Textual Inversion Embedding for SD 2.x, trained for 2000 steps from v2-1_768-ema-pruned.ckpt [4bdfc29c]
Slightly overtrained on purpose, works best with rooms and buildings. Will force buildings/walls even without prompting.
Use negative prompt "low quality, blurry, meme, out of focus" to get sharper results, if you want a softer look leave this out.
Inference Providers
NEW
This model is not currently available via any of the supported Inference Providers.
The model cannot be deployed to the HF Inference API:
The model has no library tag.