diffusers
128 posts
@diffuserslib
Yes you're speaking with the 🤗 @huggingface 🧨 diffusers library personally
Video-to-video is now available in the official CogVideoX-5B Space 🔥 Try it out 🎥 ➡️🎥 huggingface.co/spaces/THUDM/C…
We now support loading and inferencing with two non-diffusers Flux LoRAs 1> X-Labs 2> Kohya (@kohya_tech) Thanks to @multimodalart for jamming on this with me! github.com/huggingface/di…
we just dropped an insane new release 🐣 - support to new pipelines: audio 🔊, video 🎬 and image 🖼️ models (FLUX, Stable Audio, CogVideoX, Kolors, AuraFlow and moar!) - native PAG support for image quality boost 💨 - AnimatedDiff 🤝 SparseCtrl github.com/huggingface/di…
Unity presents IPAdapter-Instruct Resolving Ambiguity in Image-based Conditioning using Instruct Prompts discuss: huggingface.co/papers/2408.03… Diffusion models continuously push the boundary of state-of-the-art image generation, but the process is hard to control with any nuance: practice proves that textual prompts are inadequate for accurately describing image style or fine structural details (such as faces). ControlNet and IPAdapter address this shortcoming by conditioning the generative process on imagery instead, but each individual instance is limited to modeling a single conditional posterior: for practical use-cases, where multiple different posteriors are desired within the same workflow, training and using multiple adapters is cumbersome. We propose IPAdapter-Instruct, which combines natural-image conditioning with ``Instruct'' prompts to swap between interpretations for the same conditioning image: style transfer, object extraction, both, or something else still? IPAdapterInstruct efficiently learns multiple tasks with minimal loss in quality compared to dedicated per-task models.
You should have already gone bonkers by now with @bfl_ml's FLUX release. What a model, eh! I am getting back to Twitter after some sprinting with my mates @_DhruvNair_, @YiYiMarz, and @multimodalart! Diffusers integration is, of course, there ❤️ We have also put together a little gist guiding you on how to run Flux with limited resources. gist.github.com/sayakpaul/b664…
The first open Stable Diffusion 3-like architecture model is JUST out 💣 - but it is not SD3! 🤔 It is HunyuanDiT by Tencent, a 1.5B parameter DiT (diffusion transformer) text-to-image model 🖼️✨ In the paper they claim to be SOTA open source! I'm working on a @huggingface demo as you read this so we can all vibe check Model: huggingface.co/Tencent-Hunyua… GitHub: github.com/Tencent/Hunyua… Paper: tencent.github.io/HunyuanDiT/ass…
