Recent Advancements in Attention Mechanism for Long-Sequence Understanding and Generation

Survey of Current Modified Transformer Attention Designs

Some exciting details from Apple announcements at WWDC 2024

AI now stands for Apple Intelligence

AWS offers several options for LLM usage, which will you choose?

Picking the right LLM deployment configuration in AWS

LLMs promise high levels of performance without the need for task specific training, but that doesn't mean that custom datasets are unnecessary.

Is your data more or less important in the age of LLMs?

Last week a bedrock agent was deployed manually in AWS Console. Let's now do it the smart way, with Infrastructure as Code using AWS CDK

How to Deploy Bedrock Agents using AWS CDK

A few weeks ago, I used Google's Gemini LLM for function calling. let's see how it works with AWS Bedrock and Anthropic's Claude

AI Agents with AWS Bedrock (Claude-3 Haiku LLM)

When you need an LLM to access new information, LLM function calling is here to help. Let's explore how to use the Gemini function calling to access real information.

Function Calling with Google's Gemini LLM

Langchain seems to be a popular choice for developing LLM applications, but when extending it beyond basic use cases, issues emerge.

Langchain: An Overview of my Experience with the LLM Framework

A foundational problem with LLMs: a takeaway from Andrej Karpathy’s lecture on GPT Tokenization.

Why LLMs mess up the simple things

The release of Gemini 1.5 Pro, a glimpse into the difficulty in fine-tuning LLM guardrails

Whose Bias Do You Want?

So much is happening in AI, how do you keep up? Here are some of the best resources to follow to stay up to date on AI research and applications.

Fantastic ML/AI and Where to Find Them: A Guide to the Best ML/AI Resources

Are there easy ways to improve the inference speed of an auto-regressive model?

Exploration of Inference Speed of 🤗 Transformers Auto-Regressive Model

Using cURL to imitate web browser HTTP traffic

Reverse-Engineering of Website Network Requests

Early adopters of LLM integrations risk unexpected outcomes

Can we trust ChatGPT integrations into consumer sites?

Looking at the source code to see what makes Mixtral-8x7B so powerful

Diving deep into Mistral AI and their newest model, Mixtral-8x7B

In a year full of AI news and hype, which events are going to have the biggest enduring impact on AI?

The most important AI development of 2023 was maybe not GPT-4

Generative AI is showing up everywhere: how do we avoid trusting it more than we should?

Automation Bias in the use of Large Language Models

How I use Github Copilot

How to find your iMessage audio messages and send them through distil-whisper ASR

Find and Transcribe iMessage Voice Memos

How to use Retrieval Augmented Generation (RAG) to connect LlaMA-2 with a large number of internet sources

Who am I? How to give an LLM access to new data

One application of the recent Executive Order from the White House

AI Safety: Snippets of Biden's Executive Order on AI

Using Google's Bard LLM for Creating a Flask and React Recipe Review Application

Using Google's Bard LLM for Rapid Web Development

Some hidden tips and tricks to get the most out of sagemaker training jobs

Training in Sagemaker with Huggingface, Tips & Tricks

Upgrade your code with chat model templates

The New Huggingface Templates for Chat Models Feature

Train LlaMA-2 LLM on your own emails, Part 2. Model Training

Prepare a dataset to train an LLM to answer emails for you

Train a LlaMA-2 LLM for free to be an email autocomplete assistant using your own email data! Part 1. Introduction and Data Preparation

How is the pad token handled in training a transformer and what's the impact of setting the pad token to be the same as the eos token?

All Posts

Survey of Current Modified Transformer Attention Designs

AI now stands for Apple Intelligence

Picking the right LLM deployment configuration in AWS

Is your data more or less important in the age of LLMs?

How to Deploy Bedrock Agents using AWS CDK