Sana-1.6B
Linear Diffusion Transformer for High Resolution Image Synthesis
0
- -High resolution image synthesis: capable of generating high-resolution images up to 4096px.
- -Linear Diffusion Transformer: Using linear transformation to improve the efficiency and quality of image synthesis.
- -Multi GPU support: Supports parallel running on 8 GTX3090 GPUs to improve processing speed.
- -Multiple sampling steps to choose from: Users can select 5 to 40 sampling steps as needed.
- -CFG and PAG guided scale adjustment: Provides 1 to 10 and 1 to 4 CFG and PAG guided scale adjustment options.
- -Negative prompt usage: Optimize image generation results by using negative prompts.
- -Image style selection: Provides multiple image style options, such as movies, photography, anime, etc.
- -Random Seed Generation: Allow users to randomize seeds to generate different image results.
Product Details
Sana-1.6B is an efficient high-resolution image synthesis model based on linear diffusion transformer technology, capable of generating high-quality images. This model was developed by NVIDIA Labs using DC-AE technology, with 32 times the potential space and the ability to run on multiple GPUs, providing powerful image generation capabilities. Sana-1.6B is renowned for its efficient image synthesis capability and high-quality output results, making it an important technology in the field of image synthesis.
