import spaces
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch
import gradio as gr
title = "# 🚀👋🏻Welcome to Tonic's🤖SuperAGI/SAM🚀"
description = """SAM is an Agentic-Native LLM that **excels at complex reasoning**.
You can also use [🤖SuperAGI/SAM](https://huggingface.co./SuperAGI/SAM) by cloning this space. 🧬🔬🔍 Simply click here: 
Join us : 🌟TeamTonic🌟 is always making cool demos! Join our active builder's🛠️community 👻 [](https://discord.gg/GWpVpekp) On 🤗Huggingface: [TeamTonic](https://huggingface.co./TeamTonic) & [MultiTransformer](https://huggingface.co./MultiTransformer) On 🌐Github: [Tonic-AI](https://github.com/tonic-ai) & contribute to 🌟 [DataTonic](https://github.com/Tonic-AI/DataTonic) 🤗Big thanks to Yuvi Sharma and all the folks at huggingface for the community grant 🤗
To contribute to this space make a PR with a new example or cool new use-case for this one 🤗
"""
examples = [["[Question:] What is the proper treatment for buccal herpes?", "You are a medicine and public health expert, you will receive a question, answer the question, and provide a complete answer"]]
model_id = "SuperAGI/SAM"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype=torch.float16, device_map="auto")
@spaces.GPU
def generate_response(formatted_input):
inputs = tokenizer(formatted_input, return_tensors="pt")
inputs = {k: v.to("cuda") for k, v in inputs.items()}
# Generate a response using the model
output = model.generate(**inputs, max_length=512, pad_token_id=tokenizer.eos_token_id)
return tokenizer.decode(output[0], skip_special_tokens=True)
class ChatBot:
def __init__(self):
self.history = []
def predict(self, example_instruction, example_answer, user_input, system_prompt):
formatted_input = f" [INST] {example_instruction} [/INST] {example_answer} [INST] {system_prompt} {user_input} [/INST]"
return generate_response(formatted_input)
bot = ChatBot()
def main():
with gr.Blocks() as demo:
gr.Markdown(title)
gr.Markdown(description)
with gr.Row():
with gr.Column():
example_instruction = gr.Textbox(label="Example Instruction")
example_answer = gr.Textbox(label="Example Answer")
with gr.Column():
user_input = gr.Textbox(label="Your Question")
system_prompt = gr.Textbox(label="System Prompt", value="You are an expert medical analyst:")
submit_btn = gr.Button("Submit")
output = gr.Textbox(label="Response")
submit_btn.click(
fn=bot.predict,
inputs=[example_instruction, example_answer, user_input, system_prompt],
outputs=output
)
demo.launch()
if __name__ == "__main__":
main()