slimStreamQ - Clean and efficient implementation of Streaming Q($\lambda$)

Paper 👉📄 | Original code 👉👨‍💻 (in Pytorch)

User installation

CPU installation:

python3 -m venv env_cpu
source env_cpu/bin/activate
pip install --upgrade pip setuptools wheel
pip install -e .[dev]

GPU installation if needed:

python3 -m venv env
source env/bin/activate
pip install --upgrade pip setuptools wheel
pip install -e .[dev,gpu]

Running experiments

To train a Stream Q($\lambda$) agent on Breakout on your local system, run:
launch_job/atari/launch.sh

• To see the stage of training, you can check the logs in experiments/atari/logs/test_Breakout/qlambda
• The models and episodic returns are stored in experiments/atari/exp_output/test_Breakout/qlambda