Update README.md

README.md

@@ -16,7 +16,7 @@ license: apache-2.0
 # Tifa-DeepSexV2-7b-MGRPO
 
 - **原始模型**：[Qwen2.5-7B](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct)
-- **GGUF**: [F16](https://huggingface.co/ValueFX9507/Tifa-DeepsexV2-7b-MGRPO-GGUF-F16) | [Q8](https://huggingface.co/ValueFX9507/Tifa-DeepsexV2-7b-MGRPO-GGUF-Q8) | [Q4](https://huggingface.co/ValueFX9507/Tifa-DeepsexV2-7b-MGRPO-GGUF-Q4) （Q4损失非常大，极力不推荐，推荐Q8以上）
+- **GGUF**: [F16](https://huggingface.co/ValueFX9507/Tifa-DeepsexV2-7b-MGRPO-GGUF-F16) | [Q8](https://huggingface.co/ValueFX9507/Tifa-DeepsexV2-7b-MGRPO-GGUF-Q8) | [Q4](https://huggingface.co/ValueFX9507/Tifa-DeepsexV2-7b-MGRPO-GGUF-Q4) （Q4损失较大，推荐Q8以上）
 - **Demo APK**: [点击下载](http://app.visionsic.com/download/projectchat.apk)
 - **简单的前端**：[Github链接](https://github.com/Value99/Tifa-Deepsex-OllamaWebUI)
 - **必看教程**：[BiliBili视频教程](https://space.bilibili.com/483051186)
@@ -112,7 +112,9 @@ Tifa-DeepsexV2-7b-MGRPO-0218，大幅减少发散、提高推理能力、惩罚
 
 ## 模型交流群
 
-![image/png](https://cdn-uploads.huggingface.co/production/uploads/650762d0eac45ee2e420a38b/3lw3Q4FLNHNAt7Rt0hTUI.png)
+
+![image/png](https://cdn-uploads.huggingface.co/production/uploads/650762d0eac45ee2e420a38b/5U2kUEEbYPahEe048xapG.png)
+
 
 
 ## 训练过程输出