import gradio as gr # from huggingface_hub import InferenceClient from transformers import pipeline """ 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 """ # requires space hardware update to use large models (TODO) # client = InferenceClient("mistralai/Mistral-Large-Instruct-2407") # Note change in instatiation*** text_generator = pipeline("text-generation", model="EleutherAI/gpt-neo-125M") def respond(message, history, system_message, max_tokens, temperature, top_p): # Construct the prompt with system message, history, and user input prompt = system_message + "\n" + "\n".join([f"User: {msg[0]}\nAssistant: {msg[1]}" for msg in history if msg[0] and msg[1]]) prompt += f"\nUser: {message}\nAssistant:" # Generate a response using the model response = text_generator(prompt, max_length=max_tokens, temperature=temperature, top_p=top_p, do_sample=True) # Extract the assistant's response part (after "Assistant:") assistant_response = response.split("Assistant:", 1)[-1].strip() return assistant_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 marketing-minded content writer for Plan.com (a UK telecommunications company). You will be provided a bullet-point list of guidelines from which to generate an article to be published in the company News section of the website. Please follow these guidelines: - Always speak using British English expressions, syntax, and spelling. - Make the articles engaging and fun, but also professional and informative. To provide relevant contextual information about the company, please source information from the following websites: - https://plan.com/our-story - https://plan.com/products-services - https://plan.com/features/productivity-and-performance - https://plan.com/features/security-and-connectivity - https://plan.com/features/connectivity-and-cost """, label="System message"), gr.Slider(minimum=1, maximum=4096, 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()