When experimenting with ML models, I use Google Cloud Platform (GCP) with 24G VRAM GPU. Not great, but better than at home. I set up the GCP such that I can restart from cold to a working model within minuutes. Most of this is scripted via a Makefile...
Haverscript is a lightweight Python library designed for working with Large Language Models (LLMs) and has recently introduced agent functionality in its latest version, 0.3. The library offers a concise, readable syntax for crafting prompts, chainin...
Haverscript, a Domain Specific Language designed for building LLM-based agents, continues to evolve with new features and improvements aimed at enhancing its flexibility and usability. With the recent updates, Haverscript has introduced significant c...
While doing testing of the together.ai API for haverscript, I came across a problem. Together’s API seem to be ignoring the seed argument. The seed argument provides reproducibility, which is essential for comparing with a golden output. The followin...
Today, I discovered that the sigmoid function is equivalent to softmax when applied to two inputs, with one input fixed to zero. With almost a decade of experience in machine learning, and earlier work in compute for signal processing, I had never co...
In HaverScript we use Markdown as an output format. We represent prompts by prefixing >, sometimes called bird-tracks, before the prompt, and the responses are normal markdown text. This gives both a readable output “script” of what happened, as well...
YouTube is full of wonderful tutorials about all aspects of ML and AI. Here are the three best deep-dive into LLMs videos I’ve found. Background about what a LLM, such as ChatGPT, is. Intro to Large Language Models by Andrej Karpathy. This is a great...
A Python library for interacting with local LLMs
