Daniel Mendelevitch

@MiSunKw @runwayml the fun fact is that i have no fun facts

@danielmend_ @runwayml Where’s the fun fact?!

life update: i joined @runwayml this week :DD

Today, we release Inductive Moment Matching (IMM): a new pre-training paradigm breaking the algorithmic ceiling of diffusion models. Higher sample quality. 10x more efficient. Single-stage, single network, stable training. Read more: lumalabs.ai/news/imm

Image to Video with #Ray2 is here. Drop any image into Dream Machine to make it a video and bring it to life — from historic artifacts and paintings to memes, custom art and photos of any kind. Available now.

check out our new video model😄!!!

Decentralized Diffusion Models power stronger models trained on more accessible infrastructure. DDMs mitigate the networking bottleneck that locks training into expensive and power-hungry centralized clusters. They scale gracefully to billions of parameters and generate photorealistic images with just a week of training on eight independent GPU nodes. They’re easy to implement, adopt DiT hyperparameters directly and outperform standard models FLOP-for-FLOP.

👀

Say hello to the all-new #DreamMachine. 🚀 Your visual thought partner—where ideas become reality. Ideate, visualize, and share your ideas with the world. Available now.

life update: i started interning at @LumaLabsAI today :D

Excited to announce our latest paper analyzing the computational tradeoffs between image synthesis approaches. We compare the text-to-image capabilities of diffusion, next-token, and masked-token prediction across various compute budgets. We also briefly investigate the impact of several modules in the recipe - the autoencoder, conditioning method, and usage of EMA on the model weights. arxiv.org/abs/2405.13218

@daniel_c0deb0t @AnthropicAI congrats!!

Bit of a plot twist, but I'll be joining @AnthropicAI after I graduate undergrad at UCLA to work on LLM research tools

@_neelr_ @_pranavnt @lnofx seqlen/ctx len is 77, the embeddings themselves are higher dimensional

@_pranavnt @lnofx for text it’s 77 i think just arbitrary and then projected to 768 for cosine similarity

holy fuck

game on

@dluozhang @samueloshay nobody has ever said this, like ever

@samueloshay Sleeping on San Ramon

Real alpha: raise your kids in the East Bay (Lafayette, Walnut Creek, Alamo, Danville, etc).

@dluozhang the el segundo canes should be a holy site

Standing in El Segundo, Mattel in front of me, Raytheon behind me

@TheStithLord dude this is so cool

Real-time AI is here, and it's insane. Generate faster than you can scroll 🏎️ When should we ship this? #StableVideo #SDXLTurbo #aivideo

@DigThatData @CoreWeave @StabilityAI congrats!

Exciting news: I've joined @CoreWeave's ML team!! I worked closely with their lead Wes Brown while driving early DreamStudio backend development @StabilityAI and am super looking forward to working with Wes full time. Open Source AI goes brrrrrr!

@unixpickle 🤷‍♂️

Anybody else find it a bit surprising how many independent video models are suddenly dropping?

Real-time high-quality sampling is a totally different experience, try it out here: clipdrop.co/stable-diffusi…