Packing Input Frame Context in Next-Frame Prediction Models for Video Generation
Paper • 2504.12626 • Published • 51
AI & ML interests
None defined yet.
Recent Activity
View all activity
lllyasviel
authored 2
papers about 20 hours ago
Post
5372
👉 What happened in AI in 2025? 👈
We prepared the 2025 version of the HF AI Timeline Grid, highlighting open vs API-based model releases, and allowing you to browse and filter by access, modality, and release type!
Play with it here:
2025-ai-timeline/2025-ai-timeline
Here's my personal quarterly TL;DR:
1️⃣ Q1 — Learning to Reason
Deepseek not only releases a top-notch reasoning model, but shows how to train them and compete with closed frontier models. OpenAI debuts Deep Research.
Significant milestones: DeepSeek R1 & R1-Zero, Qwen 2.5 VL, OpenAI Deep Research, Gemini 2.5 Pro (experimental)
2️⃣ Q2 — Multimodality and Coding
More LLMs embrace multimodality by default, and there's a surge in coding agents. Strong vision, audio, and generative models emerge.
Significant milestones: Llama 4, Qwen 3, Imagen 4, OpenAI Codex, Google Jules, Claude 4
3️⃣ Q3 — "Gold" rush, OpenAI opens up, the community goes bananas
Flagship models get gold in Math olympiads and hard benchmarks. OpenAI releases strong open source models and Google releases the much anticipated nano-banana for image generation and editing. Agentic workflows become commonplace.
Significant milestones: Gemini and OpenAI IMO Gold, gpt-oss, Gemini 2.5 Flash Image, Grok 4, Claude Sonnet 4.5
4️⃣ Q4 — Mistral returns, leaderboard hill-climbing
Mistral is back with updated model families. All labs release impressive models to wrap up the year!
Significant milestones: Claude Opus 4.5, DeepSeek Math V2, FLUX 2, GPT 5.1, Kimi K2 Thinking, Nano Banana Pro, GLM 4.7, Gemini 3, Mistral 3, MiniMax M2.1 🤯
Credits
🙏 NHLOCAL for the source data https://github.com/NHLOCAL/AiTimeline
🫡 @reach-vb for the original idea, design and recipe
🙌 @ariG23498 and yours truly for compiling and verifying the 2025 edition
🥳 Here's to 2026, wishing it becomes the best year ever for open releases and on-device-first use-cases! 🥂
We prepared the 2025 version of the HF AI Timeline Grid, highlighting open vs API-based model releases, and allowing you to browse and filter by access, modality, and release type!
Play with it here:
2025-ai-timeline/2025-ai-timeline
Here's my personal quarterly TL;DR:
1️⃣ Q1 — Learning to Reason
Deepseek not only releases a top-notch reasoning model, but shows how to train them and compete with closed frontier models. OpenAI debuts Deep Research.
Significant milestones: DeepSeek R1 & R1-Zero, Qwen 2.5 VL, OpenAI Deep Research, Gemini 2.5 Pro (experimental)
2️⃣ Q2 — Multimodality and Coding
More LLMs embrace multimodality by default, and there's a surge in coding agents. Strong vision, audio, and generative models emerge.
Significant milestones: Llama 4, Qwen 3, Imagen 4, OpenAI Codex, Google Jules, Claude 4
3️⃣ Q3 — "Gold" rush, OpenAI opens up, the community goes bananas
Flagship models get gold in Math olympiads and hard benchmarks. OpenAI releases strong open source models and Google releases the much anticipated nano-banana for image generation and editing. Agentic workflows become commonplace.
Significant milestones: Gemini and OpenAI IMO Gold, gpt-oss, Gemini 2.5 Flash Image, Grok 4, Claude Sonnet 4.5
4️⃣ Q4 — Mistral returns, leaderboard hill-climbing
Mistral is back with updated model families. All labs release impressive models to wrap up the year!
Significant milestones: Claude Opus 4.5, DeepSeek Math V2, FLUX 2, GPT 5.1, Kimi K2 Thinking, Nano Banana Pro, GLM 4.7, Gemini 3, Mistral 3, MiniMax M2.1 🤯
Credits
🙏 NHLOCAL for the source data https://github.com/NHLOCAL/AiTimeline
🫡 @reach-vb for the original idea, design and recipe
🙌 @ariG23498 and yours truly for compiling and verifying the 2025 edition
🥳 Here's to 2026, wishing it becomes the best year ever for open releases and on-device-first use-cases! 🥂
- 3 replies
lllyasviel
submitted a
paper to Daily Papers 6 months ago
multimodalart
posted an update 9 months ago
Post
31636
Want to iterate on a Hugging Face Space with an LLM?
Now you can easily convert any HF entire repo (Model, Dataset or Space) to a text file and feed it to a language model!
multimodalart/repo2txt
Now you can easily convert any HF entire repo (Model, Dataset or Space) to a text file and feed it to a language model!
multimodalart/repo2txt
- 3 replies
multimodalart
posted an update about 1 year ago
Post
18490
Self-Forcing - a real-time video distilled model from Wan 2.1 by @adobe is out, and they open sourced it 🐐
I've built a live real time demo on Spaces 📹💨
multimodalart/self-forcing
I've built a live real time demo on Spaces 📹💨
multimodalart/self-forcing
- 6 replies
pcuenq
authored a
paper about 1 year ago
multimodalart
posted an update almost 2 years ago
Post
35666
New feature 🔥
Image models and LoRAs now have little previews 🤏
If you don't know where to start to find them, I invite you to browse cool LoRAs in the profile of some amazing fine-tuners: @artificialguybr , @alvdansen , @DoctorDiffusion , @e-n-v-y , @KappaNeuro @ostris
Image models and LoRAs now have little previews 🤏
If you don't know where to start to find them, I invite you to browse cool LoRAs in the profile of some amazing fine-tuners: @artificialguybr , @alvdansen , @DoctorDiffusion , @e-n-v-y , @KappaNeuro @ostris
- 3 replies
williamberman
authored a
paper about 2 years ago
multimodalart
posted an update about 2 years ago
Post
28660
The first open Stable Diffusion 3-like architecture model is JUST out 💣 - but it is not SD3! 🤔
It is Tencent-Hunyuan/HunyuanDiT by Tencent, a 1.5B parameter DiT (diffusion transformer) text-to-image model 🖼️✨, trained with multi-lingual CLIP + multi-lingual T5 text-encoders for english 🤝 chinese understanding
Try it out by yourself here ▶️ https://huggingface.co/spaces/multimodalart/HunyuanDiT
(a bit too slow as the model is chunky and the research code isn't super optimized for inference speed yet)
In the paper they claim to be SOTA open source based on human preference evaluation!
It is Tencent-Hunyuan/HunyuanDiT by Tencent, a 1.5B parameter DiT (diffusion transformer) text-to-image model 🖼️✨, trained with multi-lingual CLIP + multi-lingual T5 text-encoders for english 🤝 chinese understanding
Try it out by yourself here ▶️ https://huggingface.co/spaces/multimodalart/HunyuanDiT
(a bit too slow as the model is chunky and the research code isn't super optimized for inference speed yet)
In the paper they claim to be SOTA open source based on human preference evaluation!
Post
10363
OpenELM in Core ML
Apple recently released a set of efficient LLMs in sizes varying between 270M and 3B parameters. Their quality, according to benchmarks, is similar to OLMo models of comparable size, but they required half the pre-training tokens because they use layer-wise scaling, where the number of attention heads increases in deeper layers.
I converted these models to Core ML, for use on Apple Silicon, using this script: https://gist.github.com/pcuenca/23cd08443460bc90854e2a6f0f575084. The converted models were uploaded to this community in the Hub for anyone that wants to integrate inside their apps: corenet-community/openelm-core-ml-6630c6b19268a5d878cfd194
The conversion was done with the following parameters:
- Precision: float32.
- Sequence length: fixed to 128.
With swift-transformers (https://github.com/huggingface/swift-transformers), I'm getting about 56 tok/s with the 270M on my M1 Max, and 6.5 with the largest 3B model. These speeds could be improved by converting to
I'm also looking at optimizing inference using an experimental kv cache in swift-transformers. It's a bit tricky because the layers have varying number of attention heads, but I'm curious to see how much this feature can accelerate performance in this model family :)
Regarding the instruct fine-tuned models, I don't know the chat template that was used. The models use the Llama 2 tokenizer, but the Llama 2 chat template, or the default Alignment Handbook one that was used to train, are not recognized. Any ideas on this welcome!
Apple recently released a set of efficient LLMs in sizes varying between 270M and 3B parameters. Their quality, according to benchmarks, is similar to OLMo models of comparable size, but they required half the pre-training tokens because they use layer-wise scaling, where the number of attention heads increases in deeper layers.
I converted these models to Core ML, for use on Apple Silicon, using this script: https://gist.github.com/pcuenca/23cd08443460bc90854e2a6f0f575084. The converted models were uploaded to this community in the Hub for anyone that wants to integrate inside their apps: corenet-community/openelm-core-ml-6630c6b19268a5d878cfd194
The conversion was done with the following parameters:
- Precision: float32.
- Sequence length: fixed to 128.
With swift-transformers (https://github.com/huggingface/swift-transformers), I'm getting about 56 tok/s with the 270M on my M1 Max, and 6.5 with the largest 3B model. These speeds could be improved by converting to
float16. However, there's some precision loss somewhere and generation doesn't work in float16 mode yet. I'm looking into this and will keep you posted! Or take a look at this issue if you'd like to help: https://github.com/huggingface/swift-transformers/issues/95I'm also looking at optimizing inference using an experimental kv cache in swift-transformers. It's a bit tricky because the layers have varying number of attention heads, but I'm curious to see how much this feature can accelerate performance in this model family :)
Regarding the instruct fine-tuned models, I don't know the chat template that was used. The models use the Llama 2 tokenizer, but the Llama 2 chat template, or the default Alignment Handbook one that was used to train, are not recognized. Any ideas on this welcome!
- 5 replies
multimodalart
posted an update over 2 years ago
Post
The Stable Diffusion 3 research paper broken down, including some overlooked details! 📝
Model
📏 2 base model variants mentioned: 2B and 8B sizes
📐 New architecture in all abstraction levels:
- 🔽 UNet; ⬆️ Multimodal Diffusion Transformer, bye cross attention 👋
- 🆕 Rectified flows for the diffusion process
- 🧩 Still a Latent Diffusion Model
📄 3 text-encoders: 2 CLIPs, one T5-XXL; plug-and-play: removing the larger one maintains competitiveness
🗃️ Dataset was deduplicated with SSCD which helped with memorization (no more details about the dataset tho)
Variants
🔁 A DPO fine-tuned model showed great improvement in prompt understanding and aesthetics
✏️ An Instruct Edit 2B model was trained, and learned how to do text-replacement
Results
✅ State of the art in automated evals for composition and prompt understanding
✅ Best win rate in human preference evaluation for prompt understanding, aesthetics and typography (missing some details on how many participants and the design of the experiment)
Paper: https://stabilityai-public-packages.s3.us-west-2.amazonaws.com/Stable+Diffusion+3+Paper.pdf
Model
📏 2 base model variants mentioned: 2B and 8B sizes
📐 New architecture in all abstraction levels:
- 🔽 UNet; ⬆️ Multimodal Diffusion Transformer, bye cross attention 👋
- 🆕 Rectified flows for the diffusion process
- 🧩 Still a Latent Diffusion Model
📄 3 text-encoders: 2 CLIPs, one T5-XXL; plug-and-play: removing the larger one maintains competitiveness
🗃️ Dataset was deduplicated with SSCD which helped with memorization (no more details about the dataset tho)
Variants
🔁 A DPO fine-tuned model showed great improvement in prompt understanding and aesthetics
✏️ An Instruct Edit 2B model was trained, and learned how to do text-replacement
Results
✅ State of the art in automated evals for composition and prompt understanding
✅ Best win rate in human preference evaluation for prompt understanding, aesthetics and typography (missing some details on how many participants and the design of the experiment)
Paper: https://stabilityai-public-packages.s3.us-west-2.amazonaws.com/Stable+Diffusion+3+Paper.pdf
- 3 replies
multimodalart
posted an update over 2 years ago
Post
⚔️ The TIGERLab's Text2Image arena is here! ⚔️
TIGER-Lab/GenAI-Arena
Like https://huggingface.co/spaces/lmsys/chatbot-arena-leaderboard for LLMs: you prompt, two images emerge, vote for the best one 🏆
With enough votes this will lead to an Elo-based leaderboard for text-to-image models, go vote! 🗳️
TIGER-Lab/GenAI-Arena
TIGER-Lab/GenAI-Arena
Like https://huggingface.co/spaces/lmsys/chatbot-arena-leaderboard for LLMs: you prompt, two images emerge, vote for the best one 🏆
With enough votes this will lead to an Elo-based leaderboard for text-to-image models, go vote! 🗳️
TIGER-Lab/GenAI-Arena
multimodalart
posted an update over 2 years ago
Post
It seems February started with a fully open source AI renaissance 🌟
Models released with fully open dataset, training code, weights ✅
LLM - allenai/olmo-suite-65aeaae8fe5b6b2122b46778 🧠
Embedding - nomic-ai/nomic-embed-text-v1 📚 (sota!)
And it's literally February 1st - can't wait to see what else the community will bring 👀
Models released with fully open dataset, training code, weights ✅
LLM - allenai/olmo-suite-65aeaae8fe5b6b2122b46778 🧠
Embedding - nomic-ai/nomic-embed-text-v1 📚 (sota!)
And it's literally February 1st - can't wait to see what else the community will bring 👀
multimodalart
authored 2
papers over 2 years ago
lllyasviel
authored a
paper over 2 years ago
multimodalart
authored a
paper almost 3 years ago