Upload folder using huggingface_hub

.gitattributes

@@ -34,3 +34,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
 tokenizer.json filter=lfs diff=lfs merge=lfs -text
+framework.png filter=lfs diff=lfs merge=lfs -text
+performance.png filter=lfs diff=lfs merge=lfs -text

README.md

@@ -22,7 +22,7 @@ This repo contains the weights from our paper: <a href="https://arxiv.org/pdf/25
 
 ## Overview
 
-<img src="https://huggingface.co/taki555/Llama3.1-8B-Shadow-FT-BAAI-2k/blob/main/framework.png" width="100%" />
+<img src="framework.png" width="100%" />
 
 Observation:
 
@@ -37,7 +37,7 @@ $\Rightarrow$ We propose the Shadow-FT framework to tune the INSTRUCT models by
 
 This repository contains the Llama-3.2-1B tuned on BAAI-2k subsets using Shadow-FT.
 
-<img src="https://huggingface.co/taki555/Llama3.1-8B-Shadow-FT-BAAI-2k/blob/main/performance.png" width="100%" />
+<img src="performance.png" width="100%" />
 
 please refer to [our paper](https://arxiv.org/pdf/2505.12716) for details.