How to use LLAMA via Line Chatbot

LLAMA

LLAMA is an AI developed by Meta to compete with ChatGPT from OpenAI. Additionally, it is an open-source model that allows developers to further enhance it. It supports up to 8 languages, including Thai, making it very suitable for integrating with Line Chatbot via API without any additional API service charges.

Make.com

The main tool we’ll use is Make.com, a website that helps with various automation by connecting different services together into workflows, reducing user work. Importantly, users do not need coding knowledge to use it smoothly.

Line Chatbot

Line Chatbot allows us to integrate other APIs to work behind the scenes in Line Chat. We’ll use Make to extract text from Line and send it as a prompt to LLAMA, generate a response, and send it back to Line. Line acts as a UI for inputting information without needing to visit the website. With this setup, you have an easy-to-use LLAMA by simply adding a friend and chatting. For creating a Line Chatbot and using its API for automation, you can follow the guide here <How to Create a Line OA and Connect it to an API>

Getting API from LLAMA

Once you are on the homepage, select “API Keys”.

Click Create API Key and set a name.

That’s it! You will now have an API Key that can be used for further applications.

Steps for Automation

Go to https://www.make.com/ log in or create an account to use the service.
After logging in, you’ll land on the main page. Click Create a new scenario to create an automation flow
On this page, you can connect events from various services that support Make.com
The flow we use to create a Line Chatbot is as follows: Since we cannot directly connect the Line API with the LLAMA API, we need to connect through a website that uses the LLAMA model. Therefore, we need to create a JSON file to place the messages from Line into the website as a prompt, then retrieve the response as a reply message in our Line chat.
Next, for the Line configuration in the Watch Event section, click Add and set a Connection Name. Then, go to the link for the Line Developer: Message API at https://developers.line.biz/console to obtain the Access Token from Line and enter it in the Channel Access Token field.
Set a variable to extract messages from the Line chat as text to be used as a prompt for LLAMA.
Create a JSON file to place the message from Line as a prompt input on the GroqCloud website. On the JSON page, select ‘Create JSON,’ name it, and then click ‘Generate.
Set the variables Messages, Content, Model, and Max_tokens. Configure each variable according to the image below.
Set up the Prompt for LLAMA to learn the basic commands, review the Prompt from the User Query, select the Model, and configure the Max Tokens.
HTTP Make Request connects to the website to use LLAMA by entering the URL as https://console.groq.com/playground.
After creating an account, go to groq.com/playground. Try typing a message => submit => view code, then copy the link within the red box.
Create an HTTP Flow and paste the code copied from the Playground page. Then, paste the API Key created on the Playground and set up the Prompt stored in the variable.
Next, create a Line Flow as Send Reply Message and select the answer from the previous step to be the response in the Line Text field.

That’s it! You can easily use LLAMA through the Line Chatbot. It’s like having an assistant to help you with tasks without the need to go through the website, making it less complicated.

Additionally, you can connect the Line Chatbot to other AIs like Claude, ChatGPT, or DALL-E. You can subscribe to receive notifications for new blogs via the form below.

[hubspot type=”form” portal=”47406981″ id=”244b28eb-5322-4f57-80ff-1745ac75313c”]