app.py

from llama_cpp.server.app import create_app, Settings
from fastapi.responses import HTMLResponse
import os
import requests
from llama_cpp import Llama
import gradio as gr

url="https://huggingface.co/TheBloke/WizardLM-13B-V1.2-GGUF/resolve/main/wizardlm-13b-v1.2.Q4_0.gguf"
response = requests.get(url)
with open("./model.gguf", mode="wb") as file:
  file.write(response.content)

llm = Llama(model_path="./model.gguf")
def generate_text(input_text):
    output = llm(f"Q: {input_text} A:", max_tokens=256, stop=["Q:", "\n"], echo=True)
    return output['choices'][0]['text']

input_text = gr.inputs.Textbox(lines= 10, label="Enter your input text")
output_text = gr.outputs.Textbox(label="Output text")

description = "llama.cpp implementation in python [https://github.com/abetlen/llama-cpp-python]"

examples = [
    ["What is the capital of France? ", "The capital of France is Paris."],
    ["Who wrote the novel 'Pride and Prejudice'?", "The novel 'Pride and Prejudice' was written by Jane Austen."],
    ["What is the square root of 64?", "The square root of 64 is 8."]
]

gr.Interface(fn=generate_text, inputs=input_text, outputs=output_text, title="Llama Language Model", description=description, examples=examples).launch()