Update main.py

main.py

@@ -23,7 +23,6 @@ class InputData(BaseModel):
     end_token: str
     system_prompt: str
     user_input: str
-    json_prompt: str
     history: str = ""
 
 class WordCheckData(BaseModel):
@@ -34,18 +33,14 @@ class WordCheckData(BaseModel):
 async def generate_response(data: InputData) -> Dict[str, Any]:
     client = InferenceClient(model=data.model, token=HF_TOKEN)
 
-    sentences = tokenizer.tokenize(data.user_input)
+    user_sentences = tokenizer.tokenize(data.user_input)
 
-    this_is_a_string = ""
+    user_input_str = "\n".join(user_sentences)
     
-    for sentence in sentences:
-        this_is_a_string += "\n" + sentence
-
-    data.history += data.prompt_template.replace("{Prompt}", str(this_is_a_string))
+    data.history += data.prompt_template.replace("{Prompt}", user_input_str)
 
     inputs = (
         data.system_prompt_template.replace("{SystemPrompt}", data.system_prompt) +
-        data.system_prompt_template.replace("{SystemPrompt}", data.json_prompt) +
         data.history
     )
 
@@ -61,14 +56,16 @@ async def generate_response(data: InputData) -> Dict[str, Any]:
 
         response_str = str(response)
 
-        sentences = tokenizer.tokenize(response_str)
+        ai_sentences = tokenizer.tokenize(response_str)
 
         cleaned_response = {
-            "New response": sentences,
-            "Sentence count": min(len(sentences), 3)
+            "New response": ai_sentences,
+            "Sentence count": min(len(ai_sentences), 3)
         }
 
-        data.history += str(cleaned_response)
+        ai_response_str = "\n".join(ai_sentences)
+
+        data.history += ai_response_str + "\n"
 
         return {
             "response": cleaned_response,