Simple CLI chat bot that can interact with your source code.
git clone https://github.com/aldmbmtl/Fin.git
cd Fin
make setup
make installFin uses a configuration file to store the settings for the Ollama API and the project. Much like a .gitignore or
any other kind of dot file, you would create a .fin.yaml file in the root of your project.
llm:
server: https://ollama.example.com
model: llama2
project:
languages:
- python
- javascriptThis allows you to specify what languages are used in your project so Fin can better help you. You can also specify the Ollama server and model to use. For reference, you can look at the .fin.yaml.sample file.
You have 2 options
- Try out fin directly off the source
- Install fin more permanently in your $HOME/.local/bin
To try out fin directly off the source, you can run the following command:
make runThis will launch the fin chat bot and you can start chatting with it.
To install fin more permanently in your $HOME/.local/bin, you can run the following command:
make install-cliFrom here, you can type fin in your terminal and start chatting with the bot.
You can start a chat thread by just typing fin. It will quickly connect to your llm and provide
the llm with some context. For example, if you have configured that your project uses python, it will
send a message to the llm saying that you are working on a python project. You can also pass in more
than one language in the configuration file.
Once the bot is launched and the thread is started, you will see a prompt that looks like this:
Fin>:
You can start typing your message and the bot will respond to you.
You can also use the bot to run adhoc commands. For example:
fin what is the meaning of life
This will set your context, send your request and return the response. You have to remember that no context survives after the request is made. So if you want to chat with the bot, you will have to start a chat thread instead.
This is an example response:
Fin>: what is the meaning of life
─────────────────────────────────────────────────────────────────
AI: Ah, a philosophical question! chuckles The meaning of
life is a complex and deeply personal topic that has
puzzled humans for centuries. As an AI language model, I
don't have personal experiences or emotions like humans
do, so I can't provide a definitive answer to this
question. However, I can offer some insights based on
various philosophical and scientific perspectives.
From a biological perspective, the meaning of life can be
seen as fulfilling basic needs such as survival,
reproduction, and self-expression. For humans, these
needs are often met through social interactions,
relationships, and personal growth.
From a philosophical standpoint, the meaning of life has
been debated by many great thinkers throughout history.
Some argue that life has no inherent meaning and that we
must create our own purpose through our choices and
actions. Others believe that life has a predetermined
purpose or that it is part of a greater cosmic plan.
Ultimately, the meaning of life is a deeply personal and
subjective question that each individual must answer for
themselves. What do you think? Can you tell me more about
what you're looking for in terms of answers to this
question?
─────────────────────────────────────────────────────────────────
You can also ask Fin what a file does. For example:
fin
├── __init__.py
├── chat_loop.py
├── config_handler.py
├── connector.py
└── engine.pyYou can ask Fin what the chat_loop.py file does by typing:
fin what does ./chat_loop.py do
This will open and send the file up to the llm and tag it with that file path. The llm will then respond with what it thinks the file does.
You can also do this with multiple files. For example:
fin how do I use ./chat_loop.py and ./config_handler.py together
─────────────────────────────────────────────────────────────────
Loading files ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% 0:00:02
Great! It sounds like you're looking to create a simple
chatbot using the fin package, which is a Python package
for building chatbots. The chat_loop.py module provides
the primary chat loop for the bot, while the
config_handler.py module loads configuration from a YAML
file.
To use these two modules together to create a simple chat
loop, you can follow these steps:
1 First, make sure you have both chat_loop.py and
config_handler.py in the same directory as your main
code.
2 Import the necessary modules at the top of your main
code file: from fin import start_up; from
fin.config_handler import load_config.
3 Initialize the load_config() function to load
configuration from the YAML file. You can do this by
calling load_config() before starting the chat loop.
4 Start the chat loop using the start_up() function.
This will enter the main loop of the chatbot, where
you can handle user input and render responses based
on the loaded configuration.
5 Inside the chat loop, you can use the load_config()
function to access the configuration data and use it
to guide the conversation. For example, you could use
the configuration to determine which responses to show
to the user based on their inputs.
Here's an example of how this might look in code:
from fin import start_up; from fin.config_handler impor
load_config
# Load configuration from YAML file
load_config()
# Start chat loop
start_up()
while True:
# Handle user input and render response based on
config
pass
By using these two modules together, you can create a
simple chatbot that loads configuration from a YAML file
and uses it to guide the conversation. Of course, this is
just a basic example, and you can customize and extend
the fin package to build more complex chatbots with
additional features and functionality.
─────────────────────────────────────────────────────────────────