Implementation of Text-To-Image generation using Stable Diffusion on Intel CPU.
- Linux, Windows, MacOS
- Python 3.8.+
- CPU compatible with OpenVINO.
pip install -r requirements.txtusage: stable_diffusion.py [-h] [--model MODEL] [--seed SEED] [--beta-start BETA_START] [--beta-end BETA_END] [--beta-schedule BETA_SCHEDULE] [--num-inference-steps NUM_INFERENCE_STEPS]
[--guidance-scale GUIDANCE_SCALE] [--eta ETA] [--tokenizer TOKENIZER] [--prompt PROMPT] [--output OUTPUT]
optional arguments:
-h, --help show this help message and exit
--model MODEL model name
--seed SEED random seed for generating consistent images per prompt
--beta-start BETA_START
LMSDiscreteScheduler::beta_start
--beta-end BETA_END LMSDiscreteScheduler::beta_end
--beta-schedule BETA_SCHEDULE
LMSDiscreteScheduler::beta_schedule
--num-inference-steps NUM_INFERENCE_STEPS
num inference steps
--guidance-scale GUIDANCE_SCALE
guidance scale
--eta ETA eta
--tokenizer TOKENIZER
tokenizer
--prompt PROMPT prompt
--output OUTPUT output image namepython stable_diffusion.py --prompt "Street-art painting of Emilia Clarke in style of Banksy, photorealism"| CPU | Time per iter | Total time |
|---|---|---|
| Intel(R) Core(TM) i5-8279U | 7.4 s/it | 3.59 min |
| AMD Ryzen Threadripper 1900X | 5.34 s/it | 2.58 min |
| Intel(R) Xeon(R) Gold 6154 CPU @ 3.00GHz | 1 s/it | 33 s |
| Intel(R) Core(TM) i7-1165G7 @ 2.80GHz | 7.4 s/it | 3.59 min |
- Original implementation of Stable Diffusion: https://github.com/CompVis/stable-diffusion
- diffusers library: https://github.com/huggingface/diffusers
The authors are not responsible for the content generated using this project. Please, don't use this project to produce illegal, harmful, offensive etc. content.