Commit
•
7d6d2e6
1
Parent(s):
b37dba4
Update README.md
Browse files
README.md
CHANGED
|
@@ -7,11 +7,16 @@ Use the tokens **_arcane style_** in your prompts for the effect.
|
|
| 7 |
|
| 8 |
If you enjoy this model, please check out my other models on [Huggingface](https://huggingface.co/nitrosocke)
|
| 9 |
|
|
|
|
|
|
|
|
|
|
| 10 |
### Sample images from the model:
|
| 11 |
|
| 12 |
### Sample images used for training:
|
| 13 |
|
| 14 |
|
|
|
|
|
|
|
| 15 |
**Version 2** (arcane-diffusion-v2): This uses the diffusers based dreambooth training and prior-preservation loss is way more effective. The diffusers where then converted with a script to a ckpt file in order to work with automatics repo.
|
| 16 |
Training was done with 5k steps for a direct comparison to v1 and results show that it needs more steps for a more prominent result. Version 3 will be tested with 11k steps.
|
| 17 |
|
|
|
|
| 7 |
|
| 8 |
If you enjoy this model, please check out my other models on [Huggingface](https://huggingface.co/nitrosocke)
|
| 9 |
|
| 10 |
+
### Sample images from v3:
|
| 11 |
+
|
| 12 |
+
|
| 13 |
### Sample images from the model:
|
| 14 |
|
| 15 |
### Sample images used for training:
|
| 16 |
|
| 17 |
|
| 18 |
+
**Version 3** (arcane-diffusion-v3): This version uses the new _train-text-encoder_ setting and improves the quality and edibility of the model immensely. Trained on 95 images from the show in 8000 steps.
|
| 19 |
+
|
| 20 |
**Version 2** (arcane-diffusion-v2): This uses the diffusers based dreambooth training and prior-preservation loss is way more effective. The diffusers where then converted with a script to a ckpt file in order to work with automatics repo.
|
| 21 |
Training was done with 5k steps for a direct comparison to v1 and results show that it needs more steps for a more prominent result. Version 3 will be tested with 11k steps.
|
| 22 |
|