try my copy of the space

README.md

@@ -32,7 +32,7 @@ The LLM backed has a dynamic setup so it can switch between providers, for now w
 Options for LLM backend:
 1. tencent/Hunyuan-Large (has been free for a while!)
 2. yuntian-deng/ChatGPT (say it will have short term availability)
-3. hysts/zephyr-7b (runs on zero-gpu)
+3. Fraser/zephyr-7b (runs on zero-gpu)
 4. BlinkDL/RWKV-Gradio-2 (running on T4, never ran into rate limits)
 5. make my own Cohere space (20 generations per minute)
 

src/App.svelte

@@ -107,7 +107,7 @@
       );
       
       rwkvClient = await gradioClient.Client.connect(
-        "hysts/zephyr-7b",
+        "Fraser/zephyr-7b",
         opts
       );