Update README.md
Browse files
README.md
CHANGED
|
@@ -34,7 +34,8 @@ Building on these insights, it adopts Next-DiT as the foundation to design a new
|
|
| 34 |
The *NewBie image Exp0.1* model is trained within this newly constructed system, representing the first experimental release of the NewBie text-to-image generation framework.
|
| 35 |
#### Text Encoder
|
| 36 |
We use Gemma3-4B-IT as the primary text encoder (conditioning on its second-to-last layer token embeddings), and extract Jina CLIP v2 pooled text features that are projected and fused into the model conditioning path. (into the time/AdaLN conditioning pathway).
|
| 37 |
-
|
|
|
|
| 38 |
|
| 39 |
## 🖼️Task type
|
| 40 |
**NewBie image Exp0.1** is pretrain on a large corpus of high-quality anime data, enabling the model to generate remarkably detailed and visually striking anime style images.
|
|
|
|
| 34 |
The *NewBie image Exp0.1* model is trained within this newly constructed system, representing the first experimental release of the NewBie text-to-image generation framework.
|
| 35 |
#### Text Encoder
|
| 36 |
We use Gemma3-4B-IT as the primary text encoder (conditioning on its second-to-last layer token embeddings), and extract Jina CLIP v2 pooled text features that are projected and fused into the model conditioning path. (into the time/AdaLN conditioning pathway).
|
| 37 |
+
#### VAE
|
| 38 |
+
Use the FLUX.1-dev 16channel VAE to encode images into latents, delivering richer, smoother color rendering and finer texture detail—helping safeguard the stunning visual quality of NewBie image Exp0.1.
|
| 39 |
|
| 40 |
## 🖼️Task type
|
| 41 |
**NewBie image Exp0.1** is pretrain on a large corpus of high-quality anime data, enabling the model to generate remarkably detailed and visually striking anime style images.
|