Update README.md
Browse files
README.md
CHANGED
@@ -9,7 +9,10 @@ base_model:
|
|
9 |
- ByteDance-Seed/BAGEL-7B-MoT
|
10 |
pipeline_tag: any-to-any
|
11 |
---
|
12 |
-
#
|
|
|
|
|
|
|
13 |
|
14 |
This repository hosts the model weights for **BAGEL-ReAlign**. We fine-tuned BAGEL on 6 80GB NVIDIA A800 for only 27 GPU hours. While the understanding capability remains unchanged, our ReAlign method brings +3.6 **zero-shot improvement** on GenEval , +1.26 on DPGBench, +0.37 on ImgEdit and +0.33 on GEdit.
|
15 |
|
@@ -17,8 +20,6 @@ For installation, usage instructions, and further documentation, please visit BA
|
|
17 |
|
18 |
## 🧠 Method
|
19 |
|
20 |
-
We present **ReAlign**, a self-supervised training framework that aligns understanding and generation in modest compute.
|
21 |
-
|
22 |
Coming soon! Stay tuned~
|
23 |
|
24 |
## 📊 Benchmarks
|
|
|
9 |
- ByteDance-Seed/BAGEL-7B-MoT
|
10 |
pipeline_tag: any-to-any
|
11 |
---
|
12 |
+
# BAGEL-ReAlign
|
13 |
+
|
14 |
+
> A self-supervised training framework that aligns understanding and generation in modest compute, with huge **zero-shot** gain on generation and editing capability.
|
15 |
+
|
16 |
|
17 |
This repository hosts the model weights for **BAGEL-ReAlign**. We fine-tuned BAGEL on 6 80GB NVIDIA A800 for only 27 GPU hours. While the understanding capability remains unchanged, our ReAlign method brings +3.6 **zero-shot improvement** on GenEval , +1.26 on DPGBench, +0.37 on ImgEdit and +0.33 on GEdit.
|
18 |
|
|
|
20 |
|
21 |
## 🧠 Method
|
22 |
|
|
|
|
|
23 |
Coming soon! Stay tuned~
|
24 |
|
25 |
## 📊 Benchmarks
|