add model txt

app.py

@@ -4,6 +4,14 @@ from huggingface_hub import InferenceClient
 """
 For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
 """
+choices = [
+    "meta-llama/Meta-Llama-3-8B-Instruct",
+    "microsoft/Phi-3-vision-128k-instruct",
+    "Qwen/Qwen2-0.5B-Instruct",
+]
+client1 = InferenceClient(model=choices[0])
+client2 = InferenceClient(model=choices[1])
+client3 = InferenceClient(model=choices[2])
 
 
 def respond(
@@ -15,7 +23,6 @@ def respond(
     top_p,
     model,
 ):
-    client = InferenceClient(model=model)
     messages = [{"role": "system", "content": system_message}]
 
     for val in history:
@@ -28,6 +35,9 @@ def respond(
 
     response = ""
 
+    client = (
+        client1 if model == choices[0] else client2 if model == choices[1] else client3
+    )
     for message in client.chat_completion(
         messages,
         max_tokens=max_tokens,
@@ -57,7 +67,7 @@ demo = gr.ChatInterface(
             step=0.05,
             label="Top-p (nucleus sampling)",
         ),
-        gr.Textbox(value="meta-llama/Meta-Llama-3-8B-Instruct", label="Model"),
+        gr.Dropdown(choices=choices, value=choices[0], label="Model"),
     ],
     multimodal=True,
 )