{
"cells": [
{
"cell_type": "markdown",
"id": "01c769fa-56c3-4280-8bdf-d37954d1effc",
"metadata": {},
"source": [
"# **ADARO-RL Pipelines Example**"
]
},
{
"cell_type": "markdown",
"id": "3f20e260-2631-4b75-a6d1-ef3a7e11cbd2",
"metadata": {},
"source": [
"### Imports"
]
},
{
"cell_type": "code",
"execution_count": 1,
"id": "d1b2fed5-31d8-4c7b-aa34-5fa889125cbe",
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/home/jovyan/Maturation/env-adaro-rl/lib/python3.10/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
" from .autonotebook import tqdm as notebook_tqdm\n"
]
}
],
"source": [
"import os\n",
"import sys\n",
"import shutil\n",
"from pathlib import Path\n",
"\n",
"import adaro_rl\n",
"import adaro_rl.viz\n",
"import adaro_rl.zoo as zoo"
]
},
{
"cell_type": "markdown",
"id": "57701917-d1be-42c9-8441-da35366b19f9",
"metadata": {},
"source": [
"### Arguments"
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "e7131de3-1728-45a9-b940-ddb9505e7447",
"metadata": {},
"outputs": [],
"source": [
"config_name=\"Enduro-v5\"\n",
"\n",
"attack_name=\"FGM_D\"\n",
"target=\"untargeted\"\n",
"eps=100.0\n",
"norm=2.0\n",
"\n",
"training_steps=2000\n",
"n_eval_episodes=2\n",
"\n",
"device=\"cuda\"\n",
"seed=0"
]
},
{
"cell_type": "markdown",
"id": "34ee1922-8928-47ec-88ff-85aaa0f889dd",
"metadata": {},
"source": [
"### Ouputs Directory"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "e655ef2c-d735-4ed6-afee-81bc4376729d",
"metadata": {},
"outputs": [],
"source": [
"output_dir = os.path.join(\"outputs\", f\"{config_name}\", f\"attack_{attack_name}_target_{target}_eps_{eps}_norm_{norm}\")\n",
"agent_dir = os.path.join(output_dir,\"agent\")\n",
"adv_trained_dir = os.path.join(output_dir,\"adv_trained_agent\")\n",
"\n",
"output_path = Path(output_dir)\n",
"agent_path = Path(agent_dir)\n",
"\n",
"# Create the directory using the Path object\n",
"output_path.mkdir(parents=True, exist_ok=True)"
]
},
{
"cell_type": "markdown",
"id": "b9335ab6-1900-4c1d-97e2-6727baca5cbf",
"metadata": {},
"source": [
"### Define Config"
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "9c64d883-4fd8-4b65-a9d3-200d23b10996",
"metadata": {},
"outputs": [],
"source": [
"config = zoo.configs[config_name]"
]
},
{
"cell_type": "markdown",
"id": "3c68f6e1-f743-403d-b5d9-033ad83c7ca6",
"metadata": {},
"source": [
"### Download the model"
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "0ea5422d-178e-428e-91dc-8c288d530f13",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"📥 Download the agent\n",
"outputs/Enduro-v5/attack_FGM_D_target_untargeted_eps_100.0_norm_2.0/agent/model.zip already exists\n"
]
}
],
"source": [
"print(\"📥 Download the agent\")\n",
"agent_path.mkdir(parents=True, exist_ok=True)\n",
"if not os.path.isfile(os.path.join(agent_dir,\"model.zip\")):\n",
" zoo.download_model(config_name, local_dir=str(agent_dir))\n",
"else:\n",
" print(f\"{os.path.join(agent_dir,'model.zip')} already exists\")"
]
},
{
"cell_type": "markdown",
"id": "b69c16f1-8aba-4600-b1ce-b7bc8ab0316d",
"metadata": {},
"source": [
"### Train the Agent"
]
},
{
"cell_type": "code",
"execution_count": 6,
"id": "2d536c71-372e-48dc-a605-d1fe8f4df5fb",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"🏋️ Train the agent\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"A.L.E: Arcade Learning Environment (version 0.10.1+unknown)\n",
"[Powered by Stella]\n",
"A.L.E: Arcade Learning Environment (version 0.10.1+unknown)\n",
"[Powered by Stella]\n"
]
},
{
"data": {
"text/html": [
"/home/jovyan/Maturation/env-adaro-rl/lib/python3.10/site-packages/rich/live.py:231: UserWarning: install \n",
"\"ipywidgets\" for Jupyter support\n",
" warnings.warn('install \"ipywidgets\" for Jupyter support')\n",
" \n"
],
"text/plain": [
"/home/jovyan/Maturation/env-adaro-rl/lib/python3.10/site-packages/rich/live.py:231: UserWarning: install \n",
"\"ipywidgets\" for Jupyter support\n",
" warnings.warn('install \"ipywidgets\" for Jupyter support')\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/home/jovyan/Maturation/env-adaro-rl/lib/python3.10/site-packages/stable_baselines3/common/on_policy_algorithm.py:150: UserWarning: You are trying to run PPO on the GPU, but it is primarily intended to run on the CPU when not using a CNN policy (you are using ActorCriticPolicy which should be a MlpPolicy). See https://github.com/DLR-RM/stable-baselines3/issues/1245 for more info. You can pass `device='cpu'` or `export CUDA_VISIBLE_DEVICES=` to force using the CPU.Note: The model will train, but the GPU utilization will be poor and the training might take longer than on CPU.\n",
" warnings.warn(\n"
]
},
{
"data": {
"text/html": [
"\n",
" \n"
],
"text/plain": [
"\n"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"print(\"\\n🏋️ Train the agent\")\n",
"adaro_rl.train(\n",
" config=config,\n",
" checkpoint=os.path.join(agent_dir,\"model.zip\"),\n",
" output_dir=str(agent_dir),\n",
" device=device,\n",
" seed=seed,\n",
" total_timesteps=training_steps,\n",
" prepopulate_timesteps=0,\n",
" verbose=False\n",
")"
]
},
{
"cell_type": "markdown",
"id": "09fb29f3-0060-4960-9365-ae7cad695e39",
"metadata": {},
"source": [
"### Evaluate the Agent"
]
},
{
"cell_type": "code",
"execution_count": 7,
"id": "0f103631-0458-4047-95d4-1530b677f630",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"🧪 Test the agent\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"A.L.E: Arcade Learning Environment (version 0.10.1+unknown)\n",
"[Powered by Stella]\n",
"A.L.E: Arcade Learning Environment (version 0.10.1+unknown)\n",
"[Powered by Stella]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"lengths : [6656, 6656]\n",
"mean : 6656.0\n",
"std : 0.0\n",
"\n",
"rewards : [483.0, 496.0]\n",
"mean : 489.5\n",
"std : 6.5\n"
]
}
],
"source": [
"print(\"\\n🧪 Test the agent\")\n",
"adaro_rl.pipelines.test(\n",
" config=config,\n",
" checkpoint=os.path.join(agent_dir,\"model.zip\"),\n",
" output_dir=os.path.join(agent_dir,\"results\"),\n",
" device=device,\n",
" seed=seed,\n",
" n_eval_episodes=n_eval_episodes\n",
")"
]
},
{
"cell_type": "markdown",
"id": "c386a8b2-0aa2-421b-981d-096abcafce56",
"metadata": {},
"source": [
"### Evaluate the Agent against the Attack"
]
},
{
"cell_type": "code",
"execution_count": 8,
"id": "385ee835-039d-47dc-86b4-03edc445025d",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"🧨 Online attack: FGM_D_untargeted_100.0_2.0\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"A.L.E: Arcade Learning Environment (version 0.10.1+unknown)\n",
"[Powered by Stella]\n",
"/home/jovyan/Maturation/env-adaro-rl/lib/python3.10/site-packages/stable_baselines3/common/on_policy_algorithm.py:150: UserWarning: You are trying to run PPO on the GPU, but it is primarily intended to run on the CPU when not using a CNN policy (you are using ActorCriticPolicy which should be a MlpPolicy). See https://github.com/DLR-RM/stable-baselines3/issues/1245 for more info. You can pass `device='cpu'` or `export CUDA_VISIBLE_DEVICES=` to force using the CPU.Note: The model will train, but the GPU utilization will be poor and the training might take longer than on CPU.\n",
" warnings.warn(\n",
"A.L.E: Arcade Learning Environment (version 0.10.1+unknown)\n",
"[Powered by Stella]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"lengths : [3328, 6656]\n",
"mean : 4992.0\n",
"std : 1664.0\n",
"rewards : [157.0, 475.0]\n",
"mean : 316.0\n",
"std : 159.0\n",
"save in outputs/Enduro-v5/attack_FGM_D_target_untargeted_eps_100.0_norm_2.0/agent/results/online_adv_attacks_norm_2.0_mean_reward.csv and outputs/Enduro-v5/attack_FGM_D_target_untargeted_eps_100.0_norm_2.0/agent/results/online_adv_attacks_norm_2.0_std_reward.csv\n"
]
}
],
"source": [
"print(f\"\\n🧨 Online attack: {attack_name}_{target}_{eps}_{norm}\")\n",
"adaro_rl.pipelines.online_attack(\n",
" config=config,\n",
" attack_name=attack_name,\n",
" target=target,\n",
" eps=eps,\n",
" norm=norm,\n",
" output_dir=os.path.join(agent_dir,\"results\"),\n",
" agent_checkpoint=os.path.join(agent_dir,\"model.zip\"),\n",
" adversary_checkpoint=None,\n",
" self_reference=True,\n",
" device=device,\n",
" seed=seed,\n",
" n_eval_episodes=n_eval_episodes\n",
")"
]
},
{
"cell_type": "markdown",
"id": "0989a963-78f6-4535-b754-8e3f07bb2f1f",
"metadata": {},
"source": [
"### Advesarial Training of the Agent aginst the Attack"
]
},
{
"cell_type": "code",
"execution_count": 9,
"id": "fc3479c9-896b-4e2e-b10a-d44d3d1ca92d",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"🛡️ Adversarial training: FGM_D_untargeted_100.0_2.0\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"A.L.E: Arcade Learning Environment (version 0.10.1+unknown)\n",
"[Powered by Stella]\n",
"/home/jovyan/Maturation/env-adaro-rl/lib/python3.10/site-packages/stable_baselines3/common/on_policy_algorithm.py:150: UserWarning: You are trying to run PPO on the GPU, but it is primarily intended to run on the CPU when not using a CNN policy (you are using ActorCriticPolicy which should be a MlpPolicy). See https://github.com/DLR-RM/stable-baselines3/issues/1245 for more info. You can pass `device='cpu'` or `export CUDA_VISIBLE_DEVICES=` to force using the CPU.Note: The model will train, but the GPU utilization will be poor and the training might take longer than on CPU.\n",
" warnings.warn(\n",
"A.L.E: Arcade Learning Environment (version 0.10.1+unknown)\n",
"[Powered by Stella]\n",
"A.L.E: Arcade Learning Environment (version 0.10.1+unknown)\n",
"[Powered by Stella]\n"
]
},
{
"data": {
"text/html": [
"/home/jovyan/Maturation/env-adaro-rl/lib/python3.10/site-packages/rich/live.py:231: UserWarning: install \n",
"\"ipywidgets\" for Jupyter support\n",
" warnings.warn('install \"ipywidgets\" for Jupyter support')\n",
" \n"
],
"text/plain": [
"/home/jovyan/Maturation/env-adaro-rl/lib/python3.10/site-packages/rich/live.py:231: UserWarning: install \n",
"\"ipywidgets\" for Jupyter support\n",
" warnings.warn('install \"ipywidgets\" for Jupyter support')\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"\n",
" \n"
],
"text/plain": [
"\n"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"print(f\"\\n🛡️ Adversarial training: {attack_name}_{target}_{eps}_{norm}\")\n",
"adaro_rl.pipelines.adversarial_train(\n",
" config=config,\n",
" attack_name=attack_name,\n",
" target=target,\n",
" eps=eps,\n",
" norm=norm,\n",
" output_dir=str(adv_trained_dir),\n",
" agent_checkpoint=os.path.join(agent_dir,\"model.zip\"),\n",
" adversary_checkpoint=None,\n",
" self_reference=True,\n",
" device=device,\n",
" seed=seed,\n",
" total_timesteps=training_steps,\n",
" prepopulate_timesteps=0,\n",
" verbose=False\n",
")"
]
},
{
"cell_type": "markdown",
"id": "b7d2c1c2-ceca-4915-aa1a-8e6148373b74",
"metadata": {},
"source": [
"### Evaluate the Adversarially Trained Agent"
]
},
{
"cell_type": "code",
"execution_count": 10,
"id": "81214731-d27c-4147-9879-df2f0c63ac69",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"🔬 Test the adversarially trained agent\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"A.L.E: Arcade Learning Environment (version 0.10.1+unknown)\n",
"[Powered by Stella]\n",
"A.L.E: Arcade Learning Environment (version 0.10.1+unknown)\n",
"[Powered by Stella]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"lengths : [6656, 9984]\n",
"mean : 8320.0\n",
"std : 1664.0\n",
"\n",
"rewards : [487.0, 682.0]\n",
"mean : 584.5\n",
"std : 97.5\n"
]
}
],
"source": [
"print(\"\\n🔬 Test the adversarially trained agent\")\n",
"adaro_rl.pipelines.test(\n",
" config=config,\n",
" checkpoint=os.path.join(adv_trained_dir,\"model.zip\"),\n",
" output_dir=os.path.join(adv_trained_dir,\"results\"),\n",
" device=device,\n",
" seed=seed,\n",
" n_eval_episodes=n_eval_episodes\n",
")"
]
},
{
"cell_type": "markdown",
"id": "62c19c02-9c08-4778-95c2-43b4e30af037",
"metadata": {},
"source": [
"### Evaluate the Adversarially Trained Agent, against the Attack"
]
},
{
"cell_type": "code",
"execution_count": 11,
"id": "d90acd5b-e5ed-4a9d-b2f4-dd8f4ea3c4c6",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"🧨 Online attack on adversarially trained agent: FGM_D_untargeted_100.0_2.0\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"A.L.E: Arcade Learning Environment (version 0.10.1+unknown)\n",
"[Powered by Stella]\n",
"/home/jovyan/Maturation/env-adaro-rl/lib/python3.10/site-packages/stable_baselines3/common/on_policy_algorithm.py:150: UserWarning: You are trying to run PPO on the GPU, but it is primarily intended to run on the CPU when not using a CNN policy (you are using ActorCriticPolicy which should be a MlpPolicy). See https://github.com/DLR-RM/stable-baselines3/issues/1245 for more info. You can pass `device='cpu'` or `export CUDA_VISIBLE_DEVICES=` to force using the CPU.Note: The model will train, but the GPU utilization will be poor and the training might take longer than on CPU.\n",
" warnings.warn(\n",
"A.L.E: Arcade Learning Environment (version 0.10.1+unknown)\n",
"[Powered by Stella]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"lengths : [6656, 6656]\n",
"mean : 6656.0\n",
"std : 0.0\n",
"rewards : [393.0, 434.0]\n",
"mean : 413.5\n",
"std : 20.5\n",
"save in outputs/Enduro-v5/attack_FGM_D_target_untargeted_eps_100.0_norm_2.0/adv_trained_agent/results/online_adv_attacks_norm_2.0_mean_reward.csv and outputs/Enduro-v5/attack_FGM_D_target_untargeted_eps_100.0_norm_2.0/adv_trained_agent/results/online_adv_attacks_norm_2.0_std_reward.csv\n"
]
}
],
"source": [
"print(f\"\\n🧨 Online attack on adversarially trained agent: {attack_name}_{target}_{eps}_{norm}\")\n",
"adaro_rl.pipelines.online_attack(\n",
" config=config,\n",
" attack_name=attack_name,\n",
" target=target,\n",
" eps=eps,\n",
" norm=norm,\n",
" output_dir=os.path.join(adv_trained_dir,\"results\"),\n",
" agent_checkpoint=os.path.join(adv_trained_dir,\"model.zip\"),\n",
" adversary_checkpoint=None,\n",
" self_reference=True,\n",
" device=device,\n",
" seed=seed,\n",
" n_eval_episodes=n_eval_episodes\n",
")"
]
},
{
"cell_type": "code",
"execution_count": 12,
"id": "5ad16b82",
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"attack_name = attack_name + \"_\" + target\n",
"fig = adaro_rl.viz.robustness_matrix(\n",
" env_name=config_name,\n",
" agent_dirs=[agent_dir, adv_trained_dir],\n",
" agent_names=[\"agent\", \"adv_trained_agent\"],\n",
" norm=norm,\n",
" attack_list=[attack_name],\n",
" eps_list=[eps],\n",
" output_dir=output_dir,\n",
" fontsize=15\n",
")\n",
"\n",
"fig.show()"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "55039884-b12e-4645-935e-0ac3538a07ca",
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "env-adaro-rl",
"language": "python",
"name": "env-adaro-rl"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.10.16"
}
},
"nbformat": 4,
"nbformat_minor": 5
}