# Installation

## :minidisc: Installing SLM Lab

Clone the repository:

```bash
git clone https://github.com/kengz/SLM-Lab.git
```

Install the dependencies:

```bash
cd SLM-Lab/
./bin/setup
```

This runs a prepared bash script with the necessary setup steps, with Python dependencies managed through Conda. Refer to the [Help](https://slm-lab.gitbook.io/slm-lab/v4.2.0/resources/help) page if you encounter issues.

{% hint style="info" %}
Readers of the book:book:*Foundations of Deep Reinforcement Learning:* please see [this custom instruction page](https://slm-lab.gitbook.io/slm-lab/v4.2.0/publications-and-talks/instruction-for-the-book-+-intro-to-rl-section).
{% endhint %}

### Alternative Installations

#### Windows

The best way to run SLM Lab on Windows is to use a Bash shell/Linux subsystem. Credit to [**@vladimirnitu**](https://github.com/vladimirnitu) and [**@steindaian**](https://github.com/steindaian) for providing a detailed instruction PDF for doing so on Windows:

{% file src="<https://2605794550-files.gitbook.io/~/files/v0/b/gitbook-legacy-files/o/assets%2F-LterKThgiKKTvsbx4WI%2F-MaZNrSDmc574fYyhAnd%2F-MaZOQIKoLZkL95-PQnt%2FSLM_for_Windows.pdf?alt=media&token=d45d9fbb-3114-4b7a-b41e-ae69663fe915>" %}
SLM Lab for Windows (Instruction PDF)
{% endfile %}

#### Google Colab/Jupyter Notebook

Although we do not recommend running SLM Lab on Google Colab or Jupyter notebooks (notebooks come with inherent limitations, e.g. no rendering/multi-processing), we have prepared an example notebook for illustration. Credit to [**@piosif97**](https://github.com/piosif97) for helping with this:

{% hint style="info" %}
[SLM Lab Colab notebook](https://gist.github.com/kengz/6fd52a902129fb6d4509c721d71bda48)
{% endhint %}

For details on how it works, refer to [this Help section](https://slm-lab.gitbook.io/slm-lab/resources/help#google-colab-jupyter-setup).

## :desktop: Hardware Requirements

Non-image based environments can run on a laptop. Only image based environments such as the Atari games benefit from a GPU speedup. For these, we recommend 1 GPU and at least 4 CPUs. This can run a single Atari `Trial` consisting of 4 `Sessions`.

For desktop, a reference spec is GTX 1080 GPU, 4 CPUs above 3.0 GHz, and 32 GB RAM.

For cloud computing, start with an affordable instance of [AWS EC2 `p2.xlarge`](https://aws.amazon.com/ec2/instance-types/p2/) with a K80 GPU and 4 CPUs. Use the Deep Learning AMI with Conda when [creating an instance](https://aws.amazon.com/getting-started/tutorials/get-started-dlami/).&#x20;