Pinned Tweet
After 8 months of long coding nights ☕️ we finally officially release Pythae 🥳, a python library unifying generative autoencoder implementations including vaegan🥗, vqvae or RAEs.
🖥️ github repo: github.com/clementchadebec/b…
👉paper: arxiv.org/abs/2206.08309
Excited to share that I've joined @runwayml as a Member of Technical Staff, Research, working on real-time world models.
I am deeply grateful for the research opportunities I had and the people I met, first at @StabilityAI and then at @heyjasperai.
Can’t wait for what’s next!
Clément Chadebec retweeted
Introducing Praxis-1, an open-weight World Action Model that turns our video pre-training expertise into real world control for robots.
Robot demonstration data is limited. But video data is infinite. Praxis-1 extends our bet that the best policy models will learn from video, and use that scale to bring embodied intelligence to every industry.
Praxis-1 is built on our large-scale video pre-training and real-time infrastructure, and acts as a policy model for robotics developers and researchers that works across any embodiment or environment, no matter how complex.
We are actively testing with early partners including Noble Machines, Standard Bots and Ultra, and will be releasing the weights publicly in the coming months.
Learn more and request early access at runway.com/Praxis-1
Clément Chadebec retweeted
Exciting news from Jasper Research! 🚀
We built an internal 4B text-to-image model from scratch that turned out to be one of the fastest models out there!
1k image generation in less than a second!
Our technical report shares everything we learned along the way.
👇
Clément Chadebec retweeted
very nice dataset drop from the clipdrop team 🙌
📢 New @heyjasper release ! 📢
MONET 🌸 : An Apache2.0 deduped and recaptioned dataset of 105M samples unlocking reproducible text-to-image research.
Nano T2I 🖌️ : A codebase to train your own T2I model
🤗 @huggingface: huggingface.co/datasets/jasp…
💻: github.com/gojasper/nano-t2i
Very excited about this new release, pushing the boundaries of open and reproducible T2I research.
Congrats to the team!
Benjamin Aubin Gonzalo Quintana @onurxtasar @UlaLaParis @_jeev2 @dh7net @clipdropapp @heyjasperai
Clément Chadebec retweeted
We are starting to be quite bullish about getting in the data infrastructure business.
I just cloned 68 TB (while I only have a 4TB local disk) to my @huggingface training bucket in 1 minute 55 seconds, thanks to Xet deduplication and all our infra optimizations.
You can host your data processing pipelines on HF and leverage those insane optimizations 🔥
Clément Chadebec retweeted
wow 🤯
68 TB, high-quality, high-resolution, Apache 2.
Thank you! 🙌🫡
📢 New @heyjasper release ! 📢
MONET 🌸 : An Apache2.0 deduped and recaptioned dataset of 105M samples unlocking reproducible text-to-image research.
Nano T2I 🖌️ : A codebase to train your own T2I model
🤗 @huggingface: huggingface.co/datasets/jasp…
💻: github.com/gojasper/nano-t2i
Very excited about this new release, pushing the boundaries of open and reproducible T2I research.
Congrats to the team!
Benjamin Aubin Gonzalo Quintana @onurxtasar @UlaLaParis @_jeev2 @dh7net @clipdropapp @heyjasperai
Clément Chadebec retweeted
Ah! this is insane, I've been complaining for weeks about the lack of open text-to-image data! Congrats on the release @heyjasperai
📢 New @heyjasper release ! 📢
MONET 🌸 : An Apache2.0 deduped and recaptioned dataset of 105M samples unlocking reproducible text-to-image research.
Nano T2I 🖌️ : A codebase to train your own T2I model
🤗 @huggingface: huggingface.co/datasets/jasp…
💻: github.com/gojasper/nano-t2i
Very excited about this new release, pushing the boundaries of open and reproducible T2I research.
Congrats to the team!
Benjamin Aubin Gonzalo Quintana @onurxtasar @UlaLaParis @_jeev2 @dh7net @clipdropapp @heyjasperai
Clément Chadebec retweeted
Huge open release from @heyjasperai : MONET
105M curated image-text pairs, Apache 2.0, with embeddings, VAE latents, multi-VLM captions, and a companion training repo (nano-t2i) to train a T2I model end-to-end on one H200 for <$300.
Congrats @CChadebec & co 👏
📢 New @heyjasper release ! 📢
MONET 🌸 : An Apache2.0 deduped and recaptioned dataset of 105M samples unlocking reproducible text-to-image research.
Nano T2I 🖌️ : A codebase to train your own T2I model
🤗 @huggingface: huggingface.co/datasets/jasp…
💻: github.com/gojasper/nano-t2i
Very excited about this new release, pushing the boundaries of open and reproducible T2I research.
Congrats to the team!
Benjamin Aubin Gonzalo Quintana @onurxtasar @UlaLaParis @_jeev2 @dh7net @clipdropapp @heyjasperai
Clément Chadebec retweeted
With 104M of image-text pairs, this is one of the largest, if not the largest, openly-licensed image dataset
And it's on @huggingface!!
Kudos @heyjasperai
📢 New @heyjasper release ! 📢
MONET 🌸 : An Apache2.0 deduped and recaptioned dataset of 105M samples unlocking reproducible text-to-image research.
Nano T2I 🖌️ : A codebase to train your own T2I model
🤗 @huggingface: huggingface.co/datasets/jasp…
💻: github.com/gojasper/nano-t2i
Very excited about this new release, pushing the boundaries of open and reproducible T2I research.
Congrats to the team!
Benjamin Aubin Gonzalo Quintana @onurxtasar @UlaLaParis @_jeev2 @dh7net @clipdropapp @heyjasperai
📢 New @heyjasper release ! 📢
MONET 🌸 : An Apache2.0 deduped and recaptioned dataset of 105M samples unlocking reproducible text-to-image research.
Nano T2I 🖌️ : A codebase to train your own T2I model
🤗 @huggingface: huggingface.co/datasets/jasp…
💻: github.com/gojasper/nano-t2i
Very excited about this new release, pushing the boundaries of open and reproducible T2I research.
Congrats to the team!
Benjamin Aubin Gonzalo Quintana @onurxtasar @UlaLaParis @_jeev2 @dh7net @clipdropapp @heyjasperai
Using the MONET dataset exclusively, we trained a 4B T2I model from scratch. Built on an MMDiT-inspired architecture and trained via latent flow matching with a deep compression VAE, the model can generate images up to 2048x2048 resolution.
📜 : arxiv.org/abs/2605.21272
💻: github.com/gojasper/nano-t2i
We put in place a rigorous and meticulous filtering, deduplicating, and re-captioning pipeline to create MONET:
⛽ Sourced from 2.9B images from open datasets (LAION, COYO, etc.)
✅ Filtered for high-res, aesthetics & strict safety/NSFW standards
👬 Deduplicated & stripped of stock/watermarked images
💬 Re-captioned using 4 top VLMs for rich, diverse text descriptions
🕹️ Augmented with safe, permissive synthetic data
Clément Chadebec retweeted
We’re thrilled to announce that the Jasper Research team’s latest paper on Latent Bridge Matching (LBM) was accepted to the prestigious IEEE/CVF International Conference on Computer Vision! Thank you @ICCVConference for the recognition, and congratulations to the team: @onurxtasar, Sanjeev Sreetharan, Benjamin Aubin, Gonzalo Quintana, @dh7net, and @CChadebec.
LBM is a technique for rapid, high-fidelity image-to-image translation. Learn more about how LBM is changing visual marketing for retailers. bit.ly/3ZWMEcI
Clément Chadebec retweeted
Object Relighting with Latent Bridge Matching
Link to Hugging Face space below + Link to Kijai's ComfyUI wrapper in first comment 👇
Clément Chadebec retweeted
Proud to share the latest breakthrough from Jasper Research: Latent Bridge Matching (LBM) just got a major update!
Our team has released three new open-source models on Hugging Face—supporting depth, normals, and relighting—along with training code and a Gradio demo for object relighting.
This work continues our mission to advance controllable, high-quality AI image generation for enterprise marketing teams.
🔗 Explore the research, models, and demo: linkedin.com/feed/update/urn…
👏 Congrats to @CChadebec, Benjamin Aubin, @OnurXTasar, and @dh7net for pushing the field forward.
📢 Latent Bridge Matching update! 📢
New @heyjasperai release 🚀
- 3 LBM models on @huggingface (relighting, depth, normals)
- LBM training code
Also see our @Gradio space for object relighting!
🤗 huggingface.co/jasperai/LBM_…
💻 github.com/gojasper/LBM
📗 arxiv.org/abs/2503.07535
Clément Chadebec retweeted
Very proud of the team!
📢 Latent Bridge Matching update! 📢
New @heyjasperai release 🚀
- 3 LBM models on @huggingface (relighting, depth, normals)
- LBM training code
Also see our @Gradio space for object relighting!
🤗 huggingface.co/jasperai/LBM_…
💻 github.com/gojasper/LBM
📗 arxiv.org/abs/2503.07535
📢 Latent Bridge Matching update! 📢
New @heyjasperai release 🚀
- 3 LBM models on @huggingface (relighting, depth, normals)
- LBM training code
Also see our @Gradio space for object relighting!
🤗 huggingface.co/jasperai/LBM_…
💻 github.com/gojasper/LBM
📗 arxiv.org/abs/2503.07535
Today, we release on @huggingface, the model for object relighting where the model should relight a foreground object so that it blends perfectly with a target background.
🤗Ckpt: huggingface.co/jasperai/LBM_…
🤗Demo: huggingface.co/spaces/jasper…
We also release checkpoints for normal and depth estimation where the model should translate an input image into a normal or depth map.
🤗 Ckpt: huggingface.co/jasperai/LBM_…
🤗 Ckpt: huggingface.co/jasperai/LBM_…
