Update app.py

app.py

@@ -44,58 +44,23 @@ examples = [
     ["You have $100, and your goal is to turn that into as much money as possible with AI and Machine Learning. Please respond with detailed plan.", "", 300, 1.2, 0.5, 0.5, 0.5],
 ]
 
-def respond(
-    instruction,
-    input=None,
-    token_count=333,
-    temperature=1.0,
-    top_p=0.5,
-    presencePenalty = 0.5,
-    countPenalty = 0.5,
-    history=None
-):
-    args = PIPELINE_ARGS(temperature = max(0.2, float(temperature)), top_p = float(top_p),
-                     alpha_frequency = countPenalty,
-                     alpha_presence = presencePenalty,
-                     token_ban = [], # ban the generation of some tokens
-                     token_stop = [0]) # stop generation whenever you see any token here
-
-    instruction = re.sub(r'\n{2,}', '\n', instruction).strip().replace('\r\n','\n')
-    ctx = generate_prompt(instruction, input, history)
-    print(ctx + "\n")
+def respond(history=None):
+    global token_count_chat, temperature_chat, top_p_chat, presence_penalty_chat, count_penalty_chat
     
-    all_tokens = []
-    out_last = 0
-    out_str = ''
-    occurrence = {}
-    state = None
-    for i in range(int(token_count)):
-        out, state = model.forward(pipeline.encode(ctx)[-ctx_limit:] if i == 0 else [token], state)
-        for n in occurrence:
-            out[n] -= (args.alpha_presence + occurrence[n] * args.alpha_frequency)
+    # get the lastest user message and the additional parameters
+    instruction = msg.value
+    token_count = token_count_chat.value
 
-        token = pipeline.sample_logits(out, temperature=args.temperature, top_p=args.top_p)
-        if token in args.token_stop:
-            break
-        all_tokens += [token]
-        for xxx in occurrence:
-            occurrence[xxx] *= 0.996
-        if token not in occurrence:
-            occurrence[token] = 1
-        else:
-            occurrence[token] += 1
-        
-        tmp = pipeline.decode(all_tokens[out_last:])
-        if '\ufffd' not in tmp:
-            out_str += tmp
-            out_last = i + 1
-        if '\n\n' in out_str:
-            break
+    temperature = temperature_chat.value
+    top_p = top_p_chat.value
+    presence_penalty = presence_penalty_chat.value
+    count_penalty = count_penalty_chat.value
 
-    del out
-    del state
-    gc.collect()
-    return out_str.strip()
+    history[-1][1] = ""
+    
+    for character in generator(instruction, None, token_count, temperature, top_p, presence_penalty, count_penalty, history):
+        history[-1][1] += character
+        yield history
 
 def generator(
     instruction,
@@ -188,25 +153,8 @@ with gr.Blocks(title=title) as demo:
                 history = history or []
                 return "", history + [[message, None]]
     
-            def chat(history):
-                print("History: ", history)
-                global token_count_chat, temperature_chat, top_p_chat, presence_penalty_chat, count_penalty_chat
-                # get the last user message and the additional parameters
-                instruction = msg.value
-                token_count = token_count_chat.value
-    
-                temperature = temperature_chat.value
-                top_p = top_p_chat.value
-                presence_penalty = presence_penalty_chat.value
-                count_penalty = count_penalty_chat.value
-    
-                response = respond(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"
+                respond, chatbot, chatbot, api_name="chat"
             )
             
     with gr.Tab("Instruct mode"):