A Telegram bot that allows users to interact with Ollama models through a conversational interface with support for multiple conversations and system prompts.

Features

Integration with Ollama models
Multiple conversations per user
Persistent conversation memory for each conversation
Customizable system prompts per conversation
Support for text files (PDF, TXT, HTML, JSON, Markdown, CSV)
Image analysis (if model supports vision)
Simple and intuitive interface
Configurable through environment variables

Setup

Clone the repository

Create a .env file with the following variables:

# Telegram Bot Configuration
TELEGRAM_BOT_TOKEN=your_telegram_bot_token

# Ollama Configuration
OLLAMA_URL=http://localhost:11434
OLLAMA_MODEL=llama2

# Memory Configuration
MEMORY_DIR=./memory

# Default System Prompt
DEFAULT_SYSTEM_PROMPT=You are a helpful AI assistant.

Install dependencies:
```
pip install -r requirements.txt
```
Make sure Ollama is running and the specified model is available

Usage

Start the bot:
```
python src/main.py
```
Open Telegram and start a conversation with your bot
Use the /start command to begin
Use /help to see all available commands

Available Commands

Conversation Management

/conv [number] - Switch to conversation
/conv - List all conversations
/renameconv [number] [title] - Rename a conversation
/reset [number] - Reset a conversation (preserves system prompt)

System Prompt

/sysprompt [number] [prompt] - Set system prompt
/sysprompt [number] - View system prompt

Help

/help - Show all available commands

File Support

The bot can process the following file types:

Text files (.txt)
PDF documents (.pdf)
HTML files (.html)
JSON files (.json)
Markdown files (.md)
CSV files (.csv)

Image Support

If your Ollama model supports vision capabilities, you can send images for analysis.

Project Structure

src/connectors/ollama_connector.py: Handles communication with Ollama
src/managers/telegram_bot.py: Manages the Telegram bot functionality
src/memory/memory_manager.py: Handles user conversation history
src/main.py: Entry point of the application

Requirements

Python 3.7+
Ollama running locally
Telegram Bot Token
Required Python packages (see requirements.txt)

Memory Structure

Each user's conversations are stored in JSON files with the following structure:

{
  "conversations": {
    "0": {
      "messages": [
        {
          "role": "user",
          "content": "message content",
          "timestamp": "ISO timestamp"
        }
      ],
      "title": "conversation title",
      "system_prompt": "system prompt"
    }
  }
}