Update app.py

app.py

@@ -16,11 +16,17 @@ def format_prompt(message, history):
     prompt += f"[INST] {message} [/INST]"
     return prompt
 
-def compress_history(history,temperature=0.9, max_new_tokens=256, top_p=0.95, repetition_penalty=1.0):
+
+temperature=0.9
+top_p=0.95
+repetition_penalty=1.0
+
+
+def compress_history(history):
     formatted_prompt=f"{COMPRESS_HISTORY.format(history=history)}"
     generate_kwargs = dict(
         temperature=temperature,
-        max_new_tokens=max_new_tokens,
+        max_new_tokens=1024,
         top_p=top_p,
         repetition_penalty=repetition_penalty,
         do_sample=True,
@@ -35,9 +41,7 @@ def compress_history(history,temperature=0.9, max_new_tokens=256, top_p=0.95, re
 
 MAX_HISTORY=100
 
-def generate(
-    prompt, history, system_prompt, temperature=0.9, max_new_tokens=256, top_p=0.95, repetition_penalty=1.0,
-):
+def generate(prompt, history,max_new_tokens):
     temperature = float(temperature)
     if temperature < 1e-2:
         temperature = 1e-2
@@ -60,14 +64,14 @@ def generate(
             cnt+=len(l.split("\n"))
     print(f'cnt:: {cnt}')
     if cnt > MAX_HISTORY:
-        history = compress_history(history, temperature, max_new_tokens, top_p, repetition_penalty)
+        history = compress_history(str(history), temperature, max_new_tokens, top_p, repetition_penalty)
     formatted_prompt = format_prompt(f"{GAME_MASTER.format(history=history)}, {prompt}", history)
     stream = client.text_generation(formatted_prompt, **generate_kwargs, stream=True, details=True, return_full_text=False)
     output = ""
 
     for response in stream:
         output += response.token.text
-        yield output
+        yield "",[(prompt,output)]
 
     lines = output.strip().strip("\n").split("\n")
     #history=""
@@ -82,7 +86,7 @@ def generate(
             print(line)
         if line.startswith("5. "):
             print(line)            
-    return output
+    return "",[(prompt,output)]
 
 
 additional_inputs=[
@@ -129,6 +133,36 @@ additional_inputs=[
     )
 ]
 
+
+
+with gr.Blocks() as app:
+    gr.HTML("""<center><h1>Mixtral 8x7B RPG</h1><h3>Role Playing Game Master</h3>""")
+    chatbot = gr.Chatbot(label="Mixtral 8x7B Chatbot",show_copy_button=True)
+    with gr.Row():
+        with gr.Column(scale=3):
+            prompt=gr.Textbox(label = "Prompt")
+        with gr.Column(scale=1):
+            button=gr.Button()
+        
+        #models_dd=gr.Dropdown(choices=[m for m in return_list],interactive=True)
+    with gr.Row():
+        stop_button=gr.Button("Stop")
+        clear_btn = gr.Button("Clear")
+    with gr.Row():
+        tokens=additional_inputs[2]
+    json_out=gr.JSON()
+    e_box=gr.Textbox()
+    #text=gr.JSON()
+    #inp_query.change(search_models,inp_query,models_dd)
+    #test_b=test_btn.click(itt,url,e_box)
+    clear_btn.click(clear_fn,None,[prompt,chatbot])
+    go=button.click(generate,[prompt,chatbot,tokens],[prompt,chatbot])
+    stop_button.click(None,None,None,cancels=[go])
+app.launch(show_api=False) 
+
+
+
+'''
 examples=[["Start the Game", None, None, None, None, None, ],
           ["Start a Game based in the year 1322", None, None, None, None, None,],
          ]
@@ -140,4 +174,7 @@ gr.ChatInterface(
     title="Mixtral RPG Game Master",
     examples=examples,
     concurrency_limit=20,
-).launch(share=True,show_api=True)
+).launch(share=True,show_api=True)
+'''
+
+