In this blog post, we’ll explore how to create a ChatGPT-like application using Hyperdiv and Ollama. Hyperdiv provides a flexible framework for building web applications, while Ollama offers powerful local machine learning capabilities.

We will start with the Hyperdiv GPT-chatbot app template and adapt it to leverage Ollama, which runs locally. This guide will walk you through the necessary steps and code changes to integrate these technologies effectively.

TL;DR

The complete code for this tutorial is here.

Step 1: Setting Up Your Environment

Install Ollama

Download Ollama from https://ollama.com/download.

Install (Windows) or unpack (macOS) the downloaded file. This gets you an Ollama app (which allows you to start the Ollama service) and a Ollama command line.

Start the Ollama service by starting the Ollama app.

On macOS, you will see an icon for the Ollama Servce at the top bar.

Then, open a terminal and type ollama list. This command displays the install models.

ollama list

To install a model, type

ollama pull llama3

For our ChatGPT Clone, we will use the llama3 model.

If you want to use another model, then search here: https://ollama.com/library

Clone the HyperDiv Examples Repository

Start by cloning or downloading the Hyperdiv GPT-chatbot app. This app provides a basic structure for a chatbot application, which we will modify to work with Ollama.

Go to your desired local folder to store the sources and type

git clone https://github.com/hyperdiv/hyperdiv-apps

Then, go to the folder hyperdiv-apps/gpt-chatbot

Adapt app to use Ollama backend

First, we will create an ollama client to process all request:

from openai import OpenAI

client = OpenAI(
    base_url="http://localhost:11434/v1",
    api_key="ollama",
)

Then we modify the request function to use this client

We change

response = openai.ChatCompletion.create(

to

response = client.chat.completions.create(

Next step is changing the accees to the response fields. With OpenAI, the response data is a dictionary, so the way to acess the fields is like

chunk["choices"]

With Ollama, we can access the field by name

chunk.choices

The changes are

 for chunk in response:
    message = chunk.choices[0].delta
    state.current_reply += message.content

And the last step would be the change to use the correct model:

model = form.select(
    options=("codellama", "llama2", "llama3", "mistral"),
        value="llama3",
        name="gpt-model",
)

Thats is! Save all changes

Prepare Python environment and run app

Install the required modules:

pip install openai hyperdiv

Run the app:

python start.py

Open the browser at http://localhost:8888

Final Result

The complete code for this tutorial is here.

Python

print("Hello, World!")

Java

public class HelloWorld {
    public static void main(String[] args) {
        System out println("Hello, World!");
    }
}

C

#include <stdio h>

int main() {
    printf("Hello, World!\n");
    return 0;
}

4 C++:

#include <iostream>

int main() {
    std::cout << "Hello, World!" << std::endl;
    return 0;
}

JavaScript

console log("Hello, World!");

Ruby

puts "Hello, World!"

Swift

print("Hello, World!")

Go

package main

import "fmt"

func main() {
    fmt Println("Hello, World!")
}

Rust

fn main() {
    println!("Hello, World!");
}

PHP

<?php
echo "Hello, World!";
?>

Perl

print "Hello, World!\n";

Kotlin

fun main() {
    println("Hello, World!")
}

Scala

object HelloWorld {
  def main(args: Array[String]): Unit = {
    println("Hello, World!")
  }
}

Lua

print("Hello, World!")

Haskell

main :: IO ()
main = putStrLn "Hello, World!"

Dart

void main() {
  print('Hello, World!');
}

Shell

echo "Hello, World!"

Batch

@echo off
echo Hello, World!

PowerShell

Write-Output "Hello, World!"

VBScript

MsgBox "Hello, World!"

Objective-C

#import <Foundation/Foundation h>

int main(int argc, const char * argv[]) {
    @autoreleasepool {
        NSLog(@"Hello, World!");
    }
    return 0;
}

Assembly

section  data
    hello db 'Hello, World!',10
    len equ $ - hello

section  text
    global _start

_start:
    ; write our string to stdout
    mov eax, 4         ; sys_write
    mov ebx, 1         ; file descriptor 1 (stdout)
    mov ecx, hello     ; message to write
    mov edx, len       ; message length
    int 0x80           ; syscall
    ; exit
    mov eax, 1         ; sys_exit
    xor ebx, ebx       ; exit status 0
    int 0x80           ; syscall

VBA (Visual Basic for Applications)

Sub HelloWorld()
    MsgBox "Hello, World!"
End Sub

Tcl

puts "Hello, World!"

COBOL

       IDENTIFICATION DIVISION 
       PROGRAM-ID  HELLO-WORLD 
       PROCEDURE DIVISION 
           DISPLAY "Hello, World!" 
           STOP RUN 

26 F#:

printfn "Hello, World!"

Elixir

IO puts "Hello, World!"

SQL (MySQL)

SELECT 'Hello, World!';

SQL (SQLite)

SELECT 'Hello, World!';

SQL (PostgreSQL)

SELECT 'Hello, World!';

SQL (Oracle)

SELECT 'Hello, World!' FROM DUAL;

SQL (SQL Server)

PRINT 'Hello, World!';

Smalltalk

Transcript show: 'Hello, World!'; cr 

R

cat("Hello, World!\n")

Bash

echo "Hello, World!"

Erlang

-module(hello) 
-export([hello_world/0]) 

hello_world() ->
    io:fwrite("Hello, World!~n") 

Julia

println("Hello, World!")

MATLAB

disp('Hello, World!');

AutoHotkey

MsgBox, Hello, World!

Clojure

(println "Hello, World!")

Groovy

println "Hello, World!"

OCaml

print_endline "Hello, World!"

D

import std stdio;

void main()
{
    writeln("Hello, World!");
}

Crystal

puts "Hello, World!"

Nim

echo "Hello, World!"

Common Lisp

(format t "Hello, World!



Scheme



(display "Hello, World!")
(newline)
Prolog
:- initialization(main) 

main :-
    write('Hello, World!'), nl,
    halt 
ABAP
REPORT ZHELLO_WORLD 

WRITE: / 'Hello, World!' 
VB NET
vb net
Module HelloWorld
    Sub Main()
        Console WriteLine("Hello, World!")
    End Sub
End Module