Trained for 4 epochs and 48700 steps.
Browse filesTrained with datasets ['text-embeds-sd2x', 'celebrities', 'movieposters', 'normalnudes', 'propagandaposters', 'guys', 'pixel-art', 'signs', 'moviecollection', 'bookcovers', 'nijijourney', 'experimental', 'ethnic', 'sports', 'gay', 'architecture', 'shutterstock', 'cinemamix-1mp', 'nsfw-1024', 'anatomy', 'bg20k-1024', 'yoga']
Learning rate 4e-07, batch size 4, and 8 gradient accumulation steps.
Used DDPM noise scheduler for training with v_prediction prediction type and rescaled_betas_zero_snr=True
Using 'trailing' timestep spacing.
Base model: stabilityai/stable-diffusion-2-1
VAE: None
README.md
CHANGED
@@ -205,7 +205,7 @@ You may reuse the base model text encoder for inference.
|
|
205 |
### anatomy
|
206 |
- Repeats: 5
|
207 |
- Total number of images: 16352
|
208 |
-
- Total number of aspect buckets:
|
209 |
- Resolution: 1.0 megapixels
|
210 |
- Cropped: True
|
211 |
- Crop style: random
|
|
|
205 |
### anatomy
|
206 |
- Repeats: 5
|
207 |
- Total number of images: 16352
|
208 |
+
- Total number of aspect buckets: 3
|
209 |
- Resolution: 1.0 megapixels
|
210 |
- Cropped: True
|
211 |
- Crop style: random
|