HiDream.ai Releases 17B Open Image Editing Model
The new MIT-licensed model, HiDream-E1.1, allows for complex image modifications by following natural language instructions.
Startup HiDream.ai has released HiDream-E1.1, an open-source model designed for instruction-based image editing. The model allows users to modify existing images by providing simple text commands, offering a powerful new tool for developers and creative professionals working with generative AI.
What sets HiDream-E1.1 apart is its scale. At 17 billion parameters, it is a significantly larger model than many open-source alternatives in the image editing space. This substantial size suggests a capacity for understanding more nuanced and complex instructions, potentially leading to more precise and sophisticated edits than smaller models can achieve.
According to the release notes, the model is an evolution of HiDream-I1, the company's foundational text-to-image model. By fine-tuning this base for editing tasks, the new release specializes in modifying images rather than generating them from scratch.
The model is available under a permissive MIT license, encouraging broad adoption for both research and commercial applications. Developers can explore and download the HiDream-E1.1 model on the Hugging Face Hub to begin integrating it into their own projects.
Sources
- Visit
HiDream-ai/HiDream-E1-1
Hugging Face
0 comments
No comments yet. Be the first to weigh in.
More in Image Editing
ByteDance Releases Lance, a Unified Generative AI Model
The 3-billion-parameter model handles image and video generation, editing, and understanding from a single set of weights under a permissive license.
SenseTime Releases 8B 'Any-to-Any' Infographic Model
The new 8B-parameter SenseNova U1 model from SenseTime is designed for complex multimodal tasks, including the in-conversation generation and editing of infographics.
NVIDIA Releases PiD for High-Quality Image Upscaling
The new component is a specialized VAE decoder that works with Stability AI's Z-Image model to enhance super-resolution tasks.