Connect Kobold AI API with your Application

KoboldAI is a powerful and simple to use platform for a variety of AI-based text-generation experiences. Kobold API can be used to offer metadata about this API, and the model option is used to provide information about the current text generation model. world_info provides Endpoints for handling world information in the KoboldAI GUI, among other things.

In this article, we’ll go through how to get started with Kobold AI API applications using features like Info, Model, Story, Config, World Info, and other adjustable AI settings.

How to Get Started with Kobold API

This tutorial does not require any technical knowledge. This is really simple to set up, and you will have your own Kobold AI API up and running in minutes.

Colab is a Google-created research tool based on Jupyter notebooks that may be used to deploy Python or R-based models or scripts. You will also receive a complimentary Nvidia Tesla T4 GPU. In the Google Cloud, this GPU costs roughly $250. We’ll now set up Kobold AI in Google Colab.

Go to the Kobold AI with GPU link.

Step 1: Install KoboldAI on Google Colab.

If you are interested, you can visit the article about How to Install Kobold AI API: An Easy Step-by-Step Guide for a more detailed explanation of the installation procedure.

Step 2: Play an audio file to keep the tab open. If you are performing this process on a mobile phone, you must press the play button to generate an audio file. This is because the deployment takes time, and you must play the audio to keep the tab open in your browser. Start playing the audio by clicking the play icon.

Step 3: Now, in the next section, you need to choose the desired model that suits you. The default model, for example, is Nerys-V2-6 and the version is official. If you wish to alter the option for the new UI in Kobold UI United, you only need to make one change in the settings. If you want to change the model and style, you can use one of the given models.

Step 4: Click the Play button to begin the deployment. Wait for the deployment to finish installing all of the drivers and packages needed to execute the Kobold AI API.

Step 5: When the deployment has completed, you will get an output similar to the one below. Then You will then be given a URL to test your Kobold AI API in your browser.

Output for version official:

INIT | Starting | LUA bridge INIT | OK | LUA bridge INIT | Starting | LUA Scripts INIT | OK | LUA Scripts INIT | OK | Webserver MESSAGE | KoboldAI has finished loading and is available at the following link: https://petersburg-source-shark-blade.trycloudflare.com

Output for version United:

INFO | modeling.inference_models.hf:set_input_parameters:191 – {‘use_gpu’: True, ‘0_Layers’: 32, ‘CPU_Layers’: 0, ‘Disk_Layers’: 0, ‘id’: ‘KoboldAI/OPT-6B-nerys-v2’, ‘model’: ‘KoboldAI/OPT-6B-nerys-v2’, ‘path’: None, ‘menu_path’: ”} INIT | Starting | Flask INIT | OK | Flask INIT | Starting | Webserver INIT | OK | Webserver MESSAGE | KoboldAI has finished loading and is available at the following link for UI 1: https://claimed-g-jokes-undertaken.trycloudflare.com MESSAGE | KoboldAI has finished loading and is available at the following link for UI 2: https://claimed-g-jokes-undertaken.trycloudflare.com/new_ui MESSAGE | KoboldAI has finished loading and is available at the following link for KoboldAI Lite: https://claimed-g-jokes-undertaken.trycloudflare.com/lite MESSAGE | KoboldAI has finished loading and is available at the following link for the API: https://claimed-g-jokes-undertaken.trycloudflare.com/api

Then, in the list of links, select the API. If you copy the API to a new tab, it will be available.

Explore the Kobold AI API

You can now explore the Kobold AI API and feature information.

Kobold AI API applications are:

• Info
• Model
• Story
• World Info
• config

Info: Metadata about this API

In this info options, there are three options for checking the current, latest, and list API versions.

Checking the current API versions

1. First, check the API’s current version. You must click the GET option.

1. Then, in the right corner, you click the try it out option, followed by the Execute option.

1. After the execution process, it will display the following parameters, response, curl, and requested URL. Then the URL will be copied into a new tab to return the version of the API that is currently in use.

1. Returns the current version of the API.

Checking the Latest API Versions

All previous steps will be repeated. However, select the GET option to obtain the latest version information.

1. First, check the API’s latest version. You must click the Second GET option.
2. Then, in the right corner, you click the try it out option, followed by the Execute option.
3. After the execution process, it will display the following parameters, response, curl, and requested URL. Then the URL will be copied into a new tab to return the version of the API that is latest use.
4. Returns the latest API version available.

Checking the List of Available API Versions

All previous steps will be repeated. However, select the GET option to obtain the list of available API versions sorted in ascending order.

1. First, check the list of available API versions. You must click the Third GET option.
2. Then, in the right corner, you click the try it out option, followed by the Execute option.
3. After the execution process, it will display the following parameters, response, curl, and requested URL. Then the URL will be copied into a new tab to return the list of available API versions.
4. Returns a list of available API versions sorted in ascending order.

Generate: Text generation Endpoint

• First, to generate text, you must click the generate option.

• Then, in the right corner, you click the try it out option, followed by the Execute option.

• After the execution process, it will display the following parameters, response, curl, and requested URL. Then the URL will be copied into a new tab to Text is generated in response to a submission, sampler settings, soft prompt, and number of return sequences.

Model: Current Text Generation Model

• Previous steps should be taken to model information about current text generation. There are two types in this option: GET and PUT.

• Get is used for gets the current model string, which is presented in brackets in the title of the Kobold AI GUI, e.g., “Kobold AI API Client (Kobold AI API/fairseq-dense-13B-Nerys-v2)” and uses it. PUT option is used for to Loads a model based on its Hugging Face model ID, the path to a model folder (relative to the Kobold AI API root folder’s “models” folder), or “Read Only” for no model.

Story: Endpoints for Managing the Story in GUI

• The previous steps should be followed for Story and Endpoints for managing the story in the KoboldAI GUI. This option includes multiple types, including GET, DELETE, POST, and PUT.

• Each has a different purpose, but the steps required to obtain the requested URL should be similar.

World_info: Endpoints for Managing World Information in the KoboldAI GUI

• The previous steps should be followed for world_info, which is used in the KoboldAI GUI to manage world information. This option has several types, including GET, DELETE, POST, and PUT. Each has a different purpose, but the steps to obtain the necessary URL should be similar.

Config: get/set various setting values

The previous steps should be followed for config, which allows you to retrieve or set various configuration values. This option has two types, including GET and PUT.

• Each has multiple purposes, which are to retrieve and set the current author note setting value, author’s note depth setting to a specified value, template setting value, gens per action setting value, sampler order setting value, remove the current single name, set the soft prompt name, template setting, sampling setting, input formatting, and output formatting. remove special characters, remove blank links, maximum content setting value, memory setting value, and more.

However, the methods to obtain the specified URL should be same.

Also read: You might also find useful our guide on How to Use Janitor AI API – Your Ultimate Step-by-Step Guide

Conclusion

In conclusion, getting started with the Kobold AI API is simple if you follow these simple steps. By creating an account, obtaining your API key, and integrating it into your project, you can unlock the power of Kobold’s features. Explore its docs for more information and make the most of this powerful tool.

Please feel free to share your thoughts and feedback in the comment section below.