app.py

import gradio as gr
import requests
import subprocess
import time

# Start the Ollama server and wait for it to output the URL
subprocess.Popen(['bash', 'run_ollama.sh'])
time.sleep(10)  # Adjust sleep time based on how long it takes Ollama to start

# Read the URL from the file
with open('ollama_url.txt', 'r') as file:
    OLLAMA_URL = file.read().strip()

console_output = []

def call_ollama_api(prompt):
    response = requests.post(
        f'{OLLAMA_URL}/generate',
        json={"prompt": prompt}
    )
    if response.status_code == 200:
        result = response.json()
        output = result.get('text', '')
        console_output.append(f"Prompt: {prompt}\nResponse: {output}")
        return output
    else:
        error_message = f"Error: {response.status_code} - {response.text}"
        console_output.append(error_message)
        return error_message

def chat(prompt):
    output = call_ollama_api(prompt)
    return output

def show_console_output():
    return "\n\n".join(console_output)

iface = gr.Interface(
    fn=chat,
    inputs="text",
    outputs="text",
    title="Ollama Chat",
    description="Chat with Ollama API and see the responses."
)

console_iface = gr.Interface(
    fn=show_console_output,
    inputs=None,
    outputs="text",
    title="Console Output",
    description="See the console outputs here."
)

gr.TabbedInterface([iface, console_iface], ["Chat", "Console"]).launch()