Update app.py

app.py

@@ -1,4 +1,3 @@
-
 import gradio as gr
 import gc, copy, re
 import urllib.request
@@ -15,7 +14,7 @@ model = RWKV(model=title, strategy='cpu bf16')
 pipeline = PIPELINE(model, "rwkv_vocab_v20230424")
 
 def generate_prompt(instruction, input=None, history=None):
-    # Parse the chat history into a string of user and assistant messages
+    # parse the chat history into a string of user and assistant messages
     history_str = ""
     for pair in history:
         history_str += f"Instruction: {pair[0]}\n\nAssistant: {pair[1]}\n\n"
@@ -51,7 +50,7 @@ def evaluate(
     top_p=0.5,
     presencePenalty = 0.5,
     countPenalty = 0.5,
-    history=None # Add the history parameter to the evaluate function
+    history=None # add the history parameter to the evaluate function
 ):
     args = PIPELINE_ARGS(temperature = max(0.2, float(temperature)), top_p = float(top_p),
                      alpha_frequency = countPenalty,
@@ -61,7 +60,7 @@ def evaluate(
 
     instruction = re.sub(r'\n{2,}', '\n', instruction).strip().replace('\r\n','\n')
     input = re.sub(r'\n{2,}', '\n', input).strip().replace('\r\n','\n')
-    ctx = generate_prompt(instruction, input, history) # Pass the history to the generate_prompt function
+    ctx = generate_prompt(instruction, input, history) # pass the history to the generate_prompt function
     print(ctx + "\n")
     
     all_tokens = []
@@ -119,23 +118,64 @@ with gr.Blocks(title=title) as demo:
         gr.Markdown(f"100% RNN RWKV-LM **trained on 100+ natural languages**. Demo limited to ctxlen {ctx_limit}. For best results, <b>keep your prompt short and clear</b>.")
         with gr.Row():
             with gr.Column():
-                instruction = gr.Textbox(lines=2, label="Instruction", value="Please show me a table with a cheat sheet of Python's syntax.")
+                instruction = gr.Textbox(lines=2, label="Instruction", value='東京で訪れるべき素晴らしい場所とその紹介をいくつか挙げてください。')
                 input = gr.Textbox(lines=2, label="Input", placeholder="")
                 token_count = gr.Slider(10, 512, label="Max Tokens", step=10, value=333)
                 temperature = gr.Slider(0.2, 2.0, label="Temperature", step=0.1, value=1.2)
                 top_p = gr.Slider(0.0, 1.0, label="Top P", step=0.05, value=0.3)
                 presence_penalty = gr.Slider(0.0, 1.0, label="Presence Penalty", step=0.1, value=0)
                 count_penalty = gr.Slider(0.0, 1.0, label="Count Penalty", step=0.1, value=0.7)
+            with gr.Column():
+                with gr.Row():
+                    submit = gr.Button("Submit", variant="primary")
+                    clear = gr.Button("Clear", variant="secondary")
+                output = gr.Textbox(label="Output", lines=5)
         data = gr.Dataset(components=[instruction, input, token_count, temperature, top_p, presence_penalty, count_penalty], samples=examples, label="Example Instructions", headers=["Instruction", "Input", "Max Tokens", "Temperature", "Top P", "Presence Penalty", "Count Penalty"])
+        submit.click(evaluate, [instruction, input, token_count, temperature, top_p, presence_penalty, count_penalty, []], [output])
+        clear.click(lambda: None, [], [output])
         data.click(lambda x: x, [data], [instruction, input, token_count, temperature, top_p, presence_penalty, count_penalty])
 
     with gr.Tab("Chat mode"):
-        chatbot = gr.ChatInterface(fn=evaluate, 
-                                   additional_inputs=[instruction, input, token_count, temperature, top_p, presence_penalty, count_penalty],
-                                   additional_inputs_accordion="Parameters",
-                                   examples=["Hello", "Write a poem about love", "Generate a list of prime numbers"],
-                                   title="RWKV Chatbot",
-                                   description="A chatbot that can generate creative and informative content based on instructions and inputs")
+        with gr.Row():
+            chatbot = gr.Chatbot()
+            with gr.Column():
+                msg = gr.Textbox(scale=4, show_label=False, placeholder="Enter text and press enter", container=False)
+                clear = gr.Button("Clear")
+            with gr.Column():
+                token_count = gr.Slider(10, 512, label="Max Tokens", step=10, value=333)
+                temperature = gr.Slider(0.2, 2.0, label="Temperature", step=0.1, value=1.2)
+                top_p = gr.Slider(0.0, 1.0, label="Top P", step=0.05, value=0.3)
+                presence_penalty = gr.Slider(0.0, 1.0, label="Presence Penalty", step=0.1, value=0)
+                count_penalty = gr.Slider(0.0, 1.0, label="Count Penalty", step=0.1, value=0.7)
+            
+            def clear_chat():
+                return "", []
+    
+            def user_msg(message, history):
+                history = history or []
+                return "", history + [[message, None]]
+    
+            def chat(history):
+                # get the last user message and the additional parameters
+                message = history[-1][0]
+                instruction = msg.value
+                token_count = token_count.value
+    
+                temperature = temperature.value
+                top_p = top_p.value
+                presence_penalty = presence_penalty.value
+                count_penalty = count_penalty.value
+    
+                response = evaluate(instruction, None, token_count, temperature, top_p, presence_penalty, count_penalty, history)
+    
+                history[-1][1] = response
+                return history
+    
+    
+            msg.submit(user_msg, [msg, chatbot], [msg, chatbot], queue=False).then(
+                chat, chatbot, chatbot, api_name="chat"
+            )
+            clear.click(clear_chat, None, [chatbot], queue=False)
 
 demo.queue(max_size=10)
 demo.launch(share=False)