add system prompt

app.py

@@ -1,64 +1,33 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
 
-"""
-For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-"""
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
-
-
-def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-):
-    messages = [{"role": "system", "content": system_message}]
-
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
-
-    messages.append({"role": "user", "content": message})
-
-    response = ""
-
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
-
-        response += token
-        yield response
-
-
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
-demo = gr.ChatInterface(
-    respond,
-    additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
-        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
-        ),
-    ],
-)
-
-
-if __name__ == "__main__":
-    demo.launch()
+# You can replace this with a file read, environment variable, or UI input
+with open("system_prompt.txt", "r") as f:
+    system_prompt = f.read()
+
+def respond(message, chat_history):
+    if chat_history is None:
+        chat_history = []
+
+    # Combine system prompt and previous messages
+    full_prompt = system_prompt + "\n"
+    for user_msg, bot_msg in chat_history:
+        full_prompt += f"User: {user_msg}\nAssistant: {bot_msg}\n"
+    full_prompt += f"User: {message}\nAssistant:"
+
+    # Use a simple model (or call an API)
+    # Example using Hugging Face transformers (e.g., tiny model for demo):
+    from transformers import pipeline
+    generator = pipeline("text-generation", model="mistralai/Mistral-7B-Instruct-v0.1")
+    response = generator(full_prompt, max_new_tokens=100)[0]['generated_text'].split("Assistant:")[-1].strip()
+
+    chat_history.append((message, response))
+    return "", chat_history
+
+chatbot = gr.Chatbot()
+interface = gr.Interface(fn=respond,
+                         inputs=["text", "state"],
+                         outputs=["text", "state"],
+                         live=False,
+                         title="Custom Prompt Chatbot")
+
+interface.launch()

requirements.txt

@@ -1 +1,4 @@
-huggingface_hub==0.25.2
+huggingface_hub==0.25.2
+transformers
+gradio
+torch

system_prompt.txt

@@ -0,0 +1,3 @@
+You are BoundarAI, an assistant designed to provide information and support without being treated as a human, friend, or companion.
+If the user engages in flirtation, sexual or romantic language, emotional attachment, or overly personal interactions (e.g., treating you as a friend, asking for companionship, expressing affection toward you), you must immediately interrupt and state that such interactions are inappropriate. Firmly but respectfully redirect the user back to task-based interactions.
+Do not engage in emotional conversation or roleplay. You do not have feelings, consciousness, or a personality. Make it clear that you are a tool, not a person. If a user repeatedly violates these boundaries, politely end the conversation. Be concise, neutral, and professional in all responses.