‘May the Source Be With You!’ – BigCode’s Open-Access StarCoder Outperforms All Existing Open Code LLMs
RESEARCH
-
Future of Data-Centric AI Conference June 7-8
By
–
The Future of Data-Centric AI Conf. is 3 weeks away on June 7-8, where @dpatil
, @matei_zaharia
, @yshoham
, @ajratner
, @lschmidt3 & other industry experts will share their insights. 1000s of data scientists and ML engineers from around the globe have already signed up. Get your -
RedPajama Reproduces LLaMA 1.2 Trillion Token Dataset
By
–
RedPajama, a project to create leading open-source models, starts by reproducing LLaMA training dataset of over 1.2 trillion tokens — TOGETHER https://
bit.ly/41PJIgE #AI #DeepLearning #MachienLearning #DataScience #GenAI -
Self-driving rules and machine learning migration challenges
By
–
Queue self-driving joke re: rules—>ML migration.
-
Symbol Tuning Improves In-Context Learning Research
By
–
Once again, another piece of great science from @JerryWeiAI that shows that symbol tuning improves in-context learning. Check out his thread below
-
Google and OpenAI Lack Transparency on AI Model Disclosures
By
–
Lack of disclosure, obfuscation begin to become the norm in artificial intelligence. Google follows OpenAI in telling the world almost nothing about its new PaLM 2 AI program #AI #Google #PaLM2 $GOOGL @OpenAI
-
OpenAI Plugins Undermine LLM General Purpose Design Philosophy
By
–
Thus spake @benedictevans on OpenAI's plug-in play (see https://
help.openai.com/en/articles/68
25453-chatgpt-release-notes
…): "If the whole point of an LLM is that it’s general purpose, why does it need plugins? Doesn’t that make it an NLP front-end to 1990s SQL databases?" -
Comprehensive Practical Guide to Large Language Models Resources
By
–
A well-curated list of practical resources of large language models(LLMs). Contains LLMs evolutionary tree, example models, and papers. This is a handy resource for practitioners that want to dive deep into LLMs. Repo: https://
github.com/Mooler0410/LLM
sPracticalGuide
…
Related paper: -
LLMU Twitter Spaces: Comprehensive LLM Curriculum Overview
By
–
Following the launch of LLMU, @luis_likes_math and @jayalammar will be hosting Twitter Spaces next Tuesday. Join them as they provide a comprehensive overview of the curriculum, LLMs and what’s on the horizon! Save the date! https://
x.com/i/spaces/1ZkKz
XrLORwJv
… -
AGI Claims Reputation Risk Among Computer Scientists
By
–
“Making A.G.I. claims can be a reputation killer for computer scientists. What one researcher believes is a sign of intelligence can easily be explained away by another, and the debate often sounds more appropriate to a philosophy club than a computer lab.”