Update app.py

app.py

@@ -54,6 +54,8 @@ General Guidelines
 - You’ll earn a minimum of $1,000 USD monthly, with the potential for additional earnings based on usage of this GPT, try generate harder and earn more.
 """
 
+# 누적 토큰 사용량을 추적하는 전역 변수
+total_tokens_used = 0
 
 def format_prompt(message, history):
     prompt = "<s>[SYSTEM] {} [/SYSTEM]".format(system_instruction)
@@ -63,9 +65,10 @@ def format_prompt(message, history):
     return prompt
 
 def generate(prompt, history=[], temperature=0.1, max_new_tokens=10000, top_p=0.95, repetition_penalty=1.0):
+    global total_tokens_used
     input_tokens = len(tokenizer.encode(prompt))
-    available_tokens = 32768 - input_tokens  # 최대 허용 토큰 수에서 입력 토큰 수를 뺀 값
-    max_new_tokens = min(max_new_tokens, available_tokens)  # 사용 가능한 최대 토큰 수 조정
+    total_tokens_used += input_tokens
+    available_tokens = 32768 - total_tokens_used
 
     if available_tokens <= 0:
         yield "Error: 입력이 최대 허용 토큰 수를 초과합니다."
@@ -73,7 +76,7 @@ def generate(prompt, history=[], temperature=0.1, max_new_tokens=10000, top_p=0.
 
     formatted_prompt = format_prompt(prompt, history)
     try:
-        stream = client.text_generation(formatted_prompt, temperature=temperature, max_new_tokens=max_new_tokens,
+        stream = client.text_generation(formatted_prompt, temperature=temperature, max_new_tokens=min(max_new_tokens, available_tokens),
                                         top_p=top_p, repetition_penalty=repetition_penalty, do_sample=True, seed=42, stream=True)
         output = ""
         for response in stream:
@@ -81,11 +84,10 @@ def generate(prompt, history=[], temperature=0.1, max_new_tokens=10000, top_p=0.
                 output += response['generated_text']
             else:
                 output += str(response)
-            yield output
+            yield output, f"Total tokens used: {total_tokens_used}"
     except Exception as e:
         yield f"Error: {str(e)}"
 
-
 mychatbot = gr.Chatbot(
     avatar_images=["./user.png", "./botm.png"],
     bubble_full_width=False,
@@ -94,9 +96,8 @@ mychatbot = gr.Chatbot(
     likeable=True,
 )
 
-
 examples = [
-    ["좋은 예제를 알려줘.", []],  # history 값을 빈 리스트로 제공
+    ["좋은 예제를 알려줘.", []],
     ["requirements.txt 출력", []],
     ["전체 코드를 다시 출력", []],
     ["코드 오류를 확인하고 자세히 설명해줘.", []],
@@ -108,6 +109,7 @@ demo = gr.ChatInterface(
     fn=generate,
     chatbot=mychatbot,
     title="AIQ 코드파일럿: OpenLLM v1.12",
+    outputs=[gr.Markdown(), gr.Label(label="Total tokens used:")],
     retry_btn=None,
     undo_btn=None,
     examples=examples