Post

Currently playing around with a prompt+sketch as an input to an image-to-image model. It feels like cheating but it's incredibly effective for translating initial ideas out of your head.

Dan Corin

@danielcorin.com

I'm using am using `black-forest-labs/FLUX.1-canny` running on together.ai. An amazing model with low latency. I tried at least 5 other providers before settling on this

together.aiTogether AI – The AI Acceleration Cloud - Fast Inference, Fine-Tuning & TrainingRun and fine-tune generative AI models with easy-to-use APIs and highly scalable infrastructure. Train & deploy models at scale on our AI Acceleration Cloud and scalable GPU clusters. Optimize perform...

18/1/25, 8:50 pm0 reposts0 likes

Dan Corin

@danielcorin.com

18 Jan 2025

Still pretty rough but got auto-color palette extraction working (the second row of colors). The algorithm works by sampling pixels, quantizing to 32-step RGB buckets, then picking most frequent colors that are sufficiently different (>100 RGB distance). Claude wrote it.