I was able to generate some great starter images, but when I try to recreate them in other tools I'm not able to.
For example, if I run
python3.9 demo.py --seed "18012747" --num-inference-steps "100" --guidance-scale "7.5" --prompt "A Photograph of lightning in a bottle" --output "/tmp/output.png"
then I get the following image,
Can anyone explain the difference? I thought both were SD 1.4. Same prompt, seed, guidance_scale, steps. I tried all the samplers available to the sd-webui versions in case that was it.
I was able to generate some great starter images, but when I try to recreate them in other tools I'm not able to. For example, if I run python3.9 demo.py --seed "18012747" --num-inference-steps "100" --guidance-scale "7.5" --prompt "A Photograph of lightning in a bottle" --output "/tmp/output.png" then I get the following image,
But when I try the same prompt and settings on this SD tool, https://github.com/sd-webui/stable-diffusion-webui , I get the following image,
Can anyone explain the difference? I thought both were SD 1.4. Same prompt, seed, guidance_scale, steps. I tried all the samplers available to the sd-webui versions in case that was it.