s

app.py

@@ -1,33 +1,80 @@
 import gradio as gr
-from transformers import pipeline
-
-# Load a text generation pipeline with Flan-T5-small model
-llm_pipeline = pipeline(task="text-generation", model="google/flan-t5-small")
-
-# Function to generate text based on user input, with a system instruction
-def generate_text(prompt):
-    system_instruction = "Respond like a helpful AI: "  # System instruction
-    full_prompt = system_instruction + prompt  # Combine the instruction with user input
-    
-    result = llm_pipeline(
-        full_prompt, 
-        max_length=50,  
-        num_return_sequences=1,  
-        temperature=0.7,  
-        top_k=50,  
-        top_p=0.9  
-    )
-    return result[0]['generated_text']
-
-# Define the Gradio interface
-with gr.Blocks() as gradio_app:
-    gr.Markdown("## Text Generation with Flan-T5-small (Helpful AI Mode)")
-    prompt = gr.Textbox(label="Enter your prompt", placeholder="Type something here...")
-    output = gr.Textbox(label="Generated text")
-    submit_btn = gr.Button("Generate")
-
-    # Set up the button to trigger text generation
-    submit_btn.click(fn=generate_text, inputs=prompt, outputs=output)
-
-# Launch the app
-gradio_app.launch()
+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: str,
+    history: list[tuple[str, str]],
+    system_message: str,
+    max_tokens: int,
+    temperature: float,
+    top_p: float,
+):
+    """
+    Generates a response using the Hugging Face Inference API.
+
+    Args:
+        message (str): User's current input.
+        history (list[tuple[str, str]]): Previous conversation history.
+        system_message (str): Instructions for the model (e.g., persona details).
+        max_tokens (int): Maximum tokens allowed for the response.
+        temperature (float): Sampling temperature for randomness in the output.
+        top_p (float): Top-p (nucleus) sampling parameter.
+
+    Yields:
+        str: The generated chatbot response.
+    """
+    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 = ""
+
+    try:
+        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
+    except Exception as e:
+        yield f"Error: {str(e)}"
+
+
+"""
+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()