- feature: 🎉 tile in a single dimension ("x" or "y"). This enables, with a bit of luck, generation of 360 VR images. Try this for example: `imagine --tile-x -w 1024 -h 512 "360 degree equirectangular panorama photograph of the mountains" --upscale`
7.1.1
- fix: memory/speed regression introduced in 6.1.0 - fix: model switching now clears memory better, thus avoiding out of memory errors
7.1.0
- feature: 🎉 Stable Diffusion 2.1. Generated people are no longer (completely) distorted. Use with `--model SD-2.1` or `--model SD-2.0-v`
7.0.0
- feature: negative prompting. `--negative-prompt` or `ImaginePrompt(..., negative_prompt="ugly, deformed, extra arms, etc")` - feature: a default negative prompt is added to all generations. Images in SD-2.0 don't look bad anymore. Images in 1.5 look improved as well.
6.1.2
- fix: add back in memory-efficient algorithms
6.1.1
- feature: xformers will be used if available (for faster generation) - fix: version metadata was broken