31 lines
785 B
Markdown
31 lines
785 B
Markdown
# Installation & Setup for `trollagent_ai`
|
|
|
|
## 1. Prerequisites
|
|
- Python 3.10+
|
|
- An accessible [Ollama](https://ollama.com/) instance serving a reasoning model (default: `gemma4:12b` or any OpenAI-compatible Ollama model).
|
|
|
|
## 2. Environment Setup
|
|
```bash
|
|
cd trollagent_ai
|
|
python3 -m venv venv
|
|
source venv/bin/activate
|
|
pip install -r requirements.txt
|
|
```
|
|
|
|
## 3. Ollama Configuration
|
|
By default, the agent connects to:
|
|
- `OLLAMA_BASE_URL`: `http://192.168.1.220:11434` (or `http://localhost:11434` for local instance)
|
|
- `OLLAMA_MODEL`: `gemma4:12b`
|
|
|
|
Make sure the model is pulled and running:
|
|
```bash
|
|
ollama pull gemma4:12b
|
|
# Or test with llama3 / mistral
|
|
ollama pull llama3
|
|
```
|
|
|
|
## 4. Run the Bot
|
|
```bash
|
|
python3 bot.py --name BloodTroll --color "#15803d" --strength 4 --health 10
|
|
```
|