Update README.md

README.md

@@ -15,7 +15,7 @@ tags:
 
 LlamaCare is a medical language model designed to enhance medical knowledge sharing.
 
-Based on Meta LlamaV2. We used Lora with 4 bit quantization to finetune the model on a 16G gpu.
+Based on Meta LlamaV2. We used Lora with 4 bit quantization to finetune the model on a 20G GPU.
 
 Current fine-tuning datasets contained:
 
@@ -24,5 +24,5 @@ shibing624/medical
 BI55/MedText
 
 ## Try LlamaCare
-[LlamaCare](//8248258926c1432e9753b96faf6115edv1igv1v13oo.ds.megaease.cn/)
-Try LlamaCare on https://8248258926c1432e9753b96faf6115edv1igv1v13oo.ds.megaease.cn/
+[LlamaCare](https://de004c8a79fc4412b2d82a5b4d18d98ev2lbv1v1500.ds.megaease.cn/)
+Try LlamaCare on https://de004c8a79fc4412b2d82a5b4d18d98ev2lbv1v1500.ds.megaease.cn/