Logo: 
Areas Served: 
Meet MPT-7B: A Suite of Open Source, Commercially Available LLMs that Supports 65k Tokens
Author(s): Jesus Rodriguez Originally published on Towards AI. The new suite of models was released by MosaicML and support models optimized for Instructions, Chats, Stories and More. Image Credit: MosaicML I recently started an AI-focused educational newsletter, that already has over 150,000 …
How OpenAI Uses GPT-4 to Interpret Neurons in LLMs
Author(s): Jesus Rodriguez Originally published on Towards AI. A new interpretability method based on GPT-4 can derive explanations about specific neurons in LLMs. Created Using Midjourney I recently started an AI-focused educational newsletter, that already has over 150,000 subscribers. TheSequence is a …
Inside Lamini: A New Framework for Fine-Tuning LLMs
Author(s): Jesus Rodriguez Originally published on Towards AI. The framework streamlines the process of using techniques such as RLHF in your LLM models. Top highlight Image Credit: Lamini I recently started an AI-focused educational newsletter, that already has over 150,000 subscribers. TheSequence …
Inside Low-Code LLM: Microsoft Research’s Novel Prompt Engineering Method Based on Human-LLM Interactions
Author(s): Jesus Rodriguez Originally published on Towards AI. Created Using Midjourney I recently started an AI-focused educational newsletter, that already has over 150,000 subscribers. TheSequence is a no-BS (meaning no hype, no news etc) ML-oriented newsletter that takes 5 minutes to read. …
Meet DeepSpeed-Chat: Microsoft’s New Framework to Create ChatGPT-Like Models Using RLHF Training
Author(s): Jesus Rodriguez Originally published on Towards AI. Image Credit: Microsoft Research I recently started an AI-focused educational newsletter, that already has over 150,000 subscribers. TheSequence is a no-BS (meaning no hype, no news etc) ML-oriented newsletter that takes 5 minutes to …
Meet Koala: Berkeley University’s LLaMA-Based Model Fine-Tuned with ChatGPT Dialogues
Author(s): Jesus Rodriguez Originally published on Towards AI. Created Using Midjourney I recently started an AI-focused educational newsletter, that already has over 150,000 subscribers. TheSequence is a no-BS (meaning no hype, no news etc) ML-oriented newsletter that takes 5 minutes to read. …
Meet Dolly: How Databricks Finetuned a Two-Year-Old LLM to Follow Instructions like ChatGPT
Author(s): Jesus Rodriguez Originally published on Towards AI. Created Using Midjourney I recently started an AI-focused educational newsletter, that already has over 150,000 subscribers. TheSequence is a no-BS (meaning no hype, no news etc) ML-oriented newsletter that takes 5 minutes to read. …
The Crown Jewel Behind ChatGPT: Reinforcement Learning with Human Feedback
Author(s): Jesus Rodriguez Originally published on Towards AI. One of the core ideas behind ChatGPT dates back to a research paper from 2017. Created with Stable Diffusion I recently started an AI-focused educational newsletter, that already has over 150,000 subscribers. TheSequence …
Meet BLIP-2: Salesforce New Open Source Visual-Language Model that is Faster the Simpler than GPT-4
Author(s): Jesus Rodriguez Originally published on Towards AI. The model radically improves in the cost and efficiency of pretraining for visual-language models. Created Using Midjourney I recently started an AI-focused educational newsletter, that already has over 150,000 subscribers. TheSequence is a …
Meet MM-REACT: Microsoft Research New Model that Enables Visual Reasoning on top of ChatGPT
Author(s): Jesus Rodriguez Originally published on Towards AI. The model combines language and computer vision to enable sophisticated reasoning capabilities. Image Credit: Microsoft Research I recently started an AI-focused educational newsletter, that already has over 150,000 subscribers. TheSequence is a no-BS …
Microsofts MathPrompter Shows How to Use Mathematical Reasoning with Large Language Models
Author(s): Jesus Rodriguez Originally published on Towards AI. Microsoft’s MathPrompter Shows How to Use Mathematical Reasoning with Large Language ModelsThe model uses a four-step process to improve trust and reasoning in mathematical problems. I recently started an AI-focused educational newsletter, that already …
OpenChatKit is an Open Source Alternative to ChatGPT
Author(s): Jesus Rodriguez Originally published on Towards AI. The framework was created by the collaboration of Togeter, LAION, and Ontocord.Created Using Midjourney I recently started an AI-focused educational newsletter, that already has over 150,000 subscribers. TheSequence is a no-BS (meaning no hype, …
Popular posts
Updates
Recent Posts
