CODE - AI Dynamics

The Transformer Blueprint: A Holistic Guide to Neural Networks

By

–

29 July 2023 17h39

A NEW ARTICLE The first article of Deep Learning Revision Research Blog(introduced recently) is out: "The Transformer Blueprint: A Holistic Guide to the Transformer Neural Network Architecture". In the article, we discuss the core mechanics of transformer neural

→ View original post on X — @jeande_d

29 July 2023

Using PyTorch Vision ViT as Baseline for Model Training

By

@rasbt

–

29 July 2023 14h03

No that I am aware (but I also haven't looked too deeply). I would probably use the original ViT in Torchvision as a baseline: https://
pytorch.org/vision/main/_m
odules/torchvision/models/vision_transformer.html#vit_b_16
… I have a companion article here showing how to accelerate the training on that model:

→ View original post on X — @rasbt

29 July 2023

Open-Source Model Fine-tuning Methods and Training Flexibility

By

@rasbt

–

29 July 2023 14h01

Yes, any finetuning method is allowed, afaik. Technically, you could also just further (pre)train the model. Anything that runs <24 h on open source code and openly available data.

→ View original post on X — @rasbt

29 July 2023

10x Engineers Must Embrace AI Tools to Stay Relevant

By

@alexandr_wang

–

29 July 2023 0h07

weird effect of AI developer tools— the 10x engineers of today will NOT be the 10x engineers of tomorrow because they’ll be too proud to adopt the new AI tools and workflows, which are meaningfully different. the solution—have no ego, let the AI tell you what to do.

→ View original post on X — @alexandr_wang

29 July 2023

OpenAI Extends Deprecation Window Amid Developer Backlash

By

@mattlynley

–

28 July 2023 19h55

We also saw OpenAI capitulating to developers by extending a deprecation window to a full year (up from 3mo) due to "feedback" (aka backlash) from developers. It also saw the rise of a narrative of GPT-4's performance dropping, which was a bit of a misinterpretation of a paper.

→ View original post on X — @mattlynley

28 July 2023

LangSmith Updates and Open Source Recap Released

By

@hwchase17

–

28 July 2023 18h15

First edition out LangSmith updates, Open Source recap, Use-cases we love… and some sneak previews of things on the horizon Concise, high density. s/o @zebriez

→ View original post on X — @hwchase17

28 July 2023

Llama 2 Coding Limitations Impact on Agentic AI Tasks

By

@hwchase17

–

28 July 2023 17h30

Had an awesome webinar with @RLanceMartin and @jamescalam on Weds around using Llama 2 Llama2 is significantly worse on coding tasks – is that why it's not nearly as good at agentic tasks as GPT models? Up on YouTube now:

→ View original post on X — @hwchase17

28 July 2023