Now I know what I’m watching this evening

😍

me: I'm blending a model called "capybara" with a "smaug" model. What would be a nice name for a new model? ChatGPT: Dragonbara! me: ok, now make a cool, detailed, fantasy illustration of Dragonbara. ChatGPT:

Too many cool projects, not enough hours in a day

I love this autogen usecase! I’m curious to see how good the results might be.

Maya not only has awesome technical content, but her delivery is on point Entertaining and informative, thanks Maya Akim !

For $5 and 20min I made a model better than Mixtral. my model scored way better on arc, hellaswag, truthfulqa, winogrande and gsm8k. all thanks to model blending. I'm sharing everything I learned in the video youtu.be/byf-y0P4hMg

I decided to cancel my chatGPT subscription. I was using gpt4 mostly for simple tasks like "check grammar" and "summarize this paper". But in the last few weeks I realized that mistral's le chat and claude do a much better job. I never used custom gpts or plugins.

For those who are keen on understanding LLMs and doing fun experiments with them- I highly recommend Maya Akim's channel. Its definitely not one of the opinionated tech-bros channel, rather every video is a hands-on experiment and an unbiased real-world review !! Apart from

This looks like a really cool way to evaluate code generation capabilities of LLMs

After listening to Sam Altman for 2h, the only truly new piece of information is that Ilya is really silly around puppies.

Today we're excited to introduce Multi-Model Sessions! — Load & prompt multiple local LLMs simultaneously — Use the API to build networks, pipelines of LLMs that can talk with each other — Use the new JSON Mode to constrain each model's output format LM Studio 0.2.17 is

Thanks David Ondrej for having me on the pod! David and I summarized months and months of our experiences with AI agents and why we think they are a big deal. youtu.be/fsIipBuM4Nc?si…

If you didn’t follow all, the situation has dramatically changed on the arena of LLMs recently: - Anthropic’s Claude 3 opus is now the undefeated winner of all closed-source models (just look at this win-rate line!) - cohere Command R+ is the new super strong leader of

The Small Language Model (SLM) Show is giving you another full episode devoted to the Magic of Model Merging. Join hosts Mark McQuade & @marymaccarthy, MergeKit founder Charles Goddard, & special guest Maya Akim next Weds 11am PT/2pm EST: youtube.com/watch?v=6IWPG1… #nlp #LLM #LLMs

Today's open source (or open-weight to be more accurate) models are better than closed source models from 1 year ago.

"Many people had a "ChatGPT moment" shortly after ChatGPT was released, when they ... were surprised that it significantly exceeded their expectation of what AI can do. If you have not yet had a similar "AI Agentic moment," I hope you will soon. I had one several months ago..."