AWS Deep Racer Ubuntu setup with NVIDIA setup

aws-deepracer-ubuntu-local-installs

sudo apt update
sudo apt upgrade
nvidia-smi
sudo apt install nvidia-utils-515
nvidia-smi
sudo apt install nvidia-driver-515
sudo prime-select nvidia
nvidia-smi
sudo reboot
nvidia-smi
git clone https://github.com/aws-deepracer-community/deepracer-for-cloud
sudo apt install git
git clone https://github.com/aws-deepracer-community/deepracer-for-cloud
docker
sudo apt update
sudo apt-get install ca-certificates curl gnupg lsb-release
sudo mkdir -p /etc/apt/keyrings
echo   "deb [arch=$(dpkg --print-architecture) signed-by=/etc/apt/keyrings/docker.gpg] https://download.docker.com/linux/ubuntu \
$(lsb_release -cs) stable" | sudo tee /etc/apt/sources.list.d/docker.list > /dev/null
sudo apt update
sudo apt-get install docker-ce docker-ce-cli containerd.io docker-compose-plugin
sudo usermod -aG docker $USER
sudo apt install docker-compose
docker ps
sudo systemctl enable docker.service
sudo systemctl enable containerd.service
sudo reboot
docker ps

distribution=$(. /etc/os-release;echo $ID$VERSION_ID)
curl -s -L https://nvidia.github.io/nvidia-docker/gpgkey | sudo apt-key add -
curl -s -L https://nvidia.github.io/nvidia-docker/$distribution/nvidia-docker.list | sudo tee /etc/apt/sources.list.d/nvidia-docker.list
sudo apt-get update && sudo apt-get install -y nvidia-container-toolkit
sudo systemctl restart docker
sudo docker run --gpus all nvidia/cuda:11.0-base nvidia-smi
docker pull awsdeepracercommunity/deepracer-sagemaker:5.0.0-gpu
ls
cd deepracer-for-cloud/
ls
bin/init.sh 
sudo apt install jq
bin/init.sh 
source bin/activate.sh 
sudo apt install awscli
pip install boto3
pip install --upgrade awscli
aws configure --profile minio
source bin/activate.sh
docker swarm init --advertise-addr 127.0.0.1 #if incase multiple network interfaces issue
source bin/activate.sh 
bin/init.sh 
source bin/activate.sh 
docker ps

aws configure --profile minio

  -> edit deepracer-for-cloud
  -> system.env to change to gpu for sagemaker image
  -> change hyperparameters.json, model_metadata.json, reward_function.py for getting better model
  
 dr-upload-custom-files #otherwise 404 HeadObject operation: Not Found
 dr-start-training (dont stop docker ps if feel struck.. instead open new terminal and activate, issue dr-stop-training)

if training started -> localhost:8080 -> click kvs_stream to see live training and evaluvation

  -> To run policy training update below settings in /etc/docker/daemon.json
  {
    "runtimes": {
      "nvidia": {
        "path": "nvidia-container-runtime",
        "runtimeArgs": []
      }
    },
    "default-runtime": "nvidia"
  }

How can I stop a model and then restart it on a slightly different reward function without going all the way back to iteration 0

You can also just update the run.env files manually:

DR_LOCAL_S3_MODEL_PREFIX=Experiment11c
DR_LOCAL_S3_PRETRAINED=True
DR_LOCAL_S3_PRETRAINED_PREFIX=Experiment11b

I like using a different python file for each reward function I am testing...
I am constantly updating the file manually. Here you can change the name of the reward function to whatever you want (assuming you have an actual file with that same name in your custom_files folder)
DR_LOCAL_S3_REWARD_KEY=$DR_LOCAL_S3_CUSTOM_FILES_PREFIX/reward_function_returnspeed_exp11b.py

dr-update
dr-increment-training