notebook.ipynb

{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"name":"notebook.ipynb","provenance":[],"collapsed_sections":[]},"kernelspec":{"name":"python3","display_name":"Python 3"},"accelerator":"GPU"},"cells":[{"cell_type":"markdown","metadata":{"id":"uGRz9iW3qC1x","colab_type":"text"},"source":["# MuZero\n","\n","This notebook is just an example to show how to run [MuZero (https://github.com/sinaai/muzero)](https://github.com/sinaai/muzero) in Google Colab or Jupyter Notebook. You can also launch MuZero directly by cloning the github repository and running the command `python muzero.py`."]},{"cell_type":"code","metadata":{"id":"0CHey99GnA7n","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":1000},"executionInfo":{"status":"ok","timestamp":1596399359487,"user_tz":-270,"elapsed":57764,"user":{"displayName":"Sina Bagherinezhad","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14Gg0M0q3-4wmzD32zN50yrO-Lp5vSiEUYTYAkFBNCg=s64","userId":"10828062040523127623"}},"outputId":"d6b61ca7-c47d-439d-e5c4-6a8c553bc2a0"},"source":["!git clone https://github.com/sinaai/MuZero.git\n","%cd MuZero/\n","!pip install -r requirements.txt"],"execution_count":1,"outputs":[{"output_type":"stream","text":["Cloning into 'MuZero'...\n","remote: Enumerating objects: 8, done.\u001b[K\n","remote: Counting objects: 100% (8/8), done.\u001b[K\n","remote: Compressing objects: 100% (6/6), done.\u001b[K\n","remote: Total 8 (delta 0), reused 0 (delta 0), pack-reused 0\u001b[K\n","Unpacking objects: 100% (8/8), done.\n","/content/MuZero\n","Requirement already satisfied: numpy in /usr/local/lib/python3.6/dist-packages (from -r requirements.txt (line 1)) (1.18.5)\n","Collecting tensorflow==2.0.1\n","\u001b[?25l  Downloading https://files.pythonhosted.org/packages/43/16/b07e3f7a4a024b47918f7018967eb984b0c542458a6141d8c48515aa81d4/tensorflow-2.0.1-cp36-cp36m-manylinux2010_x86_64.whl (86.3MB)\n","\u001b[K     |████████████████████████████████| 86.3MB 40kB/s \n","\u001b[?25hRequirement already satisfied: gym in /usr/local/lib/python3.6/dist-packages (from -r requirements.txt (line 3)) (0.17.2)\n","Requirement already satisfied: six>=1.10.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.1->-r requirements.txt (line 2)) (1.15.0)\n","Requirement already satisfied: keras-applications>=1.0.8 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.1->-r requirements.txt (line 2)) (1.0.8)\n","Collecting gast==0.2.2\n","  Downloading https://files.pythonhosted.org/packages/4e/35/11749bf99b2d4e3cceb4d55ca22590b0d7c2c62b9de38ac4a4a7f4687421/gast-0.2.2.tar.gz\n","Requirement already satisfied: protobuf>=3.6.1 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.1->-r requirements.txt (line 2)) (3.12.2)\n","Requirement already satisfied: grpcio>=1.8.6 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.1->-r requirements.txt (line 2)) (1.30.0)\n","Requirement already satisfied: absl-py>=0.7.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.1->-r requirements.txt (line 2)) (0.9.0)\n","Requirement already satisfied: wrapt>=1.11.1 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.1->-r requirements.txt (line 2)) (1.12.1)\n","Requirement already satisfied: google-pasta>=0.1.6 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.1->-r requirements.txt (line 2)) (0.2.0)\n","Requirement already satisfied: opt-einsum>=2.3.2 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.1->-r requirements.txt (line 2)) (3.3.0)\n","Requirement already satisfied: wheel>=0.26; python_version >= \"3\" in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.1->-r requirements.txt (line 2)) (0.34.2)\n","Requirement already satisfied: keras-preprocessing>=1.0.5 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.1->-r requirements.txt (line 2)) (1.1.2)\n","Requirement already satisfied: termcolor>=1.1.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.1->-r requirements.txt (line 2)) (1.1.0)\n","Collecting tensorboard<2.1.0,>=2.0.0\n","\u001b[?25l  Downloading https://files.pythonhosted.org/packages/76/54/99b9d5d52d5cb732f099baaaf7740403e83fe6b0cedde940fabd2b13d75a/tensorboard-2.0.2-py3-none-any.whl (3.8MB)\n","\u001b[K     |████████████████████████████████| 3.8MB 38.9MB/s \n","\u001b[?25hRequirement already satisfied: astor>=0.6.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow==2.0.1->-r requirements.txt (line 2)) (0.8.1)\n","Collecting tensorflow-estimator<2.1.0,>=2.0.0\n","\u001b[?25l  Downloading https://files.pythonhosted.org/packages/fc/08/8b927337b7019c374719145d1dceba21a8bb909b93b1ad6f8fb7d22c1ca1/tensorflow_estimator-2.0.1-py2.py3-none-any.whl (449kB)\n","\u001b[K     |████████████████████████████████| 450kB 48.2MB/s \n","\u001b[?25hRequirement already satisfied: scipy in /usr/local/lib/python3.6/dist-packages (from gym->-r requirements.txt (line 3)) (1.4.1)\n","Requirement already satisfied: pyglet<=1.5.0,>=1.4.0 in /usr/local/lib/python3.6/dist-packages (from gym->-r requirements.txt (line 3)) (1.5.0)\n","Requirement already satisfied: cloudpickle<1.4.0,>=1.2.0 in /usr/local/lib/python3.6/dist-packages (from gym->-r requirements.txt (line 3)) (1.3.0)\n","Requirement already satisfied: h5py in /usr/local/lib/python3.6/dist-packages (from keras-applications>=1.0.8->tensorflow==2.0.1->-r requirements.txt (line 2)) (2.10.0)\n","Requirement already satisfied: setuptools in /usr/local/lib/python3.6/dist-packages (from protobuf>=3.6.1->tensorflow==2.0.1->-r requirements.txt (line 2)) (49.2.0)\n","Requirement already satisfied: markdown>=2.6.8 in /usr/local/lib/python3.6/dist-packages (from tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.1->-r requirements.txt (line 2)) (3.2.2)\n","Requirement already satisfied: google-auth-oauthlib<0.5,>=0.4.1 in /usr/local/lib/python3.6/dist-packages (from tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.1->-r requirements.txt (line 2)) (0.4.1)\n","Requirement already satisfied: requests<3,>=2.21.0 in /usr/local/lib/python3.6/dist-packages (from tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.1->-r requirements.txt (line 2)) (2.23.0)\n","Requirement already satisfied: google-auth<2,>=1.6.3 in /usr/local/lib/python3.6/dist-packages (from tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.1->-r requirements.txt (line 2)) (1.17.2)\n","Requirement already satisfied: werkzeug>=0.11.15 in /usr/local/lib/python3.6/dist-packages (from tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.1->-r requirements.txt (line 2)) (1.0.1)\n","Requirement already satisfied: future in /usr/local/lib/python3.6/dist-packages (from pyglet<=1.5.0,>=1.4.0->gym->-r requirements.txt (line 3)) (0.16.0)\n","Requirement already satisfied: importlib-metadata; python_version < \"3.8\" in /usr/local/lib/python3.6/dist-packages (from markdown>=2.6.8->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.1->-r requirements.txt (line 2)) (1.7.0)\n","Requirement already satisfied: requests-oauthlib>=0.7.0 in /usr/local/lib/python3.6/dist-packages (from google-auth-oauthlib<0.5,>=0.4.1->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.1->-r requirements.txt (line 2)) (1.3.0)\n","Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.6/dist-packages (from requests<3,>=2.21.0->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.1->-r requirements.txt (line 2)) (3.0.4)\n","Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.6/dist-packages (from requests<3,>=2.21.0->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.1->-r requirements.txt (line 2)) (2020.6.20)\n","Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.6/dist-packages (from requests<3,>=2.21.0->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.1->-r requirements.txt (line 2)) (2.10)\n","Requirement already satisfied: urllib3!=1.25.0,!=1.25.1,<1.26,>=1.21.1 in /usr/local/lib/python3.6/dist-packages (from requests<3,>=2.21.0->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.1->-r requirements.txt (line 2)) (1.24.3)\n","Requirement already satisfied: rsa<5,>=3.1.4; python_version >= \"3\" in /usr/local/lib/python3.6/dist-packages (from google-auth<2,>=1.6.3->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.1->-r requirements.txt (line 2)) (4.6)\n","Requirement already satisfied: cachetools<5.0,>=2.0.0 in /usr/local/lib/python3.6/dist-packages (from google-auth<2,>=1.6.3->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.1->-r requirements.txt (line 2)) (4.1.1)\n","Requirement already satisfied: pyasn1-modules>=0.2.1 in /usr/local/lib/python3.6/dist-packages (from google-auth<2,>=1.6.3->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.1->-r requirements.txt (line 2)) (0.2.8)\n","Requirement already satisfied: zipp>=0.5 in /usr/local/lib/python3.6/dist-packages (from importlib-metadata; python_version < \"3.8\"->markdown>=2.6.8->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.1->-r requirements.txt (line 2)) (3.1.0)\n","Requirement already satisfied: oauthlib>=3.0.0 in /usr/local/lib/python3.6/dist-packages (from requests-oauthlib>=0.7.0->google-auth-oauthlib<0.5,>=0.4.1->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.1->-r requirements.txt (line 2)) (3.1.0)\n","Requirement already satisfied: pyasn1>=0.1.3 in /usr/local/lib/python3.6/dist-packages (from rsa<5,>=3.1.4; python_version >= \"3\"->google-auth<2,>=1.6.3->tensorboard<2.1.0,>=2.0.0->tensorflow==2.0.1->-r requirements.txt (line 2)) (0.4.8)\n","Building wheels for collected packages: gast\n","  Building wheel for gast (setup.py) ... \u001b[?25l\u001b[?25hdone\n","  Created wheel for gast: filename=gast-0.2.2-cp36-none-any.whl size=7540 sha256=d5662abdc99bf359d0842ad2fedcfca65c2eb2e990de87624d1ba4a5e3f39d1f\n","  Stored in directory: /root/.cache/pip/wheels/5c/2e/7e/a1d4d4fcebe6c381f378ce7743a3ced3699feb89bcfbdadadd\n","Successfully built gast\n","\u001b[31mERROR: tensorflow-probability 0.10.0 has requirement gast>=0.3.2, but you'll have gast 0.2.2 which is incompatible.\u001b[0m\n","Installing collected packages: gast, tensorboard, tensorflow-estimator, tensorflow\n","  Found existing installation: gast 0.3.3\n","    Uninstalling gast-0.3.3:\n","      Successfully uninstalled gast-0.3.3\n","  Found existing installation: tensorboard 2.2.2\n","    Uninstalling tensorboard-2.2.2:\n","      Successfully uninstalled tensorboard-2.2.2\n","  Found existing installation: tensorflow-estimator 2.2.0\n","    Uninstalling tensorflow-estimator-2.2.0:\n","      Successfully uninstalled tensorflow-estimator-2.2.0\n","  Found existing installation: tensorflow 2.2.0\n","    Uninstalling tensorflow-2.2.0:\n","      Successfully uninstalled tensorflow-2.2.0\n","Successfully installed gast-0.2.2 tensorboard-2.0.2 tensorflow-2.0.1 tensorflow-estimator-2.0.1\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"y7ozabyBmk6z","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":1000},"executionInfo":{"status":"ok","timestamp":1596398137679,"user_tz":-270,"elapsed":2028765,"user":{"displayName":"Sina Bagherinezhad","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14Gg0M0q3-4wmzD32zN50yrO-Lp5vSiEUYTYAkFBNCg=s64","userId":"10828062040523127623"}},"outputId":"bdb8204f-4045-4b41-dae4-f9ea8aba8501"},"source":["from helper import SharedStorage, MuZeroConfig, ReplayBuffer, make_cartpole_config\n","from self_play import run_selfplay, run_eval\n","from training import train_network\n","\n","# MuZero training is split into two independent parts: Network training and\n","# self-play data generation.\n","# These two parts only communicate by transferring the latest network checkpoint\n","# from the training to the self-play, and the finished games from the self-play\n","# to the training.\n","train_scores = []\n","eval_scores = []\n","def muzero(config: MuZeroConfig):\n","\n","    storage = SharedStorage(config.new_network(), config.uniform_network(), config.new_optimizer())\n","    replay_buffer = ReplayBuffer(config)\n","\n","    for loop in range(config.nb_training_loop):\n","        print(\"Training loop\", loop)\n","        score_train = run_selfplay(config, storage, replay_buffer, config.nb_episodes)\n","        train_network(config, storage, replay_buffer, config.nb_epochs)\n","        score_eval = run_eval(config, storage, 20)\n","\n","        print(\"Train score:\", score_train)\n","        print(\"Eval score:\", score_eval)\n","        print(f\"MuZero played {config.nb_episodes * (loop + 1)} \"\n","              f\"episodes and trained for {config.nb_epochs * (loop + 1)} epochs.\\n\")\n","        train_scores.append(score_train)\n","        eval_scores.append(score_eval)\n","\n","    return storage.latest_network()\n","\n","config = make_cartpole_config()\n","muzero(config)"],"execution_count":9,"outputs":[{"output_type":"stream","text":["Training loop 0\n","2020-08-02 19:21:52.828729: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA\n","2020-08-02 19:21:52.832882: I tensorflow/core/platform/profile_utils/cpu_utils.cc:94] CPU Frequency: 2300000000 Hz\n","2020-08-02 19:21:52.833109: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x1d44bc0 executing computations on platform Host. Devices:\n","2020-08-02 19:21:52.833178: I tensorflow/compiler/xla/service/service.cc:175]   StreamExecutor device (0): Host, Default Version\n","WARNING:tensorflow:Layer initial_model is casting an input tensor from dtype float64 to the layer's dtype of float32, which is new behavior in TensorFlow 2.  The layer has dtype float32 because it's dtype defaults to floatx.\n","\n","If you intended to run this layer in float32, you can safely ignore this warning. If in doubt, this warning is likely only an issue if you are porting a TensorFlow 1.X model to TensorFlow 2.\n","\n","To change all layers to have dtype float64 by default, call `tf.keras.backend.set_floatx('float64')`. To change just this layer, pass dtype='float64' to the layer constructor. If you are the author of this layer, you can disable autocasting by passing autocast=False to the base Layer constructor.\n","\n","WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/ops/array_grad.py:502: _EagerTensorBase.cpu (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n","Instructions for updating:\n","Use tf.identity instead.\n","Train score: 18.75\n","Eval score: 9.6\n","MuZero played 20 episodes and trained for 20 epochs.\n","\n","Training loop 1\n","Train score: 21.9\n","Eval score: 9.6\n","MuZero played 40 episodes and trained for 40 epochs.\n","\n","Training loop 2\n","Train score: 10.35\n","Eval score: 9.2\n","MuZero played 60 episodes and trained for 60 epochs.\n","\n","Training loop 3\n","Train score: 10.1\n","Eval score: 47.2\n","MuZero played 80 episodes and trained for 80 epochs.\n","\n","Training loop 4\n","Train score: 19.1\n","Eval score: 135.3\n","MuZero played 100 episodes and trained for 100 epochs.\n","\n","Training loop 5\n","Train score: 40.95\n","Eval score: 112.8\n","MuZero played 120 episodes and trained for 120 epochs.\n","\n","Training loop 6\n","Train score: 75.3\n","Eval score: 145.3\n","MuZero played 140 episodes and trained for 140 epochs.\n","\n","Training loop 7\n","Train score: 105.8\n","Eval score: 200.0\n","MuZero played 160 episodes and trained for 160 epochs.\n","\n","Training loop 8\n","Train score: 157.35\n","Eval score: 200.0\n","MuZero played 180 episodes and trained for 180 epochs.\n","\n","Training loop 9\n","Train score: 143.1\n","Eval score: 114.9\n","MuZero played 200 episodes and trained for 200 epochs.\n","\n","Training loop 10\n","Train score: 123.2\n","Eval score: 145.4\n","MuZero played 220 episodes and trained for 220 epochs.\n","\n","Training loop 11\n","Train score: 133.8\n","Eval score: 118.3\n","MuZero played 240 episodes and trained for 240 epochs.\n","\n","Training loop 12\n","Train score: 116.9\n","Eval score: 147.7\n","MuZero played 260 episodes and trained for 260 epochs.\n","\n","Training loop 13\n","Train score: 144.75\n","Eval score: 148.4\n","MuZero played 280 episodes and trained for 280 epochs.\n","\n","Training loop 14\n","Train score: 141.6\n","Eval score: 125.4\n","MuZero played 300 episodes and trained for 300 epochs.\n","\n","Training loop 15\n","Train score: 125.4\n","Eval score: 200.0\n","MuZero played 320 episodes and trained for 320 epochs.\n","\n","Training loop 16\n","Train score: 200.0\n","Eval score: 141.7\n","MuZero played 340 episodes and trained for 340 epochs.\n","\n","Training loop 17\n","Train score: 141.0\n","Eval score: 191.0\n","MuZero played 360 episodes and trained for 360 epochs.\n","\n","Training loop 18\n","Train score: 192.35\n","Eval score: 153.0\n","MuZero played 380 episodes and trained for 380 epochs.\n","\n","Training loop 19\n","Train score: 148.75\n","Eval score: 200.0\n","MuZero played 400 episodes and trained for 400 epochs.\n","\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"agqqJD6_wAvn","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":279},"executionInfo":{"status":"ok","timestamp":1596399096946,"user_tz":-270,"elapsed":1091,"user":{"displayName":"Sina Bagherinezhad","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14Gg0M0q3-4wmzD32zN50yrO-Lp5vSiEUYTYAkFBNCg=s64","userId":"10828062040523127623"}},"outputId":"ab73f4fe-301e-4018-f1e3-8c23b8bbe631"},"source":["import matplotlib.pyplot as plt\n","\n","plt.plot(train_scores, color=\"red\")\n","plt.plot(eval_scores, color=\"blue\")\n","plt.xticks(range(20))\n","plt.xlabel('Loops')\n","plt.ylabel('Score')\n","plt.show()"],"execution_count":44,"outputs":[{"output_type":"display_data","data":{"image/png":"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\n","text/plain":["<Figure size 432x288 with 1 Axes>"]},"metadata":{"tags":[],"needs_background":"light"}}]}]}