4 ms·
A little pointless, considering that Stable Diffusion is already at or near human levels. Worst case scenario, just train on pre-2022 images.
by matteoraso 3y ago
A little pointless, considering that Stable Diffusion is already at or near human levels. Worst case scenario, just train on pre-2022 images.
- BitwiseFool 3y agoLow background jpegs! https://en.wikipedia.org/wiki/Low-background_steel https://en.wikipedia.org/wiki/Low-background_steel
- aussieguy1234 3y agoStable Diffusion is capable of training on its own images
- TerrifiedMouse 3y agoIs that advisable? It won’t learn anything new and might reinforce its errors.
- ranguna 3y agoIt probably won't learn anything new, but it will learn not to generate bad images if you cherry pick the best ones for training.
- raxxorraxor 3y agoNo, it isn't if you don't curate meticulously. Perhaps by accident something new could emerge which is worth it to use as input for an advanced model. But the rule is that new concepts are very hard to produce, although thanks to countless models, stable diffusion is probably the most flexible approach by quite some margin.
- micahcc 3y agoLikely they have humans review the images (or even touch up). Same thing with the dataset scraped from the internet.
- aussieguy1234 3y agoYes, that's the technique. Generate 10 images, then choose the ones that turned out well for the next round. That's the standard way to create a LoRA.