LLMS - AI Dynamics

OpenAI Improves Math Reasoning with RLHF and Chain-of-Thought

By

–

01 June 2023 0h31

research from @OpenAI on improving math reasoning by RLHF with a reward model trained on 800k human-generated chain-of-thought data (which @scale_AI partnered w/
@OpenAI on!) RLHF seems to be a scalable technique for making LLMs smarter in many ways

→ View original post on X — @alexandr_wang

1 June 2023

LLMs Prompts Output Parsers and Evaluation Fundamentals

By

@hwchase17

–

31 May 2023 23h28

we covered a lot but i liked the prompts/llms/output parsers section (basic building blocks, important to understand) and the evaluation section (very practical regardless of use case)

→ View original post on X — @hwchase17

31 May 2023

Falcon-40b Language Capabilities and Supported Languages Overview

By

@thom_wolf

–

31 May 2023 21h16

It’s not trained on Arabic though. Covers English, German, Spanish, French (and limited capabilities in Italian, Portuguese, Polish, Dutch, Romanian, Czech, Swedish) from the model card https://
huggingface.co/tiiuae/falcon-
40b
…

→ View original post on X — @thom_wolf

31 May 2023

Falcon 40B Now Apache-2 Licensed Free Commercial Use

By

@thom_wolf

–

31 May 2023 21h10

The license of the Falcon 40B model has just been changed to… Apache-2 which means that this model is now free for any usage including commercial use (and same for the 7B)

→ View original post on X — @thom_wolf

31 May 2023

LangChain for Building LLM Applications: Memory, QA, Agents

By

@andrewyng

–

31 May 2023 21h03

3/LangChain for LLM Application Development: Learn this powerful open-source tool for building applications using LLMs, including memory for chatbots, Question answering over a doc, and an LLM agent that can decide what action to take next. https://
learn.deeplearning.ai/langchain

→ View original post on X — @andrewyng

31 May 2023

Building Complex Systems with ChatGPT API

By

@andrewyng

–

31 May 2023 21h03

2/Building Systems with the ChatGPT API: Go beyond individual prompts, and learn to build complex applications that use multiple API calls to an LLM. Also learn to evaluate an LLM's outputs for safety and accuracy, and drive iterative improvements. https://
learn.deeplearning.ai/chatgpt-buildi
ng-system
…

→ View original post on X — @andrewyng

31 May 2023

DeepLearning.AI Launches 3 New Generative AI Courses

By

@andrewyng

–

31 May 2023 21h02

1/Thrilled to announce: 3 new Generative AI courses!
* Building Systems with the ChatGPT API, with OpenAI’s @isafulf * LangChain for LLM Application Development, with LangChain’s @hwchase17 * How Diffusion Models Work, by @realSharonZhou Check them out: http://
deeplearning.ai/short-courses/

→ View original post on X — @andrewyng

31 May 2023

Wordcraft: AI Text Editor for Short Story Writing

By

@miketamir

–

31 May 2023 21h02

GitHub – PAIR-code/wordcraft: Wordcraft is an AI-powered text editor with an emphasis on short story writing https://github.com/PAIR-code/wordcraft?utm_source=substack&utm_medium=email #AI #MachineLearning #DeepLearning #LLMs #DataScience

→ View original post on X — @miketamir

31 May 2023

OpenAI Releases Process Supervision Method for Mathematical Reasoning

By

@_akhaliq

–

31 May 2023 20h52

Open AI releases paper + dataset Let’s Verify Step by Step trained a model to achieve a new state-of-the-art in mathematical problem solving by rewarding each correct step of reasoning (“process supervision”) instead of simply rewarding the correct final answer (“outcome

→ View original post on X — @_akhaliq

31 May 2023