Instructions to use AiArtLab/sdxs-2b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use AiArtLab/sdxs-2b with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("AiArtLab/sdxs-2b", dtype=torch.bfloat16, device_map="cuda") prompt = "sdxs-2b" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
- Local Apps
- Draw Things
- DiffusionBee
Initial model weights
Browse files
README.md
CHANGED
|
@@ -20,7 +20,7 @@ At AiArtLab, we strive to create a free, compact and fast model that can be trai
|
|
| 20 |
- CosmosTransformer3D: 2b parameters
|
| 21 |
- Qwen3.5: 0.8b parameters
|
| 22 |
- Qwen vae: 16ch8x
|
| 23 |
-
- Resolution: Default 768x1152, trained from
|
| 24 |
- Limitations: trained on small datasets ~2.5kk, focused on art / illustrations / anime and photo
|
| 25 |
- Captions: danbooru, natural (short/medium), trained with 250 max toks
|
| 26 |
|
|
|
|
| 20 |
- CosmosTransformer3D: 2b parameters
|
| 21 |
- Qwen3.5: 0.8b parameters
|
| 22 |
- Qwen vae: 16ch8x
|
| 23 |
+
- Resolution: Default 768x1152, trained from 576 to 1152 with step 64
|
| 24 |
- Limitations: trained on small datasets ~2.5kk, focused on art / illustrations / anime and photo
|
| 25 |
- Captions: danbooru, natural (short/medium), trained with 250 max toks
|
| 26 |
|