing realistic task-oriented dialogues, which includes about half a million realistic conversations in six languages, many of which exhibit common phenomena like disfluencies, user revisions and code-mixing. → https://t.co/Jum6oI8sKR pic.twitter.com/DtHALJRIfL
— Google AI (@GoogleAI) 27 mars 2023Introducing a multilingual dataset for parsing realistic task-oriented dialogues, which includes about half a million realistic conversations in six languages, many of which exhibit common phenomena like disfluencies, user revisions and code-mixing. → https://
goo.gle/3za7aZx -
Waiting for API 4 Access to Upgrade Systems
By
–
I don't have access to the 4 API yet, but will upgrade when I get it