Update README.md
Browse files- release tech report
README.md
CHANGED
|
@@ -13,7 +13,8 @@ Valley [github](https://github.com/bytedance/Valley) is a cutting-edge multimoda
|
|
| 13 |
when evaluated against models of the same scale.
|
| 14 |
|
| 15 |
## Release
|
| 16 |
-
- [
|
|
|
|
| 17 |
|
| 18 |
## Valley-Eagle
|
| 19 |
The foundational version of Valley is a multimodal large model aligned with Siglip and Qwen2.5, incorporating LargeMLP and ConvAdapter to construct the projector.
|
|
|
|
| 13 |
when evaluated against models of the same scale.
|
| 14 |
|
| 15 |
## Release
|
| 16 |
+
- [01/13] 🔥 Release TechReport. [Valley2: Exploring Multimodal Models with Scalable Vision-Language Design](https://arxiv.org/abs/2501.05901)
|
| 17 |
+
- [12/23] Announcing [Valley-Qwen2.5-7B](https://huggingface.co/ByteDance)!
|
| 18 |
|
| 19 |
## Valley-Eagle
|
| 20 |
The foundational version of Valley is a multimodal large model aligned with Siglip and Qwen2.5, incorporating LargeMLP and ConvAdapter to construct the projector.
|