EARLY EXPLORTATION ON LOCAL RAN STABLE-DIFFUSION MODELS

I’m in the early stages of exploring AI-powered image generation, one of the most exciting developments has been the ability to run Stable Diffusion models locally, even on average consumer-grade hardware like the NVIDIA GeForce RTX 3060 Ti. Despite not being the top-tier GPU, this graphics card still provides enough power to experiment with high-quality image refinement techniques.

For this demonstration, I started with a low-quality reference image of a race-car (512x512 pixels).

CREATING STICKERS WITH AI ASSISTANCE

Recently I’ve seen loads of arts and crafts on my social platforms, and I wanted to see how accessible some of these online sticker printing services were.

As someone without a talent for drawning I also wanted to see how far AI, most specifically DALL-E could take me.

I’m no stranger to photography, and vector art softwares, and for this challenege I used Gimp and Inkscape as I’m running a pure Linux environment as my daily driver (and wanted to use open source software).

BUILD AI-GENERATED MICRO-SITES WITH LANDER!

Excited to unveil my latest passion project, Lander!

Combining the power of AI with web site management, Lander is not just another CMS, its integrated cron scheduling feature empowers users to automate content creation seamlessly.

AI GENERATED BLOGS

Found myself with a bit of extra time on my hands lately, so I decided to take the opportunity to explore some AI generated content in the form of micro blogs.

My goal was to write some python that interfaces with OpenAI’s api (gpt3.5 turbo model), writes focused content on randomized topics, and then warps that up in a Hugo static site, aka micro blog.

I started with an expandable YaML configuration. With this, I could define multiple terms which get randomized into search queries, this ultimately makes as many posts as I ask for:

CAN NOSTR EVENTS BE MANIPULATED?

If you haven’t been following along, I have a couple of nostr posts at this point.

https://nessy.info/tags/nostr/

I’ve been trying to further understand nostr by deep diving the protocol. At this point my thought is how mutable are nostr messages (events), I understand that during broadcast the relay verifies the signature, but then they need to store these events in some centralized database, right? Could a rogue relay for example accept your event, then alter it at a later time? Well in this post I hope to explore just that.

SIMPLE TOOLS FOR INTERACTING WITH THE NOSTR PROTOCOL

If you haven’t had a chance to ready my previous post about nostr, it is probably worth checking out as it give a detailed, step by step explanation.

https://nessy.info/post/2023-02-16-deciphering-nostr-and-its-private-keys/

With knowledge gained from the previous post I decided to put together a couple of rough python scripts, this is to handle a few of our previously manual steps.

Head on over to my Github and check out my nostr_stuff repository:

SIMPLE CLI FOR CATEGORIZING AND SENTIMENT OF TEXT

Now that I’ve spent some time with huggingface.co, specifically their NPL Course (natural language processing) I wanted to combine a couple of the learnings into a simple python script.

What I ended up with was a script that could both categorize using a zero-shot-classification model, as well as get sentiment using a sentiment-analysis model.

You can interact with this script in one of two ways, first by sending a string as input during execution:

EARLY EXPLORATION OF LARGE LANGUAGE MODELS ON PYTHON

With the current hype on artificial intelligence and platforms like OpenAI’s ChatGPT, I decided it’s about time I explore.

I’m not going to lie, I’m not exactly excited about a company named OpenAI being a closed source, for profit platform; however, this has been one of the reason I decided to explore offline, open, and self hosted solutions (in the hopes of creating my own models).

If you recall I’ve experimented with python and machine learning in the past, look no further than my 2016 Test your machine learning blog post.

KUBERNETES RASPBERRY PI LAB ENVIRONMENT SETUP USING ANSIBLE

After manually setting up a Lightweight Kubernetes cluster on a few of my Raspberry Pi’s I decided to tear it all down, and rebuild it from scratch using Ansible, and an infrastructure as code strategy.

This gives me a chance to catchup a bit on ansible, and keep my documentation and notes for this project in a replay-able Github repository. The project at the moment is rough, and contains sensitive details; however, this is a local lab environment that will most likely be trashed later on, so I’m not worried.

RASPBERRY PI KUBERNETES CLUSTER

I figured now the perfect time for me to explore the lightweight Kubernetes project k3s.

The documentation was really solid, plus I found Alex Ortner’s Medium blog post very helpful.

I dusted of a Raspberry Pi 4, and three Raspberry Pi 3’s for this setup.

I’m still in the mists of my kubernetes journey, but I wanted to share some of the early primitives, and some of the notes I’ve written for myself.