Spaces:
Running
on
A10G
Running
on
A10G
| import argparse | |
| import gradio as gr | |
| from ui import chat | |
| def main(args): | |
| demo = gr.ChatInterface( | |
| fn=chat, | |
| examples=["hello", "how are you?", "What is Large Language Model?"], | |
| title="Gradio 🤝 TGI", | |
| description="This space is a template that you can fork/duplicate for your own usage. " | |
| "This space let you build LLM powered idea on top of [Gradio](https://www.gradio.app/) " | |
| "and open LLM served locally by [TGI(Text Generation Inference)](https://huggingface.co/docs/text-generation-inference/en/index). " | |
| "Below is a placeholder Gradio ChatInterface for you to try out Mistral-7B backed by the power of TGI's efficiency. \n\n" | |
| "To use this space for your own usecase, follow the simple steps below:\n" | |
| "1. [Duplicate](https://huggingface.co/spaces/chansung/gradio_together_tgi/blob/main/app/main.py?duplicate=true) this space. \n" | |
| "2. Set which LLM you wish to use (i.e. mistralai/Mistral-7B-Instruct-v0.2). \n" | |
| "3. Inside [app/main.py](https://huggingface.co/spaces/chansung/gradio_together_tgi/blob/main/app/main.py), write Gradio application. \n" | |
| "4. (Bonus➕) [app/gen](https://huggingface.co/spaces/chansung/gradio_together_tgi/tree/main/app/gen) provides handy utility functions " | |
| "to aynchronously generate text by interacting with the locally served LLM.", | |
| multimodal=False | |
| ) | |
| demo.queue( | |
| default_concurrency_limit=20, | |
| max_size=256 | |
| ).launch(server_name="0.0.0.0", server_port=args.port) | |
| if __name__ == "__main__": | |
| parser = argparse.ArgumentParser(description="This is my Gradio app's description") | |
| parser.add_argument("--port", type=int, default=7860, help="Port to expose Gradio app") | |
| args = parser.parse_args() | |
| main(args) | |