SynthLight: Portrait Relighting with Diffusion Model by Learning to Re-render Synthetic Faces
Paper
โข
2501.09756
โข
Published
โข
12
่ฐข้
sanaka87
AI & ML interests
Image Generation
Recent Activity
upvoted
a
paper
about 21 hours ago
SynthLight: Portrait Relighting with Diffusion Model by Learning to
Re-render Synthetic Faces
updated
a model
about 23 hours ago
sanaka87/3DIS
Organizations
None yet
sanaka87's activity
upvoted
a
paper
about 21 hours ago
reacted to
m-ric's
post with ๐
about 22 hours ago
Post
735
๐ ๐ถ๐ป๐ถ๐ ๐ฎ๐
'๐ ๐ป๐ฒ๐ ๐ ๐ผ๐ ๐๐๐ ๐ฟ๐ฒ๐ฎ๐ฐ๐ต๐ฒ๐ ๐๐น๐ฎ๐๐ฑ๐ฒ-๐ฆ๐ผ๐ป๐ป๐ฒ๐ ๐น๐ฒ๐๐ฒ๐น ๐๐ถ๐๐ต ๐ฐ๐ ๐๐ผ๐ธ๐ฒ๐ป๐ ๐ฐ๐ผ๐ป๐๐ฒ๐
๐ ๐น๐ฒ๐ป๐ด๐๐ต ๐ฅ
This work from Chinese startup @MiniMax-AI introduces a novel architecture that achieves state-of-the-art performance while handling context windows up to 4 million tokens - roughly 20x longer than current models. The key was combining lightning attention, mixture of experts (MoE), and a careful hybrid approach.
๐๐ฒ๐ ๐ถ๐ป๐๐ถ๐ด๐ต๐๐:
๐๏ธ MoE with novel hybrid attention:
โฃ Mixture of Experts with 456B total parameters (45.9B activated per token)
โฃ Combines Lightning attention (linear complexity) for most layers and traditional softmax attention every 8 layers
๐ Outperforms leading models across benchmarks while offering vastly longer context:
โฃ Competitive with GPT-4/Claude-3.5-Sonnet on most tasks
โฃ Can efficiently handle 4M token contexts (vs 256K for most other LLMs)
๐ฌ Technical innovations enable efficient scaling:
โฃ Novel expert parallel and tensor parallel strategies cut communication overhead in half
โฃ Improved linear attention sequence parallelism, multi-level padding and other optimizations achieve 75% GPU utilization (that's really high, generally utilization is around 50%)
๐ฏ Thorough training strategy:
โฃ Careful data curation and quality control by using a smaller preliminary version of their LLM as a judge!
Overall, not only is the model impressive, but the technical paper is also really interesting! ๐
It has lots of insights including a great comparison showing how a 2B MoE (24B total) far outperforms a 7B model for the same amount of FLOPs.
Read it in full here ๐ MiniMax-01: Scaling Foundation Models with Lightning Attention (2501.08313)
Model here, allows commercial use <100M monthly users ๐ MiniMaxAI/MiniMax-Text-01
This work from Chinese startup @MiniMax-AI introduces a novel architecture that achieves state-of-the-art performance while handling context windows up to 4 million tokens - roughly 20x longer than current models. The key was combining lightning attention, mixture of experts (MoE), and a careful hybrid approach.
๐๐ฒ๐ ๐ถ๐ป๐๐ถ๐ด๐ต๐๐:
๐๏ธ MoE with novel hybrid attention:
โฃ Mixture of Experts with 456B total parameters (45.9B activated per token)
โฃ Combines Lightning attention (linear complexity) for most layers and traditional softmax attention every 8 layers
๐ Outperforms leading models across benchmarks while offering vastly longer context:
โฃ Competitive with GPT-4/Claude-3.5-Sonnet on most tasks
โฃ Can efficiently handle 4M token contexts (vs 256K for most other LLMs)
๐ฌ Technical innovations enable efficient scaling:
โฃ Novel expert parallel and tensor parallel strategies cut communication overhead in half
โฃ Improved linear attention sequence parallelism, multi-level padding and other optimizations achieve 75% GPU utilization (that's really high, generally utilization is around 50%)
๐ฏ Thorough training strategy:
โฃ Careful data curation and quality control by using a smaller preliminary version of their LLM as a judge!
Overall, not only is the model impressive, but the technical paper is also really interesting! ๐
It has lots of insights including a great comparison showing how a 2B MoE (24B total) far outperforms a 7B model for the same amount of FLOPs.
Read it in full here ๐ MiniMax-01: Scaling Foundation Models with Lightning Attention (2501.08313)
Model here, allows commercial use <100M monthly users ๐ MiniMaxAI/MiniMax-Text-01
updated
a
model
about 23 hours ago
posted
an
update
3 days ago
Post
1641
๐ Excited to Share Our Latest Work: 3DIS & 3DIS-FLUX for Multi-Instance Layout-to-Image Generation! โค๏ธโค๏ธโค๏ธ
๐จ Daily Paper: 3DIS-FLUX: simple and efficient multi-instance generation with DiT rendering (2501.05131)
๐ Code is now open source!
๐ Project Website: https://limuloo.github.io/3DIS/
๐ GitHub Repository: https://github.com/limuloo/3DIS
๐ 3DIS Paper: https://arxiv.org/abs/2410.12669
๐ 3DIS-FLUX Tech Report: https://arxiv.org/abs/2501.05131
๐ฅ Why 3DIS & 3DIS-FLUX?
Current SOTA multi-instance generation methods are typically adapter-based, requiring additional control modules trained on pre-trained models for layout and instance attribute control. However, with the emergence of more powerful models like FLUX and SD3.5, these methods demand constant retraining and extensive resources.
โจ Our Solution: 3DIS
We introduce a decoupled approach that only requires training a low-resolution Layout-to-Depth model to convert layouts into coarse-grained scene depth maps. Leveraging community and company pre-trained models like ControlNet + SAM2, we enable training-free controllable image generation on high-resolution models such as SDXL and FLUX.
๐ Benefits of Our Decoupled Multi-Instance Generation:
1. Enhanced Control: By constructing scenes using depth maps in the first stage, the model focuses on coarse-grained scene layout, improving control over instance placement.
2. Flexibility & Preservation: The second stage employs training-free rendering methods, allowing seamless integration with various models (e.g., fine-tuned weights, LoRA) while maintaining the generative capabilities of pre-trained models.
Join us in advancing Layout-to-Image Generation! Follow and star our repository to stay updated! โญ
๐จ Daily Paper: 3DIS-FLUX: simple and efficient multi-instance generation with DiT rendering (2501.05131)
๐ Code is now open source!
๐ Project Website: https://limuloo.github.io/3DIS/
๐ GitHub Repository: https://github.com/limuloo/3DIS
๐ 3DIS Paper: https://arxiv.org/abs/2410.12669
๐ 3DIS-FLUX Tech Report: https://arxiv.org/abs/2501.05131
๐ฅ Why 3DIS & 3DIS-FLUX?
Current SOTA multi-instance generation methods are typically adapter-based, requiring additional control modules trained on pre-trained models for layout and instance attribute control. However, with the emergence of more powerful models like FLUX and SD3.5, these methods demand constant retraining and extensive resources.
โจ Our Solution: 3DIS
We introduce a decoupled approach that only requires training a low-resolution Layout-to-Depth model to convert layouts into coarse-grained scene depth maps. Leveraging community and company pre-trained models like ControlNet + SAM2, we enable training-free controllable image generation on high-resolution models such as SDXL and FLUX.
๐ Benefits of Our Decoupled Multi-Instance Generation:
1. Enhanced Control: By constructing scenes using depth maps in the first stage, the model focuses on coarse-grained scene layout, improving control over instance placement.
2. Flexibility & Preservation: The second stage employs training-free rendering methods, allowing seamless integration with various models (e.g., fine-tuned weights, LoRA) while maintaining the generative capabilities of pre-trained models.
Join us in advancing Layout-to-Image Generation! Follow and star our repository to stay updated! โญ
liked
a
model
3 days ago
commented
a
paper
3 days ago
upvoted
a
paper
3 days ago
authored
2
papers
3 days ago
upvoted
a
paper
3 days ago
liked
a
Space
24 days ago
XD
reacted to
Flowerfan's
post with ๐คโค๏ธ๐ค๐
10 months ago
Post
Multi-Instance Generation Controller: Enjoy complete control over position generation, attribute determination, and count!
code link: https://github.com/limuloo/MIGC
project page: https://migcproject.github.io/
MIGC decouples multi-instance generation into individual single-instance generation subtasks within the cross-attention layer of Stable Diffusion.
Welcome to follow our project and use the code to create anything you imagine!
Please let us know if you have any suggestions!
code link: https://github.com/limuloo/MIGC
project page: https://migcproject.github.io/
MIGC decouples multi-instance generation into individual single-instance generation subtasks within the cross-attention layer of Stable Diffusion.
Welcome to follow our project and use the code to create anything you imagine!
Please let us know if you have any suggestions!
- 6 replies
updated
a
Space
over 1 year ago