Update app.py

app.py

@@ -20,11 +20,13 @@ from huggingface_hub import hf_hub_download
 model_id = "kawasumi/Tema_Q-R4.2-GGUF"
 model_file = "Tema_Q-R4.2-Q2_K.gguf"
 
+# 入力制限文字数
+MAX_INPUT_CHARS = 15
+
 print("Downloading model...")
 model_path = hf_hub_download(repo_id=model_id, filename=model_file)
 
 print(f"Loading model from {model_path}...")
-# メモリ節約のための設定を強化
 llm = Llama(
     model_path=model_path,
     n_ctx=1024,      
@@ -35,25 +37,35 @@ llm = Llama(
 print("Model loaded.")
 
 def chat_response(message, history):
+    # --- 文字数制限の追加 ---
+    if len(message) > MAX_INPUT_CHARS:
+        yield f"【エラー】入力が長すぎます。{MAX_INPUT_CHARS}文字以内で入力してください。（現在 {len(message)} 文字）"
+        return
+    # ----------------------
 
     prompt = f"<start_of_turn>user\n{message}<end_of_turn>\n<start_of_turn>model\n"
     
-    output = llm(
-        prompt,
-        max_tokens=512,
-        stop=["<end_of_turn>", "user"],
-        stream=True
-    )
-    
-    response = ""
-    for chunk in output:
-        text = chunk["choices"][0]["text"]
-        response += text
-        yield response
+    try:
+        output = llm(
+            prompt,
+            max_tokens=512,
+            stop=["<end_of_turn>", "user"],
+            stream=True
+        )
+        
+        response = ""
+        for chunk in output:
+            text = chunk["choices"][0]["text"]
+            response += text
+            yield response
+    except Exception as e:
+        yield f"エラーが発生しました: {str(e)}"
 
+# UIの構築（説明欄に制限の記載を追加）
 demo = gr.ChatInterface(
     fn=chat_response, 
     title="Tema_Q-R4.2 Chat",
+    description=f"※{MAX_INPUT_CHARS}文字以内で入力してください。CPU環境のため、回答開始まで時間がかかります。"
 )
 
 if __name__ == "__main__":