Update app.py

app.py

@@ -15,7 +15,7 @@ def format_prompt(message, history):
   return prompt
 
 def generate(
-    prompt, system_prompt, history, max_new_tokens, repetition_penalty, temperature, top_p, top_k, seed
+    prompt, history, system_prompt, temperature=0.9, max_new_tokens=256, top_p=0.95, repetition_penalty=1.0,
 ):
     temperature = float(temperature)
     if temperature < 1e-2:
@@ -26,10 +26,9 @@ def generate(
         temperature=temperature,
         max_new_tokens=max_new_tokens,
         top_p=top_p,
-        top_k=top_k,
         repetition_penalty=repetition_penalty,
         do_sample=True,
-        seed=seed,
+        seed=42,
     )
 
     formatted_prompt = format_prompt(f"{system_prompt}, {prompt}", history)
@@ -45,20 +44,12 @@ def generate(
 additional_inputs=[
     gr.Textbox(
         label="System Prompt",
+        max_lines=1,
         interactive=True,
     ),
-    gr.Slider(
-        label="Max new tokens",
-        value=1000,
-        minimum=100,
-        maximum=32768,
-        step=64,
-        interactive=True,
-        info="The maximum numbers of new tokens, controls how long is the output",
-    ),
     gr.Slider(
         label="Temperature",
-        value=0.7,
+        value=0.9,
         minimum=0.0,
         maximum=1.0,
         step=0.05,
@@ -66,13 +57,13 @@ additional_inputs=[
         info="Higher values produce more diverse outputs",
     ),
     gr.Slider(
-        label="Repetition penalty",
-        value=1.2,
-        minimum=1.0,
-        maximum=2.0,
-        step=0.05,
+        label="Max new tokens",
+        value=256,
+        minimum=0,
+        maximum=1048,
+        step=64,
         interactive=True,
-        info="Penalize repeated tokens, making the AI repeat less itself",
+        info="The maximum numbers of new tokens",
     ),
     gr.Slider(
         label="Top-p (nucleus sampling)",
@@ -84,19 +75,13 @@ additional_inputs=[
         info="Higher values sample more low-probability tokens",
     ),
     gr.Slider(
-        label="Top-k",
-        value=1,
-        minimum=0,
-        maximum=100,
-        step=1,
+        label="Repetition penalty",
+        value=1.2,
+        minimum=1.0,
+        maximum=2.0,
+        step=0.05,
         interactive=True,
-        info="Higher k means more diverse outputs by considering a range of tokens",
-    ),
-    gr.Number(
-        label="Seed",
-        value=42,
-        minimum=1,
-        info="A starting point to initiate the generation process",
+        info="Penalize repeated tokens",
     )
 ]
 
@@ -112,8 +97,7 @@ gr.ChatInterface(
     fn=generate,
     chatbot=gr.Chatbot(show_label=False, show_share_button=False, show_copy_button=True, likeable=True, layout="panel"),
     additional_inputs=additional_inputs,
-    title="Mixtral-8x7B-Instruct-v0.1",
-    description="If you get an erorr, you putted a too much high Max_New_Tokens or your system prompt+prompt is too long, shorten up one of these",
+    title="Mixtral 46.7B",
     examples=examples,
     concurrency_limit=20,
 ).launch(show_api=False)