Tensorblue

The tensorblue package provides an interface for interacting with OpenAI models, specifically tailored for different user tiers. It includes functions for chat-based language generation, image generation, and vision processing.

npm install tensorblue

Example Code

Try to run of the responses and comment out the other to see the result.

• Basic(tier 1) user will get to use chat function
• Pro(tier 2) user will get to use chat and image
• Enterprise(tier 3) user will get to use all 3 of them.

How to use this package?

1. Setup a node enviroment.
2. Create a .env file and put your API key with the name 'USER_KEY'.
3. You need 'dotenv' package, for accessing the API key.
4. Name it as script.js

const { chat, image, vision } = require('tensorblue');
const { config } = require('dotenv');
config();

const userkey = process.env.USER_KEY;

//Comment out one of the three response 

const response = async () => {
    return await chat("What is npm?", userkey);
}

const response = async() => {
    return await image("Ancient India map", userkey)
}

const response = async() => {
    return await vision(
        "Explain this image for me.",   
 "https://upload.wikimedia.org/wikipedia/commons/thumb/b/b6/Image_created_with_a_mobile_phone.png/1024px-Image_created_with_a_mobile_phone.png",
        userkey,
    )
};

response().then(result => {
    console.log(result);
});

• Run this file

node script.js

Usage

Text to text (for all tiers)

• Function:

async function chat(prompt, key)

• Description: Generates a response based on a user prompt and it's tensorblue key.

• Parameters:

  • prompt (string): User's input prompt.
  • key (string): User's API key.

• Returns: A Promise that resolves to the OpenAI response.

• Sample usage:

const response = async () => {
    return await chat("What is ISL?", userkey);
}

response().then(result => {
    console.log(result);
});

Text to Image (For pro and enterprise tier)

• Function:

async function image(prompt, key)

• Description: Generates an image based on a user prompt using tensorblue key.

• Parameters:

  • prompt (string): User's input prompt.
  • key (string): User's API key.

• Returns: A Promise that resolves to the generated image URL.

• Sample usage:

const response = async() => {
    return await image(prompt, userkey)
}

response().then(result => {
    console.log(result);
});

Vision Functioning (For enterprise tier)

• Function:

async function vision(prompt, url, key)

• Description: Processes an image and provides a description based on a user prompt.

• Parameters:

  • prompt (string): User's input prompt.
  • url (string): URL of the image for vision processing.
  • key (string): User's API key.

• Returns: A Promise that resolves to the OpenAI vision response.

• Sample usage:

const response = async() => {
    return await vision(
        "Explain this image for me.", 
        "https://upload.wikimedia.org/wikipedia/commons/thumb/b/b6/Image_created_with_a_mobile_phone.png/1024px-Image_created_with_a_mobile_phone.png",
        userkey,
    )
};

response().then(result => {
    console.log(result);
});

Error handling

• Error Handling
  • Invalid Tier Number: Throws an error if the provided tier number is not valid for a specific function based on your API key.