MikeBee

11 ms·

1.
Show HN: Routed Attention – 75-99% savings by routing between O(N) and O(N²)(zenodo.org)
1 points by MikeBee 8mo ago|1 comments
2.
by MikeBee 9mo ago
Friston Was Right, But Implementations Have Been Wrong.
3.
by MikeBee 10mo ago
Towers of Hanoi
4.
The Impossible Mathematics of Tiny Recursive Models(medium.com)
1 points by MikeBee 11mo ago|0 comments
5.
by MikeBee 1y ago
speed up transformer inference by 60%
6.
What vs. How(5boiryzrsw74q.ok.kimi.link)
2 points by MikeBee 1y ago|0 comments
7.
by MikeBee 1y ago
"I've discovered that attention heads in transformer models can be approximated by simple MLPs using only 5% of the original parameters while maintaining nearly identical performance. This could significantly reduce the power cons
8.
by MikeBee 2y ago
I've built an open-source framework that lets anyone generate synthetic AI training data locally using Ollama. The project tackles the challenge of creating high-quality, structured examples for teaching specialized capabilities to lan
9.
Why logits are not log-odds in neural nets(medium.com)
2 points by MikeBee 2y ago|1 comments
10.
by MikeBee 2y ago
Ollama Agents - Advanced AI Assistant Builder with Graph Knowledge-base We're excited to announce major updates to Ollama_Agents, our open-source toolkit for building sophisticated AI assistants. New features include: 1. JSON-based Gra
11.
by MikeBee 2y ago
This is the first agent framework just for Ollama. Stop paying huge sums of money to develop agents. Develop them on Ollama and port them later.
12.
by MikeBee 2y ago
n the ever-evolving field of machine learning, a recent discovery by Timothy Nguyen offers fresh insights into how we might improve the training of large language models.
13.
CodeMapper: AI's Guide to Understanding Code(medium.com)
2 points by MikeBee 2y ago|0 comments
14.
Revolutionize LLM Coding with Enhanced Context Mapping(github.com)
1 points by MikeBee 2y ago|2 comments
15.
by MikeBee 2y ago
Hey Hacker News! I'm excited to share a new tool I've been working on called `codemapper`. This Python package is designed to map a code repository for LLM (Large Language Model) editing, providing a more efficient and context-awa
16.
by MikeBee 2y ago
That's cool. You can switch the model to whatever you like. type /help and find the command to switch models.
17.
by MikeBee 2y ago
Thanks. I think I figured out how to do this for new posts. The main differences between RAGAGENT and openwebui is that chat history is vector searched for context assemble and fabric is used for system and user messages. It's not j
18.
by MikeBee 2y ago
Thanks! I've been using gemma 2 9B and Llamma3 8B interchangeably. They're both really good models.
19.
RAGAgent Now Has Fabric(github.com)
4 points by MikeBee 2y ago|1 comments
20.
by MikeBee 2y ago
As of this week, there are new updates including the latest system and user messages from fabric, an updated architecture guide, and new slash commands. Two things make RAGAgent unique: It's an all local solution, and it's made to
21.
New FOSS All-Local AI with Ollama, Gemma 2, and Fabric
33 points by MikeBee 2y ago|6 comments