
Heurist has integrated the state-of-the-art image generation model, Flux.1 dev, developed by bfl_ml, into its Imagine and Miner platforms. This model excels in prompt following, text generation, and image composition, matching the performance of closed-source alternatives. Additionally, the FLUX fine-tuner, based on Ostrisai's AI toolkit, has been open-sourced, adding capabilities for captioning, LoRA inference, and HF upload. Efforts are underway to enhance both training and inference speeds, with all optimizations to be made open source.
Check this great prompt shared @ralphlentjes 👇👇 Mine done with @Sezam_AI Ultra 2 Model https://t.co/ceICI6OR9t https://t.co/lLVM2uFI1U
3 images with similar prompts. The original was generated using stable diffusion and appears within the official documentation. Using Super Describe Pro, I reverse engineered the prompt and generated 2 further images using Flux locally. Which is your favourite? #Flux https://t.co/f3hNO3qZSW
training a Flux Lora on the bizarre little details in Hieronymous Bosch's Garden of Earthly Delights can you tell which of these were generated? 😅 link to glif below https://t.co/sUuLY87Hlt







