DALL-E Homebrew In this post, we’ll look at getting setup with running your own A.I. image generator. You need a Linux system with a CUDA enabled card to get this working through to the end. I did get the basic txt-2img script working in Windows. Unfortunately, for the July released retrieval-augmented diffusion models, you need scann in order to index the openimages dataset, which is only available on Linux. WSL worked fine for me....
GPT-3 Raps A while back on Reddit, a rap about pencils written by GPT-3 was posted, which prompted me to do some Googling to try to generate and vocode the lyrics. I tried this with music myself a while back with Coqui-aiTTS and didn’t have great results. This time, I found a service called UberDuck that doesn’t work perfectly but is pretty good compared to what I had before. A link to the pencil vocoded rap can be found in that thread....
The Famous and Hilarious Comedian GPT-3’s davinci model was used. It took a few add-ons to my originally simple prompt to get anything other than descriptions of people being lost or dragon pictures strangely enough. MidJourney and Craiyon were used to generate an image for each prompt. You are a hilarious and famous comedian with incredible perception and imagination. create a set of photographs showing humorous and absurd scenarios. Describe each in detail....
Reproducing Vinyl Stickers In today’s post, we’ll look at creating and recreating vinyl stickers. This is one of the things Midjourney is particularly good at that could fairly easily translate into real or existing use cases. Services Used Today, and probably in the future, we’re just going to look at results with Craiyon and MidJourney. When I get access to DALL-E 2, I will start including that. Starry AI may make a reappearance when we look at art reproduction or creating new artworks in an artist’s style....
Comparing Apples to Apples To kick off the Daily A.I. section of this site, I wanted to start with something simple like how well the different services can render an apple. In the future, we’ll be doing other simple tests like this one, exploring different use cases for GPT-3 and/or AI Image generation, and work through prompt-crafting. DALL-E 2 recently went into beta and I’m keen to get an invite and begin including that tool in these tests....