Update app.py

app.py

@@ -4,7 +4,7 @@ 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
 """
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
+client = InferenceClient("model.guff")
 
 
 def respond(
@@ -14,6 +14,7 @@ def respond(
     max_tokens,
     temperature,
     top_p,
+    top_k
 ):
     messages = [{"role": "system", "content": system_message}]
 
@@ -33,6 +34,9 @@ def respond(
         stream=True,
         temperature=temperature,
         top_p=top_p,
+        top_k=top_k,
+        num_beams=10,
+        num_return_sequences=10
     ):
         token = message.choices[0].delta.content
 
@@ -47,17 +51,18 @@ demo = gr.ChatInterface(
     additional_inputs=[
         gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
         gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
+        gr.Slider(minimum=0.1, maximum=4.0, value=1.0, step=0.1, label="Temperature"),
+        gr.Slider(minimum=1, maximum=100, value=50, step=1, label="Top_k"),
         gr.Slider(
             minimum=0.1,
             maximum=1.0,
-            value=0.95,
+            value=1.0,
             step=0.05,
-            label="Top-p (nucleus sampling)",
+            label="Top_p (nucleus sampling)",
         ),
     ],
 )
 
 
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch(share=True)