🦙 llama — Terminal Chatbot using LLaMA 3.2 90B

Lightweight CLI assistant powered by DeepInfra's LLaMA-3.2-90B-Vision-Instruct model.

Ask questions in the terminal like this:

llama "What is a monad in functional programming?"

Or enter an interactive REPL:

llama

---

🔧 Installation

Step 1: Clone the Repo

git clone https://github.com/notsopreety/llama.git
cd llama

Step 2: Install Python Dependencies

pip install -r requirements.txt

---

Step 3: Make llama Globally Executable

chmod +x llama.py
mv llama.py /data/data/com.termux/files/usr/bin/llama  # for Termux
# OR for Linux/macOS
sudo mv llama.py /usr/local/bin/llama

Now you can run it anywhere using:

llama "tell me a joke"

---

🧪 Features

Command	Description
llama	Start interactive REPL
llama <query>	Ask a one-off question
llama clear	Clear current conversation (preserves system prompt)
llama context <prompt>	Set custom system prompt
llama reset	Clear history and reset to default system prompt

---

💻 Termux Support

If you're using Termux, follow these additional steps:

pkg install git python
pip install -r requirements.txt
chmod +x llama.py
mv llama.py /data/data/com.termux/files/usr/bin/llama

Now use llama anywhere from your Termux shell!

---

📁 requirements.txt

requests

---

🧠 Default System Prompt

"Be a helpful assistant"

Use llama context "<your custom prompt>" to customize it.

---

📂 Conversation History

History is saved at:

~/.llama_conversation.json

You can inspect, edit, or delete it if needed.

---

📜 License

LICENSE

---

🙋‍♂️ Author

@notsopreety

---