langchain_js_console_application

This repository contains a console application made up of a set of tools to automate the daily routine tasks using LangChain JS and OpenAI's LLMs.

You can generate text by these tasks:

• Simplify text,
• Translate from one language to another,
• Create summaries,
• Find synonys/antonyms,
• Create a Linkedin post as a summary by providing a text.

Instructions:

1. create a new folder in your local computer, navigate to it via the command prompt/terminal,
2. type 'git clone https://github.com/valeadami/LangChainJSConsoleApplication.git',
3. type 'npm install',
4. add an .env file and a folder named 'files' under the root, in which you will create a text file named 'blog.txt',
5. run 'node index.js' to start the console application,
6. select your option: for tasks from #1-#5 just enter the text via console. For task #6, the program reads blog.txt so the LLM can create a Linkedin post after reading its content,
7. select your options, see the result, press S to continue the interaction, N to end.

.env file

This file is a configuration file where you set up paths and folders required by the application.

You can create first a env.txt, add the following content and then rename it as .env.

In order to use OpenAI's models it is mandatory to add the value of your OpenAI key:

OPENAI_API_KEY=INSERT HERE YOUR OPENAI KEY

These keys are optional

  LOG_ENABLED=false
  PATH_HISTORY_FILE= path to a file to trace the output from LLM
  BLOG_FILE=path to a file for task #6

If LOG_ENABLED=true you need to setup a PATH_HISTORY_FILE, i.e:

PATH_HISTORY_FILE=./history/historyfile.txt

Historyfile.txt tracks your activity: prompt, LLM output, date and time.

A complete example of .env might be the following:

OPENAI_API_KEY=sk-AAA123434
DEBUG=false
MYDEBUG=true
LOG_ENABLED=true
PATH_HISTORY_FILE=./history/historyfile.txt
BLOG_FILE=./files/blog.txt

Feel free to give the names to folder and files according to your needs.