Log inSign up
Ostris
2,109 posts
user avatar
Ostris
@ostrisai
AI / ML researcher and developer. Creator of AI Toolkit - github.com/ostris/ai-tool… Support my work - ostris.com/support
Denver, Co
ostris.com
Joined August 2023
320
Following
13.2K
Followers
  • Pinned
    user avatar
    Ostris
    @ostrisai
    May 23, 2025
    If you appreciate my contributions to open source software and models, please consider becoming a sponsor. Supporting open source projects and their creators is vital for open innovation, and financial support from people like you makes my work possible.
    50K
  • user avatar
    Ostris
    @ostrisai
    Jan 31, 2025
    If someone wants to beat Nvidia, the ONLY thing you need to do is make GPUs with 128, 256, 512, 1024 GB of VRAM. Doesn't need to be faster. Doesn't even need great tooling. We need more VRAM. Nvidia doesn't seem to get this, and the first company that does will dethrone them.
    596K
  • user avatar
    Ostris
    @ostrisai
    Oct 1, 2025
    So, generating Sora 2 videos and scrolling the feed is my new favorite thing. I love it. The ultimate AI slop machine.
    00:00
    157K
  • user avatar
    Ostris
    @ostrisai
    Jan 19, 2025
    I am happy to release the next iteration of OpenFlux, Flex.1-alpha. Flex.1 is an 8B param model with an Apache 2.0 license. It has been trained a significant amount since OpenFlux, and has a fancy new guidance embedder. Still a WIP. More to come.
    ostris/Flex.1-alpha · Hugging Face
    From huggingface.co
    77K
  • user avatar
    Ostris
    @ostrisai
    Mar 28, 2025
    ChatGPT’s multi model image gen is absolutely revolutionary. It can do some amazing things, but why is doing something we’ve been able to do with control nets and LoRAs for a very long time going viral? It is like getting a self driving car and being impressed with the horn.
    104K
  • user avatar
    Ostris
    @ostrisai
    Aug 9, 2024
    Did a lot of testing on my LoRA training script for @bfl_ml FLUX.1 dev model. Amazing model! I think it is finally ready. Running smooth on a single 4090. Posting a guide tomorrow. Special thanks to @araminta_k for helping me test on her amazing original character and artwork.
    00:00
    981K
  • user avatar
    Ostris
    @ostrisai
    Feb 15, 2024
    I teared up a bit. I am extremely excited, but also feel completly inadequate in literally everything I have ever worked on. Ever. It is absolutely stunning and humbling to watch. I need a drink.
    user avatar
    OpenAI
    @OpenAI
    Feb 15, 2024
    Replying to @OpenAI
    Prompt: “A movie trailer featuring the adventures of the 30 year old space man wearing a red wool knitted motorcycle helmet, blue sky, salt desert, cinematic style, shot on 35mm film, vivid colors.”
    00:00
    103K
  • user avatar
    Ostris
    @ostrisai
    Aug 28, 2024
    Testing Flux.1 schnell embedding interpolation. I told claude some details about my life and had it generate 40 prompts from being born -> being dead using the data I gave it. Then interpolated between the embeddings for this video. Uses @araminta_k softpasty LoRA
    00:00
    46K
  • user avatar
    Ostris
    @ostrisai
    Apr 22, 2025
    Flex.2-preview is here with text to image, universal control (line, pose, depth), and inpainting all baked into one model. Fine tunable with AI-Toolkit, Apache 2.0 license, 8B parameters. Link in 🧵
    40K
  • user avatar
    Ostris
    @ostrisai
    Jan 31, 2025
    Max VRAM Consumer GPUS: 2016 - 1080 - 8GB 2020 - 3090 - 24GB (3x) - AI revolution of huge consumer models requiring significantly more VRAM happens - 2025 - 5090 - 32GB (1.5x) The 5090 should have had a minimum of 72GB (3x).
    user avatar
    Ostris
    @ostrisai
    Jan 31, 2025
    If someone wants to beat Nvidia, the ONLY thing you need to do is make GPUs with 128, 256, 512, 1024 GB of VRAM. Doesn't need to be faster. Doesn't even need great tooling. We need more VRAM. Nvidia doesn't seem to get this, and the first company that does will dethrone them.
    22K
  • user avatar
    Ostris
    @ostrisai
    Aug 5, 2024
    IT WORKS!! I trained an IKEA instruction LoRA on a decompressed version of FLUX.1 schnell , and it works on schnell at 4 step! The choppiness should go away when I get the guidance embedding trained. Still a WIP, but it works! There is hope! huggingface.co/ostris/FLUX.1-…
    36K
  • user avatar
    Ostris
    @ostrisai
    Sep 29, 2023
    New Stable Diffusion XL LoRA, Ikea Instructions. SDXL does an amazingly hilarious job at coming up with how to make things. Special thanks to @multimodalart and @huggingface for the GPU grant!! HF -> huggingface.co/ostris/ikea-in… Civitai -> civitai.com/models/153586/…
    163K
  • user avatar
    Ostris
    @ostrisai
    Aug 17, 2024
    AI-Toolkit now officially supports training LoRAs directly on FLUX.1-schnell. 🥳 How to do it here -> github.com/ostris/ai-tool… The adapter that makes it possible here -> huggingface.co/ostris/FLUX.1-…
    36K
  • user avatar
    Ostris
    @ostrisai
    Aug 29, 2024
    FLUX.1 schnell text embedding interpolation test. Silver balloons floating in the street.
    00:00
    33K

New to X?

Sign up now to get your own personalized timeline!

Create account

By signing up, you agree to the Terms of Service and Privacy Policy, including Cookie Use.

Terms·Privacy·Cookies·Accessibility·Ads Info·© 2026 X Corp.
Don't miss what's happening
People on X are the first to know.
Log inSign up