Transformer - NASA.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Citation:**\n",
    "\n",
    "```\n",
    "@article{chen2022transformer,\n",
    "  title={Transformer network for remaining useful life prediction of lithium-ion batteries},\n",
    "  author={Chen, Daoquan and Hong, Weicong and Zhou, Xiuze},\n",
    "  journal={Ieee Access},\n",
    "  volume={10},\n",
    "  pages={19621--19628},\n",
    "  year={2022},\n",
    "  publisher={IEEE}\n",
    "}\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "D:\\Anaconda3\\envs\\torch_11\\lib\\site-packages\\numpy\\_distributor_init.py:32: UserWarning: loaded more than 1 DLL from .libs:\n",
      "D:\\Anaconda3\\envs\\torch_11\\lib\\site-packages\\numpy\\.libs\\libopenblas.WCDJNK7YVMPZQ2ME2ZZHJJRJ3JIKNDB7.gfortran-win_amd64.dll\n",
      "D:\\Anaconda3\\envs\\torch_11\\lib\\site-packages\\numpy\\.libs\\libopenblas.XWYDX2IKJW2NMTWSFYNGFUWKQU3LYTCZ.gfortran-win_amd64.dll\n",
      "  stacklevel=1)\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import random\n",
    "import math\n",
    "import os\n",
    "import scipy.io\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "import torchvision\n",
    "import transformers\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "from tqdm.notebook import tqdm\n",
    "from math import sqrt\n",
    "from datetime import datetime\n",
    "from sklearn.metrics import mean_absolute_error\n",
    "from sklearn.metrics import mean_squared_error"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 1. load data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# convert str to datatime \n",
    "def convert_to_time(hmm):\n",
    "    year, month, day, hour, minute, second = int(hmm[0]), int(hmm[1]), int(hmm[2]), int(hmm[3]), int(hmm[4]), int(hmm[5])\n",
    "    return datetime(year=year, month=month, day=day, hour=hour, minute=minute, second=second)\n",
    "\n",
    "\n",
    "# load .mat data\n",
    "def loadMat(matfile):\n",
    "    data = scipy.io.loadmat(matfile)\n",
    "    filename = matfile.split(\"/\")[-1].split(\".\")[0]\n",
    "    col = data[filename]\n",
    "    col = col[0][0][0][0]\n",
    "    size = col.shape[0]\n",
    "\n",
    "    data = []\n",
    "    for i in range(size):\n",
    "        k = list(col[i][3][0].dtype.fields.keys())\n",
    "        d1, d2 = {}, {}\n",
    "        if str(col[i][0][0]) != 'impedance':\n",
    "            for j in range(len(k)):\n",
    "                t = col[i][3][0][0][j][0];\n",
    "                l = [t[m] for m in range(len(t))]\n",
    "                d2[k[j]] = l\n",
    "        d1['type'], d1['temp'], d1['time'], d1['data'] = str(col[i][0][0]), int(col[i][1][0]), str(convert_to_time(col[i][2][0])), d2\n",
    "        data.append(d1)\n",
    "\n",
    "    return data\n",
    "\n",
    "\n",
    "# get capacity data\n",
    "def getBatteryCapacity(Battery):\n",
    "    cycle, capacity = [], []\n",
    "    i = 1\n",
    "    for Bat in Battery:\n",
    "        if Bat['type'] == 'discharge':\n",
    "            capacity.append(Bat['data']['Capacity'][0])\n",
    "            cycle.append(i)\n",
    "            i += 1\n",
    "    return [cycle, capacity]\n",
    "\n",
    "\n",
    "# get the charge data of a battery\n",
    "def getBatteryValues(Battery, Type='charge'):\n",
    "    data=[]\n",
    "    for Bat in Battery:\n",
    "        if Bat['type'] == Type:\n",
    "            data.append(Bat['data'])\n",
    "    return data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1.1 Extract raw data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Load Dataset B0005.mat ...\n",
      "Load Dataset B0006.mat ...\n",
      "Load Dataset B0007.mat ...\n",
      "Load Dataset B0018.mat ...\n"
     ]
    }
   ],
   "source": [
    "Battery_list = ['B0005', 'B0006', 'B0007', 'B0018']\n",
    "dir_path = 'datasets/NASA/'\n",
    "\n",
    "Battery = {}\n",
    "for name in Battery_list:\n",
    "    print('Load Dataset ' + name + '.mat ...')\n",
    "    path = dir_path + name + '.mat'\n",
    "    data = loadMat(path)\n",
    "    Battery[name] = getBatteryCapacity(data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## or 1.1 load the extracted data\n",
    "\n",
    "**note: If the above read fails, please load the .npy file.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "Battery_list = ['B0005', 'B0006', 'B0007', 'B0018']\n",
    "Battery = np.load('datasets/NASA/NASA.npy', allow_pickle=True)\n",
    "Battery = Battery.item()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1.2 plot figures for capacity degradation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x2612e7dc9e8>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(1, figsize=(12, 8))\n",
    "color_list = ['b:', 'g--', 'r-.', 'c.']\n",
    "c = 0\n",
    "for name,color in zip(Battery_list, color_list):\n",
    "    df_result = Battery[name]\n",
    "    ax.plot(df_result[0], df_result[1], color, label=name)\n",
    "ax.set(xlabel='Discharge cycles', ylabel='Capacity (Ah)', title='Capacity degradation at ambient temperature of 24°C')\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2. data processing and model\n",
    "\n",
    "## 2.1 data processing for model training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def build_instances(sequence, window_size):\n",
    "    #sequence: list of capacity\n",
    "    x, y = [],[]\n",
    "    for i in range(len(sequence) - window_size):\n",
    "        features = sequence[i:i+window_size]\n",
    "        target = sequence[i+window_size]\n",
    "\n",
    "        x.append(features)\n",
    "        y.append(target)\n",
    "        \n",
    "    return np.array(x).astype(np.float32), np.array(y).astype(np.float32)\n",
    "\n",
    "\n",
    "def split_dataset(data_sequence, train_ratio=0.0, capacity_threshold=0.0):\n",
    "    if capacity_threshold > 0:\n",
    "        max_capacity = max(data_sequence)\n",
    "        capacity = max_capacity * capacity_threshold\n",
    "        point = [i for i in range(len(data_sequence)) if data_sequence[i] < capacity]\n",
    "    else:\n",
    "        point = int(train_ratio + 1)\n",
    "        if 0 < train_ratio <= 1:\n",
    "            point = int(len(data_sequence) * train_ratio)\n",
    "    train_data, test_data = data_sequence[:point], data_sequence[point:]\n",
    "    \n",
    "    return train_data, test_data\n",
    "\n",
    "\n",
    "# leave-one-out evaluation: one battery is sampled randomly; the remainder are used for training.\n",
    "def get_train_test(data_dict, name, window_size=8):\n",
    "    data_sequence=data_dict[name][1]\n",
    "    train_data, test_data = data_sequence[:window_size+1], data_sequence[window_size+1:]\n",
    "    train_x, train_y = build_instances(train_data, window_size)\n",
    "    for k, v in data_dict.items():\n",
    "        if k != name:\n",
    "            data_x, data_y = build_instances(v[1], window_size)\n",
    "            train_x, train_y = np.r_[train_x, data_x], np.r_[train_y, data_y]\n",
    "            \n",
    "    return train_x, train_y, list(train_data), list(test_data)\n",
    "\n",
    "\n",
    "def relative_error(y_test, y_predict, threshold):\n",
    "    true_re, pred_re = len(y_test), 0\n",
    "    \n",
    "    for i in range(len(y_test)-1):\n",
    "        if y_test[i] <= threshold >= y_test[i+1]:\n",
    "            true_re = i - 1\n",
    "            break\n",
    "            \n",
    "    for i in range(len(y_predict)-1):\n",
    "        if y_predict[i] <= threshold:\n",
    "            pred_re = i - 1\n",
    "            break\n",
    "            \n",
    "    score = abs(true_re - pred_re)/true_re\n",
    "    if score > 1: score = 1\n",
    "        \n",
    "    return score\n",
    "\n",
    "\n",
    "def evaluation(y_test, y_predict):\n",
    "    mse = mean_squared_error(y_test, y_predict)\n",
    "    rmse = sqrt(mean_squared_error(y_test, y_predict))\n",
    "    return rmse\n",
    "    \n",
    "    \n",
    "def setup_seed(seed):\n",
    "    np.random.seed(seed)  # Numpy module.\n",
    "    random.seed(seed)  # Python random module.\n",
    "    os.environ['PYTHONHASHSEED'] = str(seed)\n",
    "    torch.manual_seed(seed) \n",
    "    if torch.cuda.is_available():\n",
    "        torch.cuda.manual_seed(seed) \n",
    "        torch.cuda.manual_seed_all(seed)\n",
    "        torch.backends.cudnn.benchmark = False\n",
    "        torch.backends.cudnn.deterministic = True"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.2 build net"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Autoencoder(nn.Module):\n",
    "    def __init__(self, input_size=16, hidden_dim=8, noise_level=0.01):\n",
    "        '''\n",
    "        Args:\n",
    "            input_size: the feature size of input data (required).\n",
    "            hidden_dim: the hidden size of AutoEncoder (required).\n",
    "            noise_level: the noise level added in Autoencoder (optional).\n",
    "        '''\n",
    "        super(Autoencoder, self).__init__()\n",
    "        self.noise_level = noise_level\n",
    "        self.fc1 = nn.Linear(input_size, hidden_dim)\n",
    "        self.fc2 = nn.Linear(hidden_dim, input_size)\n",
    "        \n",
    "    def encoder(self, x):\n",
    "        x = self.fc1(x)\n",
    "        h1 = F.relu(x)\n",
    "        return h1\n",
    "    \n",
    "    def mask(self, x):\n",
    "        corrupted_x = x + self.noise_level * torch.randn_like(x)\n",
    "        return corrupted_x\n",
    "    \n",
    "    def decoder(self, x):\n",
    "        h2 = self.fc2(x)\n",
    "        return h2\n",
    "    \n",
    "    def forward(self, x):\n",
    "        out = self.mask(x)\n",
    "        encode = self.encoder(out)\n",
    "        decode = self.decoder(encode)\n",
    "        return encode, decode\n",
    "    \n",
    "    \n",
    "class PositionalEncoding(nn.Module):\n",
    "    def __init__(self, feature_len, feature_size, dropout=0.0):\n",
    "        '''\n",
    "        Args:\n",
    "            feature_len: the feature length of input data (required).\n",
    "            feature_size: the feature size of input data (required).\n",
    "            dropout: the dropout rate (optional).\n",
    "        '''\n",
    "        super(PositionalEncoding, self).__init__()\n",
    "\n",
    "        pe = torch.zeros(feature_len, feature_size)\n",
    "        position = torch.arange(0, feature_len, dtype=torch.float).unsqueeze(1)\n",
    "\n",
    "        div_term = torch.exp(torch.arange(0, feature_size, 2).float() * (-math.log(10000.0) / feature_size))\n",
    "\n",
    "        pe[:, 0::2] = torch.sin(position * div_term)\n",
    "        pe[:, 1::2] = torch.cos(position * div_term)\n",
    "\n",
    "        pe = pe.unsqueeze(0)\n",
    "\n",
    "        self.register_buffer('pe', pe)\n",
    "\n",
    "    def forward(self, x):\n",
    "        x = x + self.pe\n",
    "        \n",
    "        return x\n",
    "    \n",
    "\n",
    "class Net(nn.Module):\n",
    "    def __init__(self, feature_size=16, hidden_dim=32, feature_num=1, num_layers=1, nhead=1, dropout=0.0, noise_level=0.01):\n",
    "        '''\n",
    "        Args:\n",
    "            feature_size: the feature size of input data (required).\n",
    "            hidden_dim: the hidden size of Transformer block (required).\n",
    "            feature_num: the number of features, such as capacity, voltage, and current; set 1 for only sigle feature (optional).\n",
    "            num_layers: the number of layers of Transformer block (optional).\n",
    "            nhead: the number of heads of multi-attention in Transformer block (optional).\n",
    "            dropout: the dropout rate of Transformer block (optional).\n",
    "            noise_level: the noise level added in Autoencoder (optional).\n",
    "        '''\n",
    "        super(Net, self).__init__()\n",
    "        self.auto_hidden = int(feature_size / 2)\n",
    "        input_size = self.auto_hidden \n",
    "        \n",
    "        if feature_num == 1:\n",
    "            # Transformer treated as an Encoder when modeling for a sigle feature like only capacity data\n",
    "            self.pos = PositionalEncoding(feature_len=feature_num, feature_size=input_size)\n",
    "            encoder_layers = nn.TransformerEncoderLayer(d_model=input_size, nhead=nhead, dim_feedforward=hidden_dim, dropout=dropout, batch_first=True)\n",
    "        elif feature_num > 1:\n",
    "            # Transformer treated as a sequence model when modeling for multi-features like capacity, voltage, and current data\n",
    "            self.pos = PositionalEncoding(feature_len=input_size, feature_size=feature_num)\n",
    "            encoder_layers = nn.TransformerEncoderLayer(d_model=feature_num, nhead=nhead, dim_feedforward=hidden_dim, dropout=dropout, batch_first=True)\n",
    "        self.cell = nn.TransformerEncoder(encoder_layers, num_layers=num_layers)\n",
    "        \n",
    "        self.linear = nn.Linear(feature_num*self.auto_hidden, 1)\n",
    "        self.autoencoder = Autoencoder(input_size=feature_size, hidden_dim=self.auto_hidden, noise_level=noise_level)\n",
    " \n",
    "    def forward(self, x): \n",
    "        batch_size, feature_num, feature_size  = x.shape\n",
    "        out, decode = self.autoencoder(x)\n",
    "        if feature_num > 1:\n",
    "            out = out.reshape(batch_size, -1, feature_num)\n",
    "        out = self.pos(out)\n",
    "        out = self.cell(out)              # sigle feature: (batch_size, feature_num, auto_hidden) or multi-features: (batch_size, auto_hidden, feature_num)\n",
    "        out = out.reshape(batch_size, -1) # (batch_size, feature_num*auto_hidden)\n",
    "        out = self.linear(out)            # out shape: (batch_size, 1)\n",
    "        \n",
    "        return out, decode"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.3 train for model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def train(lr=0.01, feature_size=8, feature_num=1, hidden_dim=32, num_layers=1, nhead=1, dropout=0.0, epochs=1000, \n",
    "          weight_decay=0.0, seed=0, alpha=0.0, noise_level=0.0, metric='re', device='cpu'):\n",
    "    '''\n",
    "        Args:\n",
    "            lr: learning rate for training (required).\n",
    "            feature_size: the feature size of input data (required).\n",
    "            feature_num: the number of features, such as capacity, voltage, and current; set 1 for only sigle feature (optional).\n",
    "            hidden_dim: the hidden size of Transformer block (required).\n",
    "            num_layers: the number of layers of Transformer block (optional).\n",
    "            nhead: the number of heads of multi-attention in Transformer block (optional).\n",
    "            dropout: the dropout rate of Transformer block (optional).\n",
    "            epochs: \n",
    "            weight_decay: \n",
    "            seed: (optional).\n",
    "            alpha: (optional).\n",
    "            noise_level: the noise level added in Autoencoder (optional).\n",
    "            metric: (optional).\n",
    "            device: the device for training (optional).\n",
    "        '''\n",
    "    score_list, fixed_result_list, moving_result_list = [], [], []\n",
    "    setup_seed(seed)\n",
    "    for i in range(4):\n",
    "        name = Battery_list[i]\n",
    "        train_x, train_y, train_data, test_data = get_train_test(Battery, name, feature_size)\n",
    "        test_sequence = train_data + test_data\n",
    "        # print('sample size: {}'.format(len(train_x)))\n",
    "\n",
    "        model = Net(feature_size=feature_size, hidden_dim=hidden_dim, feature_num=K, num_layers=num_layers, \n",
    "                    nhead=nhead, dropout=dropout, noise_level=noise_level)\n",
    "        model = model.to(device)\n",
    "        optimizer = torch.optim.Adam(model.parameters(), lr=lr, weight_decay=weight_decay)\n",
    "        criterion = nn.MSELoss()\n",
    "\n",
    "        test_x = train_data.copy()\n",
    "        loss_list, y_fixed_slice, y_moving_slice = [0], [], []\n",
    "        rmse, re = 1, 1\n",
    "        score_, score = [1],[1]\n",
    "        for epoch in range(epochs):\n",
    "            x, y = np.reshape(train_x/Rated_Capacity,(-1, feature_num, feature_size)), np.reshape(train_y/Rated_Capacity,(-1,1))\n",
    "            x, y = torch.from_numpy(x).to(device), torch.from_numpy(y).to(device)\n",
    "            x = x.repeat(1, K, 1)\n",
    "            output, decode = model(x)\n",
    "            output = output.reshape(-1, 1)\n",
    "            loss = criterion(output, y) + alpha * criterion(decode, x)\n",
    "            optimizer.zero_grad() \n",
    "            loss.backward()  \n",
    "            optimizer.step()                   \n",
    "\n",
    "            if (epoch + 1)%10 == 0:\n",
    "                test_x = train_data.copy() \n",
    "                fixed_point_list, moving_point_list = [], []\n",
    "                t = 0\n",
    "                while (len(test_x) - len(train_data)) < len(test_data):\n",
    "                    x = np.reshape(np.array(test_x[-feature_size:])/Rated_Capacity,(-1, feature_num, feature_size)).astype(np.float32)\n",
    "                    x = torch.from_numpy(x).to(device) \n",
    "                    x = x.repeat(1, K, 1)\n",
    "                    pred, _ = model(x) \n",
    "                    next_point = pred.data.cpu().numpy()[0,0] * Rated_Capacity\n",
    "                    test_x.append(next_point)      # The test values are added to the original sequence to continue to predict the next point\n",
    "                    fixed_point_list.append(next_point) # Saves the predicted value of the last point in the output sequence\n",
    "                    \n",
    "                    x = np.reshape(np.array(test_sequence[t:t+feature_size])/Rated_Capacity,(-1, 1, feature_size)).astype(np.float32)\n",
    "                    x = torch.from_numpy(x).to(device) \n",
    "                    x = x.repeat(1, K, 1)\n",
    "                    pred, _ = model(x) \n",
    "                    next_point = pred.data.cpu().numpy()[0,0] * Rated_Capacity\n",
    "                    moving_point_list.append(next_point) # Saves the predicted value of the last point in the output sequence\n",
    "                    t += 1\n",
    "                    \n",
    "                y_fixed_slice.append(fixed_point_list)             # Save all the predicted values\n",
    "                y_moving_slice.append(moving_point_list)\n",
    "\n",
    "                loss_list.append(loss)\n",
    "                rmse = evaluation(y_test=test_data, y_predict=y_fixed_slice[-1])\n",
    "                re = relative_error(y_test=test_data, y_predict=y_fixed_slice[-1], threshold=Rated_Capacity*0.7)\n",
    "                #print('epoch:{:<2d} | loss:{:<6.4f} | RMSE:{:<6.4f} | RE:{:<6.4f}'.format(epoch, loss, rmse, re))\n",
    "                \n",
    "            if metric == 're':\n",
    "                score = [re]\n",
    "            elif metric == 'rmse':\n",
    "                score = [rmse]\n",
    "            else:\n",
    "                score = [re, rmse]\n",
    "            if (loss < 1e-3) and (score_[0] < score[0]):\n",
    "                break\n",
    "            score_ = score.copy()\n",
    "            \n",
    "        score_list.append(score_)\n",
    "        fixed_result_list.append(train_data.copy() + y_fixed_slice[-1])\n",
    "        moving_result_list.append(train_data.copy() + y_moving_slice[-1])\n",
    "        \n",
    "    return score_list, fixed_result_list, moving_result_list"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.4 setting and training for overall performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "148826b4d06542e0a000b7c35c0c4dfd",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/5 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "seed:0\n",
      "[[0.09433962]\n",
      " [0.12222222]\n",
      " [0.03311258]\n",
      " [0.06410256]]\n",
      "refor this seed: 0.0784\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.0754717 ]\n",
      " [0.63333333]\n",
      " [0.09271523]\n",
      " [0.15384615]]\n",
      "refor this seed: 0.2388\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.23584906]\n",
      " [0.15555556]\n",
      " [0.15231788]\n",
      " [0.43589744]]\n",
      "refor this seed: 0.2449\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.06603774]\n",
      " [0.06666667]\n",
      " [0.16556291]\n",
      " [0.38461538]]\n",
      "refor this seed: 0.1707\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.12222222]\n",
      " [0.2384106 ]\n",
      " [0.29487179]]\n",
      "refor this seed: 0.1639\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.1794\n"
     ]
    }
   ],
   "source": [
    "Rated_Capacity = 2.0\n",
    "feature_size = 16\n",
    "feature_num = 1\n",
    "dropout = 0.0\n",
    "epochs = 2000\n",
    "nhead = 1\n",
    "hidden_dim = 16\n",
    "num_layers = 2\n",
    "lr = 0.01               \n",
    "weight_decay = 0.0\n",
    "noise_level = 0.0\n",
    "alpha = 0.01\n",
    "metric = 're'\n",
    "device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')\n",
    "\n",
    "K = 16\n",
    "\n",
    "SCORE = []\n",
    "for seed in tqdm(range(5)):\n",
    "    print('seed:{}'.format(seed))\n",
    "    score_list, _, _ = train(lr=lr, feature_size=feature_size, feature_num=feature_num, hidden_dim=hidden_dim, num_layers=num_layers, \n",
    "                             nhead=nhead, weight_decay=weight_decay, epochs=epochs, seed=seed, dropout=dropout, alpha=alpha, \n",
    "                             noise_level=noise_level, metric=metric, device=device)\n",
    "    print(np.array(score_list))\n",
    "    print(metric + 'for this seed: {:<6.4f}'.format(np.mean(np.array(score_list))))\n",
    "    for s in score_list:\n",
    "        SCORE.append(s)\n",
    "    print('------------------------------------------------------------------')\n",
    "print(metric + ' mean: {:<6.4f}'.format(np.mean(np.array(SCORE))))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.5 prediction for figures"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "seed:0\n",
      "[[0.09433962]\n",
      " [0.12222222]\n",
      " [0.03311258]\n",
      " [0.06410256]]\n",
      "refor this seed: 0.0784\n"
     ]
    }
   ],
   "source": [
    "Rated_Capacity = 2.0\n",
    "feature_size = 16\n",
    "feature_num = 1\n",
    "dropout = 0.0\n",
    "epochs = 2000\n",
    "nhead = 1\n",
    "hidden_dim = 16\n",
    "num_layers = 2\n",
    "lr = 0.01               \n",
    "weight_decay = 0.0\n",
    "noise_level = 0.0\n",
    "alpha = 0.01\n",
    "metric = 're'\n",
    "device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')\n",
    "K = 16\n",
    "seed = 0\n",
    "\n",
    "SCORE = []\n",
    "print('seed:{}'.format(seed))\n",
    "score_list, fixed_result_list, moving_result_list = train(lr=lr, feature_size=feature_size, feature_num=feature_num, \n",
    "                                                           hidden_dim=hidden_dim, num_layers=num_layers, nhead=nhead, \n",
    "                                                           weight_decay=weight_decay, epochs=epochs, seed=seed, dropout=dropout, \n",
    "                                                           alpha=alpha, noise_level=noise_level, metric=metric, device=device)\n",
    "print(np.array(score_list))\n",
    "print(metric + 'for this seed: {:<6.4f}'.format(np.mean(np.array(score_list))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1728x1296 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig,ax = plt.subplots(2, 2, figsize=(24, 18))\n",
    "\n",
    "for i in range(2):\n",
    "    for j in range(2):\n",
    "        t = i + j\n",
    "        battery_name = Battery_list[t]\n",
    "        test_data = Battery[battery_name][1]\n",
    "        fixed_predict_data = fixed_result_list[t]\n",
    "        moving_predict_data = moving_result_list[t]\n",
    "        x = [t for t in range(len(test_data))]\n",
    "        threshold = [Rated_Capacity*0.7] * len(test_data)\n",
    "        ax[i][j].plot(x, test_data, 'c', label='test data')\n",
    "        ax[i][j].plot(x, fixed_predict_data, 'b', label='fixed predicted data')\n",
    "        ax[i][j].plot(x, moving_predict_data, 'r', label='moving predicted data')\n",
    "        ax[i][j].plot(x, threshold, 'black', ls=':', label='stop line')\n",
    "        ax[i][j].legend()\n",
    "        ax[i][j].set_xlabel('Discharge cycles', fontsize=15)\n",
    "        ax[i][j].set_ylabel('Capacity (Ah)', fontsize=15)\n",
    "        ax[i][j].set_title('test v.s. prediction of battery ' + battery_name, fontsize=20)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3. use grid-search to determ parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "afa2190086ad49b89a3b65ae481c440d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/7 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9f55e93990ea40a6838eae039b39337e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "aa9220d89eb243bca30935d72b91bb6b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "K=1, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.10377358]\n",
      " [0.63333333]\n",
      " [0.26490066]\n",
      " [1.        ]]\n",
      "re: 0.5005\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.25471698]\n",
      " [0.64444444]\n",
      " [0.39735099]\n",
      " [0.38461538]]\n",
      "re: 0.4203\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.16037736]\n",
      " [0.56666667]\n",
      " [0.02649007]\n",
      " [1.        ]]\n",
      "re: 0.4384\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.        ]\n",
      " [0.64444444]\n",
      " [1.        ]\n",
      " [1.        ]]\n",
      "re: 0.6611\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.32075472]\n",
      " [0.43333333]\n",
      " [0.23178808]\n",
      " [0.29487179]]\n",
      "re: 0.3202\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.4681\n",
      "===================================================================\n",
      "K=1, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.10377358]\n",
      " [0.63333333]\n",
      " [0.26490066]\n",
      " [1.        ]]\n",
      "re: 0.5005\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.25471698]\n",
      " [0.64444444]\n",
      " [0.05960265]\n",
      " [0.43589744]]\n",
      "re: 0.3487\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.16981132]\n",
      " [0.62222222]\n",
      " [0.08609272]\n",
      " [0.37179487]]\n",
      "re: 0.3125\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.        ]\n",
      " [0.64444444]\n",
      " [1.        ]\n",
      " [1.        ]]\n",
      "re: 0.6611\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.17924528]\n",
      " [0.62222222]\n",
      " [0.34437086]\n",
      " [0.14102564]]\n",
      "re: 0.3217\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.4289\n",
      "===================================================================\n",
      "K=1, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.10377358]\n",
      " [0.61111111]\n",
      " [0.25827815]\n",
      " [1.        ]]\n",
      "re: 0.4933\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.25471698]\n",
      " [0.43333333]\n",
      " [0.04635762]\n",
      " [0.37179487]]\n",
      "re: 0.2766\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.46666667]\n",
      " [1.        ]\n",
      " [0.14102564]]\n",
      "re: 0.4019\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.        ]\n",
      " [0.64444444]\n",
      " [0.15231788]\n",
      " [1.        ]]\n",
      "re: 0.4492\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.16981132]\n",
      " [0.43333333]\n",
      " [0.1589404 ]\n",
      " [0.41025641]]\n",
      "re: 0.2931\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3828\n",
      "===================================================================\n",
      "K=1, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.06603774]\n",
      " [0.6       ]\n",
      " [0.0794702 ]\n",
      " [1.        ]]\n",
      "re: 0.4364\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.26415094]\n",
      " [0.23333333]\n",
      " [0.08609272]\n",
      " [0.06410256]]\n",
      "re: 0.1619\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.01886792]\n",
      " [0.52222222]\n",
      " [0.12582781]\n",
      " [0.26923077]]\n",
      "re: 0.2340\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.        ]\n",
      " [0.52222222]\n",
      " [0.33774834]\n",
      " [0.3974359 ]]\n",
      "re: 0.3144\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.24528302]\n",
      " [0.48888889]\n",
      " [0.03311258]\n",
      " [0.75641026]]\n",
      "re: 0.3809\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3055\n",
      "===================================================================\n",
      "K=1, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.63207547]\n",
      " [1.        ]\n",
      " [0.24503311]\n",
      " [0.03846154]]\n",
      "re: 0.4789\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.06603774]\n",
      " [0.5       ]\n",
      " [0.25827815]\n",
      " [0.17948718]]\n",
      "re: 0.2510\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.        ]\n",
      " [0.5       ]\n",
      " [0.80794702]\n",
      " [0.34615385]]\n",
      "re: 0.6635\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.10377358]\n",
      " [0.61111111]\n",
      " [0.0794702 ]\n",
      " [0.14102564]]\n",
      "re: 0.2338\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.01886792]\n",
      " [1.        ]\n",
      " [0.27152318]\n",
      " [0.23076923]]\n",
      "re: 0.3803\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.4015\n",
      "===================================================================\n",
      "K=1, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.63207547]\n",
      " [1.        ]\n",
      " [0.24503311]\n",
      " [0.29487179]]\n",
      "re: 0.5430\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.01886792]\n",
      " [0.42222222]\n",
      " [0.1986755 ]\n",
      " [1.        ]]\n",
      "re: 0.4099\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.        ]\n",
      " [0.5       ]\n",
      " [0.21192053]\n",
      " [0.79487179]]\n",
      "re: 0.6267\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.12264151]\n",
      " [0.53333333]\n",
      " [0.1589404 ]\n",
      " [0.43589744]]\n",
      "re: 0.3127\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.01886792]\n",
      " [1.        ]\n",
      " [0.24503311]\n",
      " [0.32051282]]\n",
      "re: 0.3961\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.4577\n",
      "===================================================================\n",
      "K=1, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.63207547]\n",
      " [1.        ]\n",
      " [0.02649007]\n",
      " [0.15384615]]\n",
      "re: 0.4531\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.08490566]\n",
      " [0.5       ]\n",
      " [0.01986755]\n",
      " [0.43589744]]\n",
      "re: 0.2602\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.        ]\n",
      " [0.5       ]\n",
      " [0.05960265]\n",
      " [0.71794872]]\n",
      "re: 0.5694\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.10377358]\n",
      " [0.64444444]\n",
      " [0.0794702 ]\n",
      " [0.38461538]]\n",
      "re: 0.3031\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.01886792]\n",
      " [0.63333333]\n",
      " [0.16556291]\n",
      " [0.23076923]]\n",
      "re: 0.2621\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3696\n",
      "===================================================================\n",
      "K=1, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.00943396]\n",
      " [0.62222222]\n",
      " [0.02649007]\n",
      " [0.14102564]]\n",
      "re: 0.1998\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.58888889]\n",
      " [0.2384106 ]\n",
      " [0.11538462]]\n",
      "re: 0.2357\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.        ]\n",
      " [0.45555556]\n",
      " [0.16556291]\n",
      " [0.35897436]]\n",
      "re: 0.4950\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.10377358]\n",
      " [0.58888889]\n",
      " [0.28476821]\n",
      " [0.28205128]]\n",
      "re: 0.3149\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.01886792]\n",
      " [0.63333333]\n",
      " [0.09933775]\n",
      " [0.41025641]]\n",
      "re: 0.2904\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3072\n",
      "===================================================================\n",
      "K=1, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.23584906]\n",
      " [0.56666667]\n",
      " [0.06622517]\n",
      " [0.41025641]]\n",
      "re: 0.3197\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.37735849]\n",
      " [0.61111111]\n",
      " [0.03311258]\n",
      " [0.06410256]]\n",
      "re: 0.2714\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.10377358]\n",
      " [0.63333333]\n",
      " [0.1986755 ]\n",
      " [0.32051282]]\n",
      "re: 0.3141\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.01886792]\n",
      " [0.4       ]\n",
      " [0.25827815]\n",
      " [0.28205128]]\n",
      "re: 0.2398\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.22641509]\n",
      " [0.56666667]\n",
      " [0.1589404 ]\n",
      " [0.07692308]]\n",
      "re: 0.2572\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2805\n",
      "===================================================================\n",
      "K=1, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.23584906]\n",
      " [0.56666667]\n",
      " [0.06622517]\n",
      " [0.41025641]]\n",
      "re: 0.3197\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.37735849]\n",
      " [0.61111111]\n",
      " [0.04635762]\n",
      " [0.06410256]]\n",
      "re: 0.2747\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.11320755]\n",
      " [0.64444444]\n",
      " [0.21854305]\n",
      " [1.        ]]\n",
      "re: 0.4940\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.01886792]\n",
      " [0.31111111]\n",
      " [0.55629139]\n",
      " [1.        ]]\n",
      "re: 0.4716\n",
      "------------------------------------------------------------------\n",
      "seed:4\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.22641509]\n",
      " [0.56666667]\n",
      " [0.1589404 ]\n",
      " [0.07692308]]\n",
      "re: 0.2572\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3635\n",
      "===================================================================\n",
      "K=1, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.23584906]\n",
      " [0.56666667]\n",
      " [0.08609272]\n",
      " [0.28205128]]\n",
      "re: 0.2927\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.37735849]\n",
      " [0.61111111]\n",
      " [0.33112583]\n",
      " [0.20512821]]\n",
      "re: 0.3812\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.16037736]\n",
      " [0.48888889]\n",
      " [0.1986755 ]\n",
      " [0.21794872]]\n",
      "re: 0.2665\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.01886792]\n",
      " [0.51111111]\n",
      " [0.05960265]\n",
      " [1.        ]]\n",
      "re: 0.3974\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.23584906]\n",
      " [0.56666667]\n",
      " [0.05298013]\n",
      " [0.06410256]]\n",
      "re: 0.2299\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3135\n",
      "===================================================================\n",
      "K=1, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.01886792]\n",
      " [0.57777778]\n",
      " [0.09933775]\n",
      " [0.26923077]]\n",
      "re: 0.2413\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.38679245]\n",
      " [0.6       ]\n",
      " [0.33112583]\n",
      " [0.33333333]]\n",
      "re: 0.4128\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.02830189]\n",
      " [0.53333333]\n",
      " [0.29801325]\n",
      " [0.35897436]]\n",
      "re: 0.3047\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.04716981]\n",
      " [0.52222222]\n",
      " [0.16556291]\n",
      " [0.29487179]]\n",
      "re: 0.2575\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.27358491]\n",
      " [0.58888889]\n",
      " [0.03311258]\n",
      " [0.41025641]]\n",
      "re: 0.3265\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3085\n",
      "===================================================================\n",
      "K=1, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.56603774]\n",
      " [0.54444444]\n",
      " [0.28476821]\n",
      " [0.43589744]]\n",
      "re: 0.4578\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.05660377]\n",
      " [0.44444444]\n",
      " [0.29801325]\n",
      " [0.16666667]]\n",
      "re: 0.2414\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.72641509]\n",
      " [0.47777778]\n",
      " [0.23178808]\n",
      " [0.32051282]]\n",
      "re: 0.4391\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.05660377]\n",
      " [0.35555556]\n",
      " [0.21854305]\n",
      " [0.29487179]]\n",
      "re: 0.2314\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.51111111]\n",
      " [0.24503311]\n",
      " [0.35897436]]\n",
      "re: 0.2788\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3297\n",
      "===================================================================\n",
      "K=1, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.56603774]\n",
      " [0.54444444]\n",
      " [0.28476821]\n",
      " [0.43589744]]\n",
      "re: 0.4578\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.05660377]\n",
      " [0.44444444]\n",
      " [0.29801325]\n",
      " [0.33333333]]\n",
      "re: 0.2831\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.74528302]\n",
      " [0.3       ]\n",
      " [0.33112583]\n",
      " [0.3974359 ]]\n",
      "re: 0.4435\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.05660377]\n",
      " [0.35555556]\n",
      " [0.21854305]\n",
      " [0.29487179]]\n",
      "re: 0.2314\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.52222222]\n",
      " [0.25827815]\n",
      " [0.37179487]]\n",
      "re: 0.2881\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3408\n",
      "===================================================================\n",
      "K=1, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.56603774]\n",
      " [0.54444444]\n",
      " [0.30463576]\n",
      " [0.34615385]]\n",
      "re: 0.4403\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.02830189]\n",
      " [0.04444444]\n",
      " [0.34437086]\n",
      " [1.        ]]\n",
      "re: 0.3543\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.78301887]\n",
      " [0.83333333]\n",
      " [0.06622517]\n",
      " [0.19230769]]\n",
      "re: 0.4687\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.04716981]\n",
      " [0.5       ]\n",
      " [0.05298013]\n",
      " [0.01282051]]\n",
      "re: 0.1532\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.52222222]\n",
      " [0.18543046]\n",
      " [0.38461538]]\n",
      "re: 0.2731\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3379\n",
      "===================================================================\n",
      "K=1, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.        ]\n",
      " [0.56666667]\n",
      " [0.13245033]\n",
      " [0.28205128]]\n",
      "re: 0.2453\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.02830189]\n",
      " [0.48888889]\n",
      " [0.21192053]\n",
      " [0.19230769]]\n",
      "re: 0.2304\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.68867925]\n",
      " [0.4       ]\n",
      " [0.11258278]\n",
      " [0.26923077]]\n",
      "re: 0.3676\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [0.6       ]\n",
      " [0.1589404 ]\n",
      " [0.28205128]]\n",
      "re: 0.2626\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.54444444]\n",
      " [0.12582781]\n",
      " [0.14102564]]\n",
      "re: 0.2028\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2617\n",
      "===================================================================\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5c96504967d04f5f94bf1b81e4d35c10",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "K=1, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.0754717 ]\n",
      " [0.58888889]\n",
      " [0.37748344]\n",
      " [0.        ]]\n",
      "re: 0.2605\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.22641509]\n",
      " [0.4       ]\n",
      " [0.21192053]\n",
      " [0.34615385]]\n",
      "re: 0.2961\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.33962264]\n",
      " [0.17777778]\n",
      " [0.31125828]\n",
      " [1.        ]]\n",
      "re: 0.4572\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.14150943]\n",
      " [0.64444444]\n",
      " [0.26490066]\n",
      " [0.37179487]]\n",
      "re: 0.3557\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.23584906]\n",
      " [0.51111111]\n",
      " [0.07284768]\n",
      " [0.25641026]]\n",
      "re: 0.2691\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3277\n",
      "===================================================================\n",
      "K=1, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.05660377]\n",
      " [0.57777778]\n",
      " [0.37748344]\n",
      " [0.03846154]]\n",
      "re: 0.2626\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.22641509]\n",
      " [0.38888889]\n",
      " [0.13245033]\n",
      " [0.35897436]]\n",
      "re: 0.2767\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.35849057]\n",
      " [0.01111111]\n",
      " [0.50993377]\n",
      " [0.41025641]]\n",
      "re: 0.3224\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.14150943]\n",
      " [0.64444444]\n",
      " [0.29139073]\n",
      " [0.01282051]]\n",
      "re: 0.2725\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.39622642]\n",
      " [0.02222222]\n",
      " [0.32450331]\n",
      " [0.32051282]]\n",
      "re: 0.2659\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2800\n",
      "===================================================================\n",
      "K=1, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.05660377]\n",
      " [0.58888889]\n",
      " [0.06622517]\n",
      " [1.        ]]\n",
      "re: 0.4279\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.22641509]\n",
      " [0.42222222]\n",
      " [0.62913907]\n",
      " [0.11538462]]\n",
      "re: 0.3483\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.45283019]\n",
      " [0.42222222]\n",
      " [0.50993377]\n",
      " [0.12820513]]\n",
      "re: 0.3783\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.13207547]\n",
      " [0.64444444]\n",
      " [0.10596026]\n",
      " [0.3974359 ]]\n",
      "re: 0.3200\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.37735849]\n",
      " [0.01111111]\n",
      " [0.32450331]\n",
      " [0.12820513]]\n",
      "re: 0.2103\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3370\n",
      "===================================================================\n",
      "K=1, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.03773585]\n",
      " [0.57777778]\n",
      " [0.25827815]\n",
      " [0.29487179]]\n",
      "re: 0.2922\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.13333333]\n",
      " [0.21192053]\n",
      " [0.03846154]]\n",
      "re: 0.0959\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.22641509]\n",
      " [0.31111111]\n",
      " [1.        ]\n",
      " [0.17948718]]\n",
      "re: 0.4293\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.13207547]\n",
      " [0.64444444]\n",
      " [0.11258278]\n",
      " [0.38461538]]\n",
      "re: 0.3184\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.28301887]\n",
      " [0.42222222]\n",
      " [0.14569536]\n",
      " [0.11538462]]\n",
      "re: 0.2416\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2755\n",
      "===================================================================\n",
      "K=1, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.0754717 ]\n",
      " [0.26666667]\n",
      " [1.        ]\n",
      " [0.52564103]]\n",
      "re: 0.4669\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.05660377]\n",
      " [0.32222222]\n",
      " [0.55629139]\n",
      " [0.06410256]]\n",
      "re: 0.2498\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.        ]\n",
      " [0.07777778]\n",
      " [0.10596026]\n",
      " [0.24358974]]\n",
      "re: 0.3568\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.39622642]\n",
      " [0.57777778]\n",
      " [0.76821192]\n",
      " [0.02564103]]\n",
      "re: 0.4420\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.24444444]\n",
      " [0.0397351 ]\n",
      " [0.01282051]]\n",
      "re: 0.0743\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3180\n",
      "===================================================================\n",
      "K=1, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.0754717 ]\n",
      " [0.26666667]\n",
      " [1.        ]\n",
      " [0.21794872]]\n",
      "re: 0.3900\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.05660377]\n",
      " [0.1       ]\n",
      " [0.8013245 ]\n",
      " [0.3974359 ]]\n",
      "re: 0.3388\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.        ]\n",
      " [0.06666667]\n",
      " [0.11258278]\n",
      " [0.34615385]]\n",
      "re: 0.3814\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.39622642]\n",
      " [0.58888889]\n",
      " [0.76821192]\n",
      " [0.06410256]]\n",
      "re: 0.4544\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.24444444]\n",
      " [0.25827815]\n",
      " [0.        ]]\n",
      "re: 0.1257\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3381\n",
      "===================================================================\n",
      "K=1, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.0754717 ]\n",
      " [0.42222222]\n",
      " [1.        ]\n",
      " [0.37179487]]\n",
      "re: 0.4674\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.03333333]\n",
      " [0.37748344]\n",
      " [0.37179487]]\n",
      "re: 0.1980\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.        ]\n",
      " [0.63333333]\n",
      " [0.04635762]\n",
      " [0.1025641 ]]\n",
      "re: 0.4456\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.38679245]\n",
      " [0.27777778]\n",
      " [0.22516556]\n",
      " [0.42307692]]\n",
      "re: 0.3282\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.22222222]\n",
      " [0.70860927]\n",
      " [0.12820513]]\n",
      "re: 0.2648\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3408\n",
      "===================================================================\n",
      "K=1, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.03773585]\n",
      " [0.44444444]\n",
      " [0.45695364]\n",
      " [0.6025641 ]]\n",
      "re: 0.3854\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.20754717]\n",
      " [0.63333333]\n",
      " [0.60264901]\n",
      " [0.35897436]]\n",
      "re: 0.4506\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.        ]\n",
      " [0.57777778]\n",
      " [0.24503311]\n",
      " [0.29487179]]\n",
      "re: 0.5294\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.35849057]\n",
      " [0.1       ]\n",
      " [0.1986755 ]\n",
      " [0.11538462]]\n",
      "re: 0.1931\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.26666667]\n",
      " [0.0397351 ]\n",
      " [0.17948718]]\n",
      "re: 0.1238\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3365\n",
      "===================================================================\n",
      "K=1, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.74528302]\n",
      " [0.15555556]\n",
      " [0.71523179]\n",
      " [0.29487179]]\n",
      "re: 0.4777\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.08490566]\n",
      " [0.48888889]\n",
      " [0.1192053 ]\n",
      " [0.08974359]]\n",
      "re: 0.1957\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.01886792]\n",
      " [0.2       ]\n",
      " [0.86754967]\n",
      " [0.03846154]]\n",
      "re: 0.2812\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.1509434 ]\n",
      " [0.06666667]\n",
      " [0.26490066]\n",
      " [0.62820513]]\n",
      "re: 0.2777\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.02830189]\n",
      " [0.61111111]\n",
      " [0.35099338]\n",
      " [0.07692308]]\n",
      "re: 0.2668\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2998\n",
      "===================================================================\n",
      "K=1, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.24528302]\n",
      " [0.42222222]\n",
      " [0.20529801]\n",
      " [0.29487179]]\n",
      "re: 0.2919\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.0754717 ]\n",
      " [0.33333333]\n",
      " [0.1192053 ]\n",
      " [0.19230769]]\n",
      "re: 0.1801\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.02830189]\n",
      " [0.27777778]\n",
      " [0.1589404 ]\n",
      " [0.35897436]]\n",
      "re: 0.2060\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [0.01111111]\n",
      " [0.66887417]\n",
      " [0.37179487]]\n",
      "re: 0.2653\n",
      "------------------------------------------------------------------\n",
      "seed:4\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.04716981]\n",
      " [0.1       ]\n",
      " [0.0397351 ]\n",
      " [0.11538462]]\n",
      "re: 0.0756\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2038\n",
      "===================================================================\n",
      "K=1, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.10377358]\n",
      " [0.27777778]\n",
      " [0.15231788]\n",
      " [0.24358974]]\n",
      "re: 0.1944\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.06603774]\n",
      " [0.58888889]\n",
      " [0.0794702 ]\n",
      " [0.38461538]]\n",
      "re: 0.2798\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.05660377]\n",
      " [0.53333333]\n",
      " [0.24503311]\n",
      " [0.43589744]]\n",
      "re: 0.3177\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.10377358]\n",
      " [0.26666667]\n",
      " [0.41059603]\n",
      " [0.19230769]]\n",
      "re: 0.2433\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.37735849]\n",
      " [0.45555556]\n",
      " [0.17218543]\n",
      " [0.23076923]]\n",
      "re: 0.3090\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2688\n",
      "===================================================================\n",
      "K=1, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.03773585]\n",
      " [0.42222222]\n",
      " [0.47682119]\n",
      " [0.14102564]]\n",
      "re: 0.2695\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.16037736]\n",
      " [0.16666667]\n",
      " [0.12582781]\n",
      " [0.07692308]]\n",
      "re: 0.1324\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.01886792]\n",
      " [0.51111111]\n",
      " [0.21192053]\n",
      " [0.15384615]]\n",
      "re: 0.2239\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.31132075]\n",
      " [0.46666667]\n",
      " [0.09271523]\n",
      " [0.42307692]]\n",
      "re: 0.3234\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.17924528]\n",
      " [0.04444444]\n",
      " [1.        ]\n",
      " [0.16666667]]\n",
      "re: 0.3476\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2594\n",
      "===================================================================\n",
      "K=1, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.00943396]\n",
      " [0.23333333]\n",
      " [0.46357616]\n",
      " [0.32051282]]\n",
      "re: 0.2567\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.39622642]\n",
      " [0.02222222]\n",
      " [0.0794702 ]\n",
      " [0.01282051]]\n",
      "re: 0.1277\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.27358491]\n",
      " [0.6       ]\n",
      " [0.43046358]\n",
      " [0.26923077]]\n",
      "re: 0.3933\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.        ]\n",
      " [0.48888889]\n",
      " [0.92715232]\n",
      " [0.17948718]]\n",
      "re: 0.3989\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.50943396]\n",
      " [0.23333333]\n",
      " [0.20529801]\n",
      " [0.16666667]]\n",
      "re: 0.2787\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2911\n",
      "===================================================================\n",
      "K=1, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.00943396]\n",
      " [0.17777778]\n",
      " [1.        ]\n",
      " [0.75641026]]\n",
      "re: 0.4859\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.61320755]\n",
      " [0.16666667]\n",
      " [1.        ]\n",
      " [0.46153846]]\n",
      "re: 0.5604\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.16037736]\n",
      " [0.38888889]\n",
      " [0.25827815]\n",
      " [0.57692308]]\n",
      "re: 0.3461\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.05660377]\n",
      " [0.38888889]\n",
      " [0.39735099]\n",
      " [0.29487179]]\n",
      "re: 0.2844\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.50943396]\n",
      " [0.2       ]\n",
      " [0.23178808]\n",
      " [0.5       ]]\n",
      "re: 0.3603\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.4074\n",
      "===================================================================\n",
      "K=1, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.        ]\n",
      " [0.57777778]\n",
      " [0.37086093]\n",
      " [0.28205128]]\n",
      "re: 0.3077\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.03773585]\n",
      " [0.11111111]\n",
      " [0.38410596]\n",
      " [0.05128205]]\n",
      "re: 0.1461\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.00943396]\n",
      " [0.14444444]\n",
      " [0.04635762]\n",
      " [0.29487179]]\n",
      "re: 0.1238\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.        ]\n",
      " [0.61111111]\n",
      " [0.15231788]\n",
      " [1.        ]]\n",
      "re: 0.4409\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.5       ]\n",
      " [0.08888889]\n",
      " [0.60264901]\n",
      " [0.07692308]]\n",
      "re: 0.3171\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2671\n",
      "===================================================================\n",
      "K=1, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.01886792]\n",
      " [0.06666667]\n",
      " [0.21854305]\n",
      " [0.55128205]]\n",
      "re: 0.2138\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.66037736]\n",
      " [0.48888889]\n",
      " [0.77483444]\n",
      " [0.1025641 ]]\n",
      "re: 0.5067\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.01886792]\n",
      " [0.02222222]\n",
      " [0.31788079]\n",
      " [0.01282051]]\n",
      "re: 0.0929\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.32075472]\n",
      " [0.65555556]\n",
      " [0.06622517]\n",
      " [0.08974359]]\n",
      "re: 0.2831\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.37735849]\n",
      " [0.22222222]\n",
      " [0.44370861]\n",
      " [0.21794872]]\n",
      "re: 0.3153\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2824\n",
      "===================================================================\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "90cae2c28d9c487f8e4d8a101f5a6fb6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "110c14688fa744a3a797a5e909954c11",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "K=4, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.21698113]\n",
      " [0.46666667]\n",
      " [0.21192053]\n",
      " [0.12820513]]\n",
      "re: 0.2559\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.03773585]\n",
      " [0.63333333]\n",
      " [0.18543046]\n",
      " [0.30769231]]\n",
      "re: 0.2910\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [1.        ]\n",
      " [0.27152318]\n",
      " [0.33333333]]\n",
      "re: 0.4012\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.08490566]\n",
      " [0.56666667]\n",
      " [0.22516556]\n",
      " [0.24358974]]\n",
      "re: 0.2801\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.04716981]\n",
      " [0.64444444]\n",
      " [0.21854305]\n",
      " [0.34615385]]\n",
      "re: 0.3141\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3085\n",
      "===================================================================\n",
      "K=4, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.21698113]\n",
      " [0.46666667]\n",
      " [0.21192053]\n",
      " [0.12820513]]\n",
      "re: 0.2559\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.03773585]\n",
      " [0.63333333]\n",
      " [0.18543046]\n",
      " [0.30769231]]\n",
      "re: 0.2910\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.63333333]\n",
      " [0.15231788]\n",
      " [0.28205128]]\n",
      "re: 0.2669\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.08490566]\n",
      " [0.56666667]\n",
      " [0.22516556]\n",
      " [0.24358974]]\n",
      "re: 0.2801\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.04716981]\n",
      " [0.64444444]\n",
      " [0.21192053]\n",
      " [0.34615385]]\n",
      "re: 0.3124\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2813\n",
      "===================================================================\n",
      "K=4, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.21698113]\n",
      " [0.36666667]\n",
      " [0.26490066]\n",
      " [0.16666667]]\n",
      "re: 0.2538\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.04716981]\n",
      " [0.63333333]\n",
      " [0.29139073]\n",
      " [0.38461538]]\n",
      "re: 0.3391\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.64444444]\n",
      " [0.10596026]\n",
      " [0.21794872]]\n",
      "re: 0.2421\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.09433962]\n",
      " [0.56666667]\n",
      " [0.03311258]\n",
      " [0.30769231]]\n",
      "re: 0.2505\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.06603774]\n",
      " [0.64444444]\n",
      " [0.13245033]\n",
      " [0.34615385]]\n",
      "re: 0.2973\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2765\n",
      "===================================================================\n",
      "K=4, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.21698113]\n",
      " [0.61111111]\n",
      " [0.18543046]\n",
      " [1.        ]]\n",
      "re: 0.5034\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.54444444]\n",
      " [0.27152318]\n",
      " [0.20512821]]\n",
      "re: 0.2576\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.06603774]\n",
      " [0.64444444]\n",
      " [0.05960265]\n",
      " [0.43589744]]\n",
      "re: 0.3015\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.        ]\n",
      " [0.57777778]\n",
      " [0.0397351 ]\n",
      " [1.        ]]\n",
      "re: 0.4044\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.02830189]\n",
      " [0.5       ]\n",
      " [0.04635762]\n",
      " [0.38461538]]\n",
      "re: 0.2398\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3413\n",
      "===================================================================\n",
      "K=4, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.00943396]\n",
      " [0.42222222]\n",
      " [0.15231788]\n",
      " [0.35897436]]\n",
      "re: 0.2357\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.01111111]\n",
      " [0.08609272]\n",
      " [0.43589744]]\n",
      "re: 0.1356\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.5       ]\n",
      " [0.39735099]\n",
      " [0.16666667]]\n",
      "re: 0.2660\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [0.48888889]\n",
      " [0.08609272]\n",
      " [0.14102564]]\n",
      "re: 0.1814\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.51111111]\n",
      " [0.27152318]\n",
      " [0.32051282]]\n",
      "re: 0.2758\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2189\n",
      "===================================================================\n",
      "K=4, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.03773585]\n",
      " [0.42222222]\n",
      " [0.16556291]\n",
      " [0.35897436]]\n",
      "re: 0.2461\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.01111111]\n",
      " [0.08609272]\n",
      " [0.43589744]]\n",
      "re: 0.1333\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.5       ]\n",
      " [0.39072848]\n",
      " [0.15384615]]\n",
      "re: 0.2611\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [0.48888889]\n",
      " [0.08609272]\n",
      " [0.15384615]]\n",
      "re: 0.1846\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.01886792]\n",
      " [0.45555556]\n",
      " [0.15231788]\n",
      " [0.01282051]]\n",
      "re: 0.1599\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.1970\n",
      "===================================================================\n",
      "K=4, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.        ]\n",
      " [0.42222222]\n",
      " [0.01986755]\n",
      " [0.20512821]]\n",
      "re: 0.1618\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.01111111]\n",
      " [0.13245033]\n",
      " [0.38461538]]\n",
      "re: 0.1344\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.64444444]\n",
      " [0.21192053]\n",
      " [0.28205128]]\n",
      "re: 0.2846\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [0.48888889]\n",
      " [0.0794702 ]\n",
      " [0.12820513]]\n",
      "re: 0.1765\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.02830189]\n",
      " [0.52222222]\n",
      " [0.0397351 ]\n",
      " [0.33333333]]\n",
      "re: 0.2309\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.1976\n",
      "===================================================================\n",
      "K=4, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.1509434 ]\n",
      " [0.64444444]\n",
      " [0.08609272]\n",
      " [0.37179487]]\n",
      "re: 0.3133\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.58888889]\n",
      " [0.07284768]\n",
      " [0.26923077]]\n",
      "re: 0.2327\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.54444444]\n",
      " [0.1192053 ]\n",
      " [0.19230769]]\n",
      "re: 0.2140\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.01886792]\n",
      " [0.48888889]\n",
      " [0.37086093]\n",
      " [0.43589744]]\n",
      "re: 0.3286\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.51111111]\n",
      " [0.90066225]\n",
      " [0.28205128]]\n",
      "re: 0.4235\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3024\n",
      "===================================================================\n",
      "K=4, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.        ]\n",
      " [0.36666667]\n",
      " [0.1986755 ]\n",
      " [0.57692308]]\n",
      "re: 0.2856\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.11320755]\n",
      " [0.01111111]\n",
      " [0.71523179]\n",
      " [0.01282051]]\n",
      "re: 0.2131\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.05660377]\n",
      " [1.        ]\n",
      " [0.11258278]\n",
      " [0.17948718]]\n",
      "re: 0.3372\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.09433962]\n",
      " [0.55555556]\n",
      " [0.51655629]\n",
      " [0.05128205]]\n",
      "re: 0.3044\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.31132075]\n",
      " [0.41111111]\n",
      " [0.23178808]\n",
      " [0.38461538]]\n",
      "re: 0.3347\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2950\n",
      "===================================================================\n",
      "K=4, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.        ]\n",
      " [0.        ]\n",
      " [0.56953642]\n",
      " [0.37179487]]\n",
      "re: 0.2353\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.11320755]\n",
      " [0.01111111]\n",
      " [0.25165563]\n",
      " [0.06410256]]\n",
      "re: 0.1100\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.05660377]\n",
      " [1.        ]\n",
      " [0.11258278]\n",
      " [0.17948718]]\n",
      "re: 0.3372\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.09433962]\n",
      " [0.55555556]\n",
      " [0.11258278]\n",
      " [0.02564103]]\n",
      "re: 0.1970\n",
      "------------------------------------------------------------------\n",
      "seed:4\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.31132075]\n",
      " [0.41111111]\n",
      " [0.10596026]\n",
      " [0.11538462]]\n",
      "re: 0.2359\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2231\n",
      "===================================================================\n",
      "K=4, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.        ]\n",
      " [0.54444444]\n",
      " [0.10596026]\n",
      " [0.37179487]]\n",
      "re: 0.2555\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.11320755]\n",
      " [0.13333333]\n",
      " [0.01986755]\n",
      " [0.21794872]]\n",
      "re: 0.1211\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.05660377]\n",
      " [0.06666667]\n",
      " [0.09271523]\n",
      " [0.14102564]]\n",
      "re: 0.0893\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.09433962]\n",
      " [0.55555556]\n",
      " [0.16556291]\n",
      " [0.17948718]]\n",
      "re: 0.2487\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.06603774]\n",
      " [0.62222222]\n",
      " [0.1986755 ]\n",
      " [0.12820513]]\n",
      "re: 0.2538\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.1937\n",
      "===================================================================\n",
      "K=4, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.21698113]\n",
      " [0.21111111]\n",
      " [0.26490066]\n",
      " [0.41025641]]\n",
      "re: 0.2758\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.22641509]\n",
      " [1.        ]\n",
      " [0.34437086]\n",
      " [0.33333333]]\n",
      "re: 0.4760\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.05660377]\n",
      " [1.        ]\n",
      " [0.34437086]\n",
      " [0.11538462]]\n",
      "re: 0.3791\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.        ]\n",
      " [1.        ]\n",
      " [0.25165563]\n",
      " [0.19230769]]\n",
      "re: 0.3610\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.54444444]\n",
      " [0.10596026]\n",
      " [0.34615385]]\n",
      "re: 0.2515\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3487\n",
      "===================================================================\n",
      "K=4, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.        ]\n",
      " [0.51111111]\n",
      " [0.12582781]\n",
      " [0.11538462]]\n",
      "re: 0.1881\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.05555556]\n",
      " [0.2781457 ]\n",
      " [0.21794872]]\n",
      "re: 0.1403\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.00943396]\n",
      " [0.48888889]\n",
      " [0.21854305]\n",
      " [0.12820513]]\n",
      "re: 0.2113\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.03773585]\n",
      " [0.53333333]\n",
      " [0.37748344]\n",
      " [0.24358974]]\n",
      "re: 0.2980\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.46666667]\n",
      " [0.07284768]\n",
      " [0.07692308]]\n",
      "re: 0.1565\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.1988\n",
      "===================================================================\n",
      "K=4, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.1509434 ]\n",
      " [0.61111111]\n",
      " [0.56291391]\n",
      " [0.3974359 ]]\n",
      "re: 0.4306\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.05555556]\n",
      " [0.2781457 ]\n",
      " [0.21794872]]\n",
      "re: 0.1403\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.00943396]\n",
      " [0.48888889]\n",
      " [0.21854305]\n",
      " [0.12820513]]\n",
      "re: 0.2113\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.02830189]\n",
      " [0.53333333]\n",
      " [0.09271523]\n",
      " [0.17948718]]\n",
      "re: 0.2085\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.44444444]\n",
      " [0.05960265]\n",
      " [0.        ]]\n",
      "re: 0.1284\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2238\n",
      "===================================================================\n",
      "K=4, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.11320755]\n",
      " [0.54444444]\n",
      " [0.25165563]\n",
      " [0.20512821]]\n",
      "re: 0.2786\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.01886792]\n",
      " [0.51111111]\n",
      " [0.35099338]\n",
      " [0.25641026]]\n",
      "re: 0.2843\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.00943396]\n",
      " [0.48888889]\n",
      " [0.21854305]\n",
      " [0.12820513]]\n",
      "re: 0.2113\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.        ]\n",
      " [0.53333333]\n",
      " [0.07284768]\n",
      " [0.03846154]]\n",
      "re: 0.1612\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.46666667]\n",
      " [0.01986755]\n",
      " [0.25641026]]\n",
      "re: 0.1857\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2242\n",
      "===================================================================\n",
      "K=4, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.21698113]\n",
      " [0.57777778]\n",
      " [0.35761589]\n",
      " [0.30769231]]\n",
      "re: 0.3650\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.01886792]\n",
      " [0.58888889]\n",
      " [0.20529801]\n",
      " [0.14102564]]\n",
      "re: 0.2385\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.00943396]\n",
      " [0.16666667]\n",
      " [0.14569536]\n",
      " [0.20512821]]\n",
      "re: 0.1317\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.0754717 ]\n",
      " [0.63333333]\n",
      " [0.44370861]\n",
      " [0.17948718]]\n",
      "re: 0.3330\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.01886792]\n",
      " [0.53333333]\n",
      " [0.43708609]\n",
      " [0.11538462]]\n",
      "re: 0.2762\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2689\n",
      "===================================================================\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2fcae9a46d0943728b593611b333e46f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "K=4, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.28301887]\n",
      " [0.6       ]\n",
      " [1.        ]\n",
      " [0.43589744]]\n",
      "re: 0.5797\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.09433962]\n",
      " [0.64444444]\n",
      " [0.21192053]\n",
      " [0.55128205]]\n",
      "re: 0.3755\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.19811321]\n",
      " [0.58888889]\n",
      " [0.27152318]\n",
      " [0.01282051]]\n",
      "re: 0.2678\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [0.45555556]\n",
      " [0.13907285]\n",
      " [0.23076923]]\n",
      "re: 0.4563\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.6       ]\n",
      " [0.39735099]\n",
      " [0.12820513]]\n",
      "re: 0.2837\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3926\n",
      "===================================================================\n",
      "K=4, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.28301887]\n",
      " [0.6       ]\n",
      " [1.        ]\n",
      " [0.43589744]]\n",
      "re: 0.5797\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.64444444]\n",
      " [0.21192053]\n",
      " [1.        ]]\n",
      "re: 0.4641\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.17924528]\n",
      " [0.58888889]\n",
      " [0.27152318]\n",
      " [0.16666667]]\n",
      "re: 0.3016\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [0.5       ]\n",
      " [0.10596026]\n",
      " [0.24358974]]\n",
      "re: 0.4624\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.62222222]\n",
      " [0.20529801]\n",
      " [0.3974359 ]]\n",
      "re: 0.3062\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.4228\n",
      "===================================================================\n",
      "K=4, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.28301887]\n",
      " [0.58888889]\n",
      " [1.        ]\n",
      " [0.42307692]]\n",
      "re: 0.5737\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.37777778]\n",
      " [0.2781457 ]\n",
      " [0.52564103]]\n",
      "re: 0.2954\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.14150943]\n",
      " [0.58888889]\n",
      " [0.27152318]\n",
      " [0.35897436]]\n",
      "re: 0.3402\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [0.52222222]\n",
      " [0.21854305]\n",
      " [0.34615385]]\n",
      "re: 0.5217\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.03773585]\n",
      " [0.57777778]\n",
      " [0.39735099]\n",
      " [0.12820513]]\n",
      "re: 0.2853\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.4033\n",
      "===================================================================\n",
      "K=4, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.29245283]\n",
      " [0.54444444]\n",
      " [1.        ]\n",
      " [0.30769231]]\n",
      "re: 0.5361\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.5       ]\n",
      " [0.1986755 ]\n",
      " [0.32051282]]\n",
      "re: 0.2548\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.24528302]\n",
      " [0.64444444]\n",
      " [0.23178808]\n",
      " [0.07692308]]\n",
      "re: 0.2996\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [0.53333333]\n",
      " [0.41721854]\n",
      " [0.07692308]]\n",
      "re: 0.5069\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.01886792]\n",
      " [0.44444444]\n",
      " [0.05298013]\n",
      " [0.01282051]]\n",
      "re: 0.1323\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3459\n",
      "===================================================================\n",
      "K=4, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.20754717]\n",
      " [0.73333333]\n",
      " [0.17880795]\n",
      " [0.14102564]]\n",
      "re: 0.3152\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.33962264]\n",
      " [0.68888889]\n",
      " [0.0794702 ]\n",
      " [0.11538462]]\n",
      "re: 0.3058\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.00943396]\n",
      " [0.52222222]\n",
      " [0.47019868]\n",
      " [0.16666667]]\n",
      "re: 0.2921\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.        ]\n",
      " [0.36666667]\n",
      " [0.17880795]\n",
      " [0.23076923]]\n",
      "re: 0.1941\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.0754717 ]\n",
      " [0.07777778]\n",
      " [0.19205298]\n",
      " [1.        ]]\n",
      "re: 0.3363\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2887\n",
      "===================================================================\n",
      "K=4, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.1509434 ]\n",
      " [0.13333333]\n",
      " [0.30463576]\n",
      " [0.3974359 ]]\n",
      "re: 0.2466\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.35849057]\n",
      " [0.48888889]\n",
      " [0.12582781]\n",
      " [0.21794872]]\n",
      "re: 0.2978\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.00943396]\n",
      " [0.13333333]\n",
      " [0.13245033]\n",
      " [0.24358974]]\n",
      "re: 0.1297\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.        ]\n",
      " [0.36666667]\n",
      " [0.21854305]\n",
      " [0.23076923]]\n",
      "re: 0.2040\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.0754717 ]\n",
      " [0.61111111]\n",
      " [0.04635762]\n",
      " [0.70512821]]\n",
      "re: 0.3595\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2475\n",
      "===================================================================\n",
      "K=4, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.60377358]\n",
      " [0.63333333]\n",
      " [0.16556291]\n",
      " [0.08974359]]\n",
      "re: 0.3731\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.46666667]\n",
      " [0.07284768]\n",
      " [0.15384615]]\n",
      "re: 0.1757\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.27777778]\n",
      " [0.39072848]\n",
      " [0.16666667]]\n",
      "re: 0.2088\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.        ]\n",
      " [0.36666667]\n",
      " [0.07284768]\n",
      " [0.01282051]]\n",
      "re: 0.1131\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.0754717 ]\n",
      " [0.61111111]\n",
      " [0.05298013]\n",
      " [0.29487179]]\n",
      "re: 0.2586\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2259\n",
      "===================================================================\n",
      "K=4, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.33962264]\n",
      " [0.47777778]\n",
      " [0.55629139]\n",
      " [0.28205128]]\n",
      "re: 0.4139\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.44339623]\n",
      " [0.07777778]\n",
      " [0.19205298]\n",
      " [0.32051282]]\n",
      "re: 0.2584\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.13333333]\n",
      " [0.20529801]\n",
      " [0.05128205]]\n",
      "re: 0.0975\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.        ]\n",
      " [0.63333333]\n",
      " [0.22516556]\n",
      " [0.11538462]]\n",
      "re: 0.2435\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.0754717 ]\n",
      " [0.45555556]\n",
      " [0.26490066]\n",
      " [0.08974359]]\n",
      "re: 0.2214\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2469\n",
      "===================================================================\n",
      "K=4, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.00943396]\n",
      " [0.07777778]\n",
      " [0.0397351 ]\n",
      " [0.14102564]]\n",
      "re: 0.0670\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.6       ]\n",
      " [1.        ]\n",
      " [0.58974359]]\n",
      "re: 0.5498\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.61320755]\n",
      " [0.61111111]\n",
      " [0.19205298]\n",
      " [0.19230769]]\n",
      "re: 0.4022\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.19811321]\n",
      " [0.44444444]\n",
      " [0.69536424]\n",
      " [0.44871795]]\n",
      "re: 0.4467\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.0754717 ]\n",
      " [0.42222222]\n",
      " [0.41059603]\n",
      " [0.16666667]]\n",
      "re: 0.2687\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3469\n",
      "===================================================================\n",
      "K=4, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.02830189]\n",
      " [0.06666667]\n",
      " [0.11258278]\n",
      " [0.14102564]]\n",
      "re: 0.0871\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.06603774]\n",
      " [0.6       ]\n",
      " [1.        ]\n",
      " [0.62820513]]\n",
      "re: 0.5736\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.01886792]\n",
      " [0.        ]\n",
      " [0.20529801]\n",
      " [1.        ]]\n",
      "re: 0.3060\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.21698113]\n",
      " [0.44444444]\n",
      " [0.66225166]\n",
      " [0.44871795]]\n",
      "re: 0.4431\n",
      "------------------------------------------------------------------\n",
      "seed:4\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.0754717 ]\n",
      " [0.44444444]\n",
      " [0.41059603]\n",
      " [0.16666667]]\n",
      "re: 0.2743\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3368\n",
      "===================================================================\n",
      "K=4, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.0754717 ]\n",
      " [0.06666667]\n",
      " [0.78145695]\n",
      " [0.24358974]]\n",
      "re: 0.2918\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.45555556]\n",
      " [0.02649007]\n",
      " [0.25641026]]\n",
      "re: 0.1846\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.12264151]\n",
      " [0.62222222]\n",
      " [0.64900662]\n",
      " [0.05128205]]\n",
      "re: 0.3613\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [0.08888889]\n",
      " [0.08609272]\n",
      " [0.03846154]]\n",
      "re: 0.0557\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.0754717 ]\n",
      " [0.56666667]\n",
      " [0.35761589]\n",
      " [0.16666667]]\n",
      "re: 0.2916\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2370\n",
      "===================================================================\n",
      "K=4, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.01886792]\n",
      " [0.02222222]\n",
      " [0.07284768]\n",
      " [0.35897436]]\n",
      "re: 0.1182\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.02830189]\n",
      " [0.61111111]\n",
      " [0.62913907]\n",
      " [0.29487179]]\n",
      "re: 0.3909\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.36792453]\n",
      " [0.52222222]\n",
      " [0.43046358]\n",
      " [0.        ]]\n",
      "re: 0.3302\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [0.53333333]\n",
      " [0.19205298]\n",
      " [0.3974359 ]]\n",
      "re: 0.2831\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.14150943]\n",
      " [0.21111111]\n",
      " [0.17880795]\n",
      " [0.02564103]]\n",
      "re: 0.1393\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2523\n",
      "===================================================================\n",
      "K=4, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.30188679]\n",
      " [0.38888889]\n",
      " [0.19205298]\n",
      " [0.05128205]]\n",
      "re: 0.2335\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.38679245]\n",
      " [0.06666667]\n",
      " [0.27152318]\n",
      " [0.21794872]]\n",
      "re: 0.2357\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.62264151]\n",
      " [0.38888889]\n",
      " [0.7615894 ]\n",
      " [0.20512821]]\n",
      "re: 0.4946\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.0754717 ]\n",
      " [0.57777778]\n",
      " [0.68874172]\n",
      " [0.34615385]]\n",
      "re: 0.4220\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.1509434 ]\n",
      " [0.43333333]\n",
      " [0.22516556]\n",
      " [0.17948718]]\n",
      "re: 0.2472\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3266\n",
      "===================================================================\n",
      "K=4, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.01886792]\n",
      " [0.17777778]\n",
      " [0.74172185]\n",
      " [0.58974359]]\n",
      "re: 0.3820\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.38679245]\n",
      " [0.37777778]\n",
      " [0.09933775]\n",
      " [0.41025641]]\n",
      "re: 0.3185\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.59433962]\n",
      " [0.4       ]\n",
      " [0.47019868]\n",
      " [0.30769231]]\n",
      "re: 0.4431\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.52830189]\n",
      " [1.        ]\n",
      " [0.06622517]\n",
      " [0.34615385]]\n",
      "re: 0.4852\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.02830189]\n",
      " [0.47777778]\n",
      " [0.25827815]\n",
      " [0.21794872]]\n",
      "re: 0.2456\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3749\n",
      "===================================================================\n",
      "K=4, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.26415094]\n",
      " [0.62222222]\n",
      " [0.08609272]\n",
      " [0.08974359]]\n",
      "re: 0.2656\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.37735849]\n",
      " [0.47777778]\n",
      " [0.16556291]\n",
      " [1.        ]]\n",
      "re: 0.5052\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.3490566 ]\n",
      " [0.27777778]\n",
      " [0.2384106 ]\n",
      " [0.12820513]]\n",
      "re: 0.2484\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.0754717 ]\n",
      " [0.14444444]\n",
      " [0.03311258]\n",
      " [0.07692308]]\n",
      "re: 0.0825\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.38679245]\n",
      " [0.02222222]\n",
      " [1.        ]\n",
      " [0.35897436]]\n",
      "re: 0.4420\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3087\n",
      "===================================================================\n",
      "K=4, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.1509434 ]\n",
      " [0.22222222]\n",
      " [0.07284768]\n",
      " [0.08974359]]\n",
      "re: 0.1339\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.01886792]\n",
      " [0.52222222]\n",
      " [0.29801325]\n",
      " [0.06410256]]\n",
      "re: 0.2258\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.6509434 ]\n",
      " [0.5       ]\n",
      " [0.16556291]\n",
      " [0.02564103]]\n",
      "re: 0.3355\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.80188679]\n",
      " [0.5       ]\n",
      " [0.09933775]\n",
      " [0.12820513]]\n",
      "re: 0.3824\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.19811321]\n",
      " [0.56666667]\n",
      " [0.21192053]\n",
      " [0.29487179]]\n",
      "re: 0.3179\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2791\n",
      "===================================================================\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "746b7ec0335e4b6d90500fe2f9130018",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d3443c93ac774cad9ef736fd9cd5d272",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "K=8, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.01886792]\n",
      " [0.51111111]\n",
      " [0.04635762]\n",
      " [0.25641026]]\n",
      "re: 0.2082\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.03773585]\n",
      " [0.53333333]\n",
      " [0.19205298]\n",
      " [0.20512821]]\n",
      "re: 0.2421\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.02830189]\n",
      " [0.5       ]\n",
      " [0.27152318]\n",
      " [0.3974359 ]]\n",
      "re: 0.2993\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.        ]\n",
      " [0.53333333]\n",
      " [0.09271523]\n",
      " [0.26923077]]\n",
      "re: 0.2238\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.58888889]\n",
      " [0.20529801]\n",
      " [0.33333333]]\n",
      "re: 0.2842\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2515\n",
      "===================================================================\n",
      "K=8, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.01886792]\n",
      " [0.51111111]\n",
      " [0.25827815]\n",
      " [0.25641026]]\n",
      "re: 0.2612\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.03773585]\n",
      " [0.52222222]\n",
      " [0.06622517]\n",
      " [0.34615385]]\n",
      "re: 0.2431\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.02830189]\n",
      " [0.46666667]\n",
      " [0.10596026]\n",
      " [0.25641026]]\n",
      "re: 0.2143\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.        ]\n",
      " [0.53333333]\n",
      " [0.1986755 ]\n",
      " [1.        ]]\n",
      "re: 0.4330\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.58888889]\n",
      " [0.24503311]\n",
      " [0.29487179]]\n",
      "re: 0.2846\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2872\n",
      "===================================================================\n",
      "K=8, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.00943396]\n",
      " [0.51111111]\n",
      " [0.09933775]\n",
      " [0.19230769]]\n",
      "re: 0.2030\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.03773585]\n",
      " [0.58888889]\n",
      " [0.1192053 ]\n",
      " [0.30769231]]\n",
      "re: 0.2634\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.02830189]\n",
      " [0.44444444]\n",
      " [0.2384106 ]\n",
      " [0.28205128]]\n",
      "re: 0.2483\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.        ]\n",
      " [0.53333333]\n",
      " [0.17218543]\n",
      " [0.17948718]]\n",
      "re: 0.2213\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.58888889]\n",
      " [0.12582781]\n",
      " [1.        ]]\n",
      "re: 0.4287\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2729\n",
      "===================================================================\n",
      "K=8, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.09433962]\n",
      " [0.5       ]\n",
      " [0.27152318]\n",
      " [0.20512821]]\n",
      "re: 0.2677\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.22641509]\n",
      " [0.57777778]\n",
      " [0.25827815]\n",
      " [1.        ]]\n",
      "re: 0.5156\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.58888889]\n",
      " [0.21192053]\n",
      " [0.19230769]]\n",
      "re: 0.2483\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.10377358]\n",
      " [0.63333333]\n",
      " [0.13907285]\n",
      " [0.30769231]]\n",
      "re: 0.2960\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.04716981]\n",
      " [1.        ]\n",
      " [0.19205298]\n",
      " [1.        ]]\n",
      "re: 0.5598\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3775\n",
      "===================================================================\n",
      "K=8, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.04716981]\n",
      " [0.52222222]\n",
      " [0.65562914]\n",
      " [0.42307692]]\n",
      "re: 0.4120\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.63333333]\n",
      " [0.31125828]\n",
      " [0.3974359 ]]\n",
      "re: 0.3379\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.63207547]\n",
      " [0.64444444]\n",
      " [0.1589404 ]\n",
      " [0.32051282]]\n",
      "re: 0.4390\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.20754717]\n",
      " [1.        ]\n",
      " [0.2384106 ]\n",
      " [0.29487179]]\n",
      "re: 0.4352\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.19811321]\n",
      " [0.62222222]\n",
      " [0.21192053]\n",
      " [0.30769231]]\n",
      "re: 0.3350\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3918\n",
      "===================================================================\n",
      "K=8, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.04716981]\n",
      " [0.53333333]\n",
      " [0.13245033]\n",
      " [0.41025641]]\n",
      "re: 0.2808\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.63333333]\n",
      " [0.31125828]\n",
      " [0.3974359 ]]\n",
      "re: 0.3379\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.59433962]\n",
      " [0.63333333]\n",
      " [0.13245033]\n",
      " [0.15384615]]\n",
      "re: 0.3785\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.20754717]\n",
      " [1.        ]\n",
      " [0.2384106 ]\n",
      " [0.29487179]]\n",
      "re: 0.4352\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.19811321]\n",
      " [0.62222222]\n",
      " [0.20529801]\n",
      " [0.30769231]]\n",
      "re: 0.3333\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3531\n",
      "===================================================================\n",
      "K=8, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.04716981]\n",
      " [1.        ]\n",
      " [0.09271523]\n",
      " [0.17948718]]\n",
      "re: 0.3298\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.62222222]\n",
      " [0.35761589]\n",
      " [0.32051282]]\n",
      "re: 0.3274\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.63207547]\n",
      " [0.05555556]\n",
      " [0.35761589]\n",
      " [1.        ]]\n",
      "re: 0.5113\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.20754717]\n",
      " [1.        ]\n",
      " [0.2384106 ]\n",
      " [0.29487179]]\n",
      "re: 0.4352\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.63333333]\n",
      " [0.24503311]\n",
      " [0.20512821]]\n",
      "re: 0.2732\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3754\n",
      "===================================================================\n",
      "K=8, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.04716981]\n",
      " [1.        ]\n",
      " [0.2781457 ]\n",
      " [0.15384615]]\n",
      "re: 0.3698\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.64444444]\n",
      " [0.25165563]\n",
      " [0.25641026]]\n",
      "re: 0.2905\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.63207547]\n",
      " [0.64444444]\n",
      " [0.01986755]\n",
      " [0.16666667]]\n",
      "re: 0.3658\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.05660377]\n",
      " [0.45555556]\n",
      " [0.09933775]\n",
      " [0.26923077]]\n",
      "re: 0.2202\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.61111111]\n",
      " [0.13245033]\n",
      " [0.29487179]]\n",
      "re: 0.2596\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3012\n",
      "===================================================================\n",
      "K=8, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.16981132]\n",
      " [0.54444444]\n",
      " [0.27152318]\n",
      " [0.17948718]]\n",
      "re: 0.2913\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.05660377]\n",
      " [0.45555556]\n",
      " [0.17218543]\n",
      " [0.29487179]]\n",
      "re: 0.2448\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.06666667]\n",
      " [0.14569536]\n",
      " [0.        ]]\n",
      "re: 0.0531\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [0.64444444]\n",
      " [0.1192053 ]\n",
      " [0.25641026]]\n",
      "re: 0.2574\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.64444444]\n",
      " [0.12582781]\n",
      " [0.20512821]]\n",
      "re: 0.2439\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2181\n",
      "===================================================================\n",
      "K=8, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.16037736]\n",
      " [0.54444444]\n",
      " [0.26490066]\n",
      " [0.16666667]]\n",
      "re: 0.2841\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.03773585]\n",
      " [0.47777778]\n",
      " [0.0397351 ]\n",
      " [0.20512821]]\n",
      "re: 0.1901\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.06666667]\n",
      " [0.14569536]\n",
      " [0.14102564]]\n",
      "re: 0.0883\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [0.64444444]\n",
      " [0.1192053 ]\n",
      " [0.29487179]]\n",
      "re: 0.2670\n",
      "------------------------------------------------------------------\n",
      "seed:4\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.        ]\n",
      " [0.64444444]\n",
      " [0.13245033]\n",
      " [0.20512821]]\n",
      "re: 0.2455\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2150\n",
      "===================================================================\n",
      "K=8, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.1509434 ]\n",
      " [0.54444444]\n",
      " [0.2781457 ]\n",
      " [0.17948718]]\n",
      "re: 0.2883\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.0754717 ]\n",
      " [0.47777778]\n",
      " [0.13245033]\n",
      " [0.29487179]]\n",
      "re: 0.2451\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.02222222]\n",
      " [0.13907285]\n",
      " [0.15384615]]\n",
      "re: 0.0788\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [0.47777778]\n",
      " [0.12582781]\n",
      " [0.34615385]]\n",
      "re: 0.2398\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.64444444]\n",
      " [0.13245033]\n",
      " [0.20512821]]\n",
      "re: 0.2455\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2195\n",
      "===================================================================\n",
      "K=8, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.10377358]\n",
      " [0.32222222]\n",
      " [0.37086093]\n",
      " [0.12820513]]\n",
      "re: 0.2313\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.24528302]\n",
      " [0.53333333]\n",
      " [0.25827815]\n",
      " [0.23076923]]\n",
      "re: 0.3169\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.26666667]\n",
      " [0.28476821]\n",
      " [0.41025641]]\n",
      "re: 0.2404\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.        ]\n",
      " [0.61111111]\n",
      " [0.34437086]\n",
      " [0.17948718]]\n",
      "re: 0.2837\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.5       ]\n",
      " [0.21192053]\n",
      " [0.20512821]]\n",
      "re: 0.2293\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2603\n",
      "===================================================================\n",
      "K=8, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.00943396]\n",
      " [0.58888889]\n",
      " [0.1589404 ]\n",
      " [0.23076923]]\n",
      "re: 0.2470\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.01886792]\n",
      " [0.6       ]\n",
      " [0.10596026]\n",
      " [0.15384615]]\n",
      "re: 0.2197\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.59433962]\n",
      " [0.4       ]\n",
      " [0.18543046]\n",
      " [0.23076923]]\n",
      "re: 0.3526\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.30188679]\n",
      " [0.47777778]\n",
      " [0.17218543]\n",
      " [0.01282051]]\n",
      "re: 0.2412\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.08490566]\n",
      " [0.5       ]\n",
      " [0.0397351 ]\n",
      " [0.3974359 ]]\n",
      "re: 0.2555\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2632\n",
      "===================================================================\n",
      "K=8, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.00943396]\n",
      " [0.51111111]\n",
      " [0.07284768]\n",
      " [0.32051282]]\n",
      "re: 0.2285\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.01886792]\n",
      " [0.42222222]\n",
      " [0.37748344]\n",
      " [0.17948718]]\n",
      "re: 0.2495\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.59433962]\n",
      " [0.55555556]\n",
      " [0.19205298]\n",
      " [0.43589744]]\n",
      "re: 0.4445\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.30188679]\n",
      " [0.63333333]\n",
      " [0.05960265]\n",
      " [0.24358974]]\n",
      "re: 0.3096\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.08490566]\n",
      " [0.51111111]\n",
      " [0.0397351 ]\n",
      " [0.3974359 ]]\n",
      "re: 0.2583\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2981\n",
      "===================================================================\n",
      "K=8, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.00943396]\n",
      " [0.6       ]\n",
      " [0.26490066]\n",
      " [0.29487179]]\n",
      "re: 0.2923\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.01886792]\n",
      " [0.45555556]\n",
      " [0.16556291]\n",
      " [0.01282051]]\n",
      "re: 0.1632\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.59433962]\n",
      " [0.13333333]\n",
      " [0.09933775]\n",
      " [1.        ]]\n",
      "re: 0.4568\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.30188679]\n",
      " [0.62222222]\n",
      " [0.21192053]\n",
      " [0.15384615]]\n",
      "re: 0.3225\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.58888889]\n",
      " [1.        ]\n",
      " [0.14102564]]\n",
      "re: 0.4325\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3334\n",
      "===================================================================\n",
      "K=8, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.05660377]\n",
      " [0.03333333]\n",
      " [0.01986755]\n",
      " [1.        ]]\n",
      "re: 0.2775\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.03773585]\n",
      " [0.52222222]\n",
      " [0.35761589]\n",
      " [0.14102564]]\n",
      "re: 0.2646\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.01886792]\n",
      " [0.5       ]\n",
      " [0.26490066]\n",
      " [0.14102564]]\n",
      "re: 0.2312\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.05660377]\n",
      " [0.47777778]\n",
      " [0.52980132]\n",
      " [0.3974359 ]]\n",
      "re: 0.3654\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.47777778]\n",
      " [0.17218543]\n",
      " [0.11538462]]\n",
      "re: 0.1913\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2660\n",
      "===================================================================\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8dc93119d47b4df6b4779a5d10b15309",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "K=8, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.        ]\n",
      " [0.04444444]\n",
      " [0.89403974]\n",
      " [0.3974359 ]]\n",
      "re: 0.3340\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.14150943]\n",
      " [0.56666667]\n",
      " [0.07284768]\n",
      " [0.62820513]]\n",
      "re: 0.3523\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.09433962]\n",
      " [1.        ]\n",
      " [0.15231788]\n",
      " [0.17948718]]\n",
      "re: 0.3565\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.01886792]\n",
      " [0.34444444]\n",
      " [0.12582781]\n",
      " [0.20512821]]\n",
      "re: 0.1736\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.62222222]\n",
      " [0.2384106 ]\n",
      " [0.24358974]]\n",
      "re: 0.2761\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2985\n",
      "===================================================================\n",
      "K=8, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.        ]\n",
      " [0.23333333]\n",
      " [0.1192053 ]\n",
      " [0.43589744]]\n",
      "re: 0.1971\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.14150943]\n",
      " [0.57777778]\n",
      " [0.0794702 ]\n",
      " [0.3974359 ]]\n",
      "re: 0.2990\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.09433962]\n",
      " [1.        ]\n",
      " [0.15231788]\n",
      " [0.03846154]]\n",
      "re: 0.3213\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.02830189]\n",
      " [0.47777778]\n",
      " [0.16556291]\n",
      " [0.55128205]]\n",
      "re: 0.3057\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.62222222]\n",
      " [0.18543046]\n",
      " [0.34615385]]\n",
      "re: 0.2885\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2823\n",
      "===================================================================\n",
      "K=8, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.        ]\n",
      " [0.07777778]\n",
      " [0.07284768]\n",
      " [0.24358974]]\n",
      "re: 0.0986\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.14150943]\n",
      " [0.6       ]\n",
      " [0.03311258]\n",
      " [0.55128205]]\n",
      "re: 0.3315\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.08490566]\n",
      " [0.02222222]\n",
      " [0.18543046]\n",
      " [0.78205128]]\n",
      "re: 0.2687\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.04716981]\n",
      " [0.13333333]\n",
      " [0.12582781]\n",
      " [1.        ]]\n",
      "re: 0.3266\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.62222222]\n",
      " [0.25827815]\n",
      " [0.3974359 ]]\n",
      "re: 0.3195\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2689\n",
      "===================================================================\n",
      "K=8, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.10377358]\n",
      " [0.07777778]\n",
      " [0.13907285]\n",
      " [0.37179487]]\n",
      "re: 0.1731\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.1509434 ]\n",
      " [0.58888889]\n",
      " [0.12582781]\n",
      " [0.42307692]]\n",
      "re: 0.3222\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.28888889]\n",
      " [0.15231788]\n",
      " [0.02564103]]\n",
      "re: 0.1167\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.02830189]\n",
      " [0.37777778]\n",
      " [0.61589404]\n",
      " [0.43589744]]\n",
      "re: 0.3645\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.62222222]\n",
      " [0.66887417]\n",
      " [0.32051282]]\n",
      "re: 0.4053\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2763\n",
      "===================================================================\n",
      "K=8, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.05660377]\n",
      " [0.        ]\n",
      " [0.21854305]\n",
      " [0.25641026]]\n",
      "re: 0.1329\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.08490566]\n",
      " [0.77777778]\n",
      " [0.18543046]\n",
      " [0.21794872]]\n",
      "re: 0.3165\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.11320755]\n",
      " [0.41111111]\n",
      " [0.32450331]\n",
      " [0.55128205]]\n",
      "re: 0.3500\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.02830189]\n",
      " [0.64444444]\n",
      " [0.21192053]\n",
      " [0.44871795]]\n",
      "re: 0.3333\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.30188679]\n",
      " [0.55555556]\n",
      " [0.05960265]\n",
      " [0.25641026]]\n",
      "re: 0.2934\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2852\n",
      "===================================================================\n",
      "K=8, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.05660377]\n",
      " [0.05555556]\n",
      " [0.21854305]\n",
      " [0.25641026]]\n",
      "re: 0.1468\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.01886792]\n",
      " [0.4       ]\n",
      " [0.31788079]\n",
      " [0.21794872]]\n",
      "re: 0.2387\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.05660377]\n",
      " [0.08888889]\n",
      " [0.27152318]\n",
      " [0.03846154]]\n",
      "re: 0.1139\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.02830189]\n",
      " [0.64444444]\n",
      " [0.21192053]\n",
      " [0.37179487]]\n",
      "re: 0.3141\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.30188679]\n",
      " [0.55555556]\n",
      " [0.0794702 ]\n",
      " [0.28205128]]\n",
      "re: 0.3047\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2236\n",
      "===================================================================\n",
      "K=8, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.05660377]\n",
      " [0.35555556]\n",
      " [0.17218543]\n",
      " [0.30769231]]\n",
      "re: 0.2230\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.09433962]\n",
      " [0.2       ]\n",
      " [0.56291391]\n",
      " [0.20512821]]\n",
      "re: 0.2656\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.06603774]\n",
      " [0.63333333]\n",
      " [0.28476821]\n",
      " [0.05128205]]\n",
      "re: 0.2589\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.02830189]\n",
      " [0.64444444]\n",
      " [0.21192053]\n",
      " [0.07692308]]\n",
      "re: 0.2404\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.29245283]\n",
      " [0.55555556]\n",
      " [0.15231788]\n",
      " [0.19230769]]\n",
      "re: 0.2982\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2572\n",
      "===================================================================\n",
      "K=8, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.05660377]\n",
      " [0.43333333]\n",
      " [0.03311258]\n",
      " [0.11538462]]\n",
      "re: 0.1596\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.04716981]\n",
      " [0.56666667]\n",
      " [1.        ]\n",
      " [0.37179487]]\n",
      "re: 0.4964\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.21698113]\n",
      " [0.1       ]\n",
      " [0.14569536]\n",
      " [0.42307692]]\n",
      "re: 0.2214\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.03773585]\n",
      " [0.37777778]\n",
      " [0.08609272]\n",
      " [0.3974359 ]]\n",
      "re: 0.2248\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.38679245]\n",
      " [0.        ]\n",
      " [0.16556291]\n",
      " [0.24358974]]\n",
      "re: 0.1990\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2602\n",
      "===================================================================\n",
      "K=8, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.02830189]\n",
      " [0.51111111]\n",
      " [0.29801325]\n",
      " [0.34615385]]\n",
      "re: 0.2959\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.16037736]\n",
      " [0.62222222]\n",
      " [0.08609272]\n",
      " [0.29487179]]\n",
      "re: 0.2909\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.10377358]\n",
      " [0.63333333]\n",
      " [0.1589404 ]\n",
      " [0.37179487]]\n",
      "re: 0.3170\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.12264151]\n",
      " [0.51111111]\n",
      " [0.06622517]\n",
      " [0.19230769]]\n",
      "re: 0.2231\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.76415094]\n",
      " [0.53333333]\n",
      " [0.16556291]\n",
      " [0.30769231]]\n",
      "re: 0.4427\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3139\n",
      "===================================================================\n",
      "K=8, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.04716981]\n",
      " [0.48888889]\n",
      " [0.15231788]\n",
      " [0.33333333]]\n",
      "re: 0.2554\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.53773585]\n",
      " [1.        ]\n",
      " [0.52980132]\n",
      " [0.29487179]]\n",
      "re: 0.5906\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.11320755]\n",
      " [0.63333333]\n",
      " [0.27152318]\n",
      " [0.26923077]]\n",
      "re: 0.3218\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.12264151]\n",
      " [0.11111111]\n",
      " [0.14569536]\n",
      " [0.23076923]]\n",
      "re: 0.1526\n",
      "------------------------------------------------------------------\n",
      "seed:4\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.00943396]\n",
      " [0.25555556]\n",
      " [1.        ]\n",
      " [0.        ]]\n",
      "re: 0.3162\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3273\n",
      "===================================================================\n",
      "K=8, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.04716981]\n",
      " [0.57777778]\n",
      " [0.72847682]\n",
      " [0.41025641]]\n",
      "re: 0.4409\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.58490566]\n",
      " [1.        ]\n",
      " [0.70198675]\n",
      " [0.29487179]]\n",
      "re: 0.6454\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.01886792]\n",
      " [0.01111111]\n",
      " [0.09933775]\n",
      " [0.05128205]]\n",
      "re: 0.0451\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.12264151]\n",
      " [0.6       ]\n",
      " [0.1986755 ]\n",
      " [0.15384615]]\n",
      "re: 0.2688\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.01886792]\n",
      " [0.31111111]\n",
      " [0.2384106 ]\n",
      " [0.23076923]]\n",
      "re: 0.1998\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3200\n",
      "===================================================================\n",
      "K=8, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.04716981]\n",
      " [0.43333333]\n",
      " [0.33112583]\n",
      " [0.14102564]]\n",
      "re: 0.2382\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.50943396]\n",
      " [0.54444444]\n",
      " [0.08609272]\n",
      " [0.02564103]]\n",
      "re: 0.2914\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.04716981]\n",
      " [0.34444444]\n",
      " [0.0794702 ]\n",
      " [0.23076923]]\n",
      "re: 0.1755\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.16037736]\n",
      " [0.71111111]\n",
      " [0.47019868]\n",
      " [0.24358974]]\n",
      "re: 0.3963\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.66037736]\n",
      " [0.61111111]\n",
      " [0.0397351 ]\n",
      " [0.41025641]]\n",
      "re: 0.4304\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3063\n",
      "===================================================================\n",
      "K=8, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.04716981]\n",
      " [0.23333333]\n",
      " [0.13907285]\n",
      " [0.37179487]]\n",
      "re: 0.1978\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.54444444]\n",
      " [0.06622517]\n",
      " [0.33333333]]\n",
      "re: 0.2360\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.05660377]\n",
      " [0.5       ]\n",
      " [0.06622517]\n",
      " [0.25641026]]\n",
      "re: 0.2198\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.30188679]\n",
      " [0.27777778]\n",
      " [0.20529801]\n",
      " [0.15384615]]\n",
      "re: 0.2347\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.08490566]\n",
      " [0.63333333]\n",
      " [0.2384106 ]\n",
      " [0.11538462]]\n",
      "re: 0.2680\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2313\n",
      "===================================================================\n",
      "K=8, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.03773585]\n",
      " [0.27777778]\n",
      " [0.11258278]\n",
      " [0.41025641]]\n",
      "re: 0.2096\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.6       ]\n",
      " [0.17218543]\n",
      " [1.        ]]\n",
      "re: 0.4430\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.66037736]\n",
      " [0.38888889]\n",
      " [0.03311258]\n",
      " [0.02564103]]\n",
      "re: 0.2770\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.30188679]\n",
      " [0.66666667]\n",
      " [0.48344371]\n",
      " [0.43589744]]\n",
      "re: 0.4720\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.01886792]\n",
      " [0.13333333]\n",
      " [0.14569536]\n",
      " [0.32051282]]\n",
      "re: 0.1546\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3112\n",
      "===================================================================\n",
      "K=8, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.05660377]\n",
      " [0.6       ]\n",
      " [0.76821192]\n",
      " [0.19230769]]\n",
      "re: 0.4043\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.54444444]\n",
      " [0.81456954]\n",
      " [0.34615385]]\n",
      "re: 0.4287\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.63207547]\n",
      " [0.13333333]\n",
      " [0.83443709]\n",
      " [0.20512821]]\n",
      "re: 0.4512\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.31132075]\n",
      " [0.51111111]\n",
      " [0.37748344]\n",
      " [0.23076923]]\n",
      "re: 0.3577\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.13207547]\n",
      " [0.61111111]\n",
      " [0.07284768]\n",
      " [0.41025641]]\n",
      "re: 0.3066\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3897\n",
      "===================================================================\n",
      "K=8, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.06603774]\n",
      " [0.08888889]\n",
      " [0.24503311]\n",
      " [0.19230769]]\n",
      "re: 0.1481\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.01886792]\n",
      " [0.51111111]\n",
      " [0.16556291]\n",
      " [0.20512821]]\n",
      "re: 0.2252\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.23584906]\n",
      " [0.1       ]\n",
      " [0.04635762]\n",
      " [0.17948718]]\n",
      "re: 0.1404\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.23584906]\n",
      " [0.4       ]\n",
      " [0.38410596]\n",
      " [0.33333333]]\n",
      "re: 0.3383\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.26415094]\n",
      " [0.61111111]\n",
      " [0.16556291]\n",
      " [0.26923077]]\n",
      "re: 0.3275\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2359\n",
      "===================================================================\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "aa391929718646958268f5cabda18f26",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8efa2f3a22f8479fb3f98250e1698789",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "K=16, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.04716981]\n",
      " [0.33333333]\n",
      " [0.02649007]\n",
      " [0.34615385]]\n",
      "re: 0.1883\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.61111111]\n",
      " [0.06622517]\n",
      " [0.03846154]]\n",
      "re: 0.1789\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.02830189]\n",
      " [0.6       ]\n",
      " [0.05960265]\n",
      " [0.33333333]]\n",
      "re: 0.2553\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.        ]\n",
      " [0.56666667]\n",
      " [0.68211921]\n",
      " [0.43589744]]\n",
      "re: 0.4212\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.18867925]\n",
      " [0.55555556]\n",
      " [0.25165563]\n",
      " [0.43589744]]\n",
      "re: 0.3579\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2803\n",
      "===================================================================\n",
      "K=16, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.04716981]\n",
      " [0.38888889]\n",
      " [0.02649007]\n",
      " [0.34615385]]\n",
      "re: 0.2022\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.57777778]\n",
      " [0.25165563]\n",
      " [0.26923077]]\n",
      "re: 0.2747\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.02830189]\n",
      " [0.6       ]\n",
      " [0.05960265]\n",
      " [0.30769231]]\n",
      "re: 0.2489\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.        ]\n",
      " [0.64444444]\n",
      " [0.05298013]\n",
      " [0.20512821]]\n",
      "re: 0.2256\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.18867925]\n",
      " [0.56666667]\n",
      " [0.03311258]\n",
      " [0.24358974]]\n",
      "re: 0.2580\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2419\n",
      "===================================================================\n",
      "K=16, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.04716981]\n",
      " [0.6       ]\n",
      " [0.07284768]\n",
      " [0.05128205]]\n",
      "re: 0.1928\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.61111111]\n",
      " [0.27152318]\n",
      " [0.29487179]]\n",
      "re: 0.2944\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.02830189]\n",
      " [0.55555556]\n",
      " [0.20529801]\n",
      " [0.20512821]]\n",
      "re: 0.2486\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.        ]\n",
      " [0.62222222]\n",
      " [0.27152318]\n",
      " [0.05128205]]\n",
      "re: 0.2363\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.01886792]\n",
      " [0.57777778]\n",
      " [0.34437086]\n",
      " [0.33333333]]\n",
      "re: 0.3186\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2581\n",
      "===================================================================\n",
      "K=16, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.04716981]\n",
      " [0.61111111]\n",
      " [0.1986755 ]\n",
      " [0.35897436]]\n",
      "re: 0.3040\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.01886792]\n",
      " [0.02222222]\n",
      " [0.25165563]\n",
      " [0.23076923]]\n",
      "re: 0.1309\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.56666667]\n",
      " [0.30463576]\n",
      " [0.15384615]]\n",
      "re: 0.2563\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [0.56666667]\n",
      " [0.1589404 ]\n",
      " [0.35897436]]\n",
      "re: 0.2735\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.17924528]\n",
      " [0.63333333]\n",
      " [0.04635762]\n",
      " [0.24358974]]\n",
      "re: 0.2756\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2481\n",
      "===================================================================\n",
      "K=16, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.04716981]\n",
      " [1.        ]\n",
      " [0.05298013]\n",
      " [0.11538462]]\n",
      "re: 0.3039\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.64444444]\n",
      " [0.1589404 ]\n",
      " [0.08974359]]\n",
      "re: 0.2233\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.14150943]\n",
      " [0.53333333]\n",
      " [0.17218543]\n",
      " [0.14102564]]\n",
      "re: 0.2470\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [0.04444444]\n",
      " [0.10596026]\n",
      " [0.35897436]]\n",
      "re: 0.1297\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.30188679]\n",
      " [0.55555556]\n",
      " [0.51655629]\n",
      " [0.15384615]]\n",
      "re: 0.3820\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2572\n",
      "===================================================================\n",
      "K=16, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.04716981]\n",
      " [1.        ]\n",
      " [0.05298013]\n",
      " [0.11538462]]\n",
      "re: 0.3039\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.64444444]\n",
      " [0.15231788]\n",
      " [0.23076923]]\n",
      "re: 0.2569\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.14150943]\n",
      " [0.53333333]\n",
      " [0.11258278]\n",
      " [0.25641026]]\n",
      "re: 0.2610\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [0.04444444]\n",
      " [0.10596026]\n",
      " [0.35897436]]\n",
      "re: 0.1297\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.26415094]\n",
      " [0.6       ]\n",
      " [0.47019868]\n",
      " [0.30769231]]\n",
      "re: 0.4105\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2724\n",
      "===================================================================\n",
      "K=16, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.05660377]\n",
      " [1.        ]\n",
      " [0.0397351 ]\n",
      " [0.11538462]]\n",
      "re: 0.3029\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [1.        ]\n",
      " [0.13907285]\n",
      " [0.08974359]]\n",
      "re: 0.3072\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.1509434 ]\n",
      " [0.01111111]\n",
      " [0.17880795]\n",
      " [0.14102564]]\n",
      "re: 0.1205\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [0.04444444]\n",
      " [0.28476821]\n",
      " [0.11538462]]\n",
      "re: 0.1135\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.08490566]\n",
      " [0.54444444]\n",
      " [0.09271523]\n",
      " [0.3974359 ]]\n",
      "re: 0.2799\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2248\n",
      "===================================================================\n",
      "K=16, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.        ]\n",
      " [0.53333333]\n",
      " [0.21854305]\n",
      " [0.43589744]]\n",
      "re: 0.2969\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.01111111]\n",
      " [0.22516556]\n",
      " [0.41025641]]\n",
      "re: 0.1616\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.00943396]\n",
      " [0.56666667]\n",
      " [0.06622517]\n",
      " [0.16666667]]\n",
      "re: 0.2022\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.13207547]\n",
      " [0.06666667]\n",
      " [0.26490066]\n",
      " [0.41025641]]\n",
      "re: 0.2185\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.03773585]\n",
      " [0.54444444]\n",
      " [0.2781457 ]\n",
      " [0.16666667]]\n",
      "re: 0.2567\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2272\n",
      "===================================================================\n",
      "K=16, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.        ]\n",
      " [0.6       ]\n",
      " [0.05960265]\n",
      " [0.44871795]]\n",
      "re: 0.2771\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.16666667]\n",
      " [0.1589404 ]\n",
      " [0.20512821]]\n",
      "re: 0.1327\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.02830189]\n",
      " [0.63333333]\n",
      " [0.21192053]\n",
      " [0.3974359 ]]\n",
      "re: 0.3177\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.01886792]\n",
      " [0.        ]\n",
      " [0.07284768]\n",
      " [0.37179487]]\n",
      "re: 0.1159\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.05660377]\n",
      " [0.64444444]\n",
      " [0.09933775]\n",
      " [0.16666667]]\n",
      "re: 0.2418\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2170\n",
      "===================================================================\n",
      "K=16, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.        ]\n",
      " [0.56666667]\n",
      " [0.10596026]\n",
      " [0.35897436]]\n",
      "re: 0.2579\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.14444444]\n",
      " [0.16556291]\n",
      " [0.24358974]]\n",
      "re: 0.1384\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.02830189]\n",
      " [0.53333333]\n",
      " [0.10596026]\n",
      " [0.3974359 ]]\n",
      "re: 0.2663\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.01886792]\n",
      " [0.        ]\n",
      " [0.07284768]\n",
      " [0.35897436]]\n",
      "re: 0.1127\n",
      "------------------------------------------------------------------\n",
      "seed:4\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.04716981]\n",
      " [0.02222222]\n",
      " [0.13907285]\n",
      " [0.24358974]]\n",
      "re: 0.1130\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.1776\n",
      "===================================================================\n",
      "K=16, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.        ]\n",
      " [0.63333333]\n",
      " [0.1192053 ]\n",
      " [0.02564103]]\n",
      "re: 0.1945\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.06666667]\n",
      " [0.25827815]\n",
      " [0.15384615]]\n",
      "re: 0.1197\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.02830189]\n",
      " [0.13333333]\n",
      " [0.0397351 ]\n",
      " [0.32051282]]\n",
      "re: 0.1305\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.01886792]\n",
      " [0.        ]\n",
      " [0.07284768]\n",
      " [0.29487179]]\n",
      "re: 0.0966\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.03773585]\n",
      " [0.02222222]\n",
      " [0.45033113]\n",
      " [0.15384615]]\n",
      "re: 0.1660\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.1415\n",
      "===================================================================\n",
      "K=16, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.05660377]\n",
      " [0.61111111]\n",
      " [0.28476821]\n",
      " [0.24358974]]\n",
      "re: 0.2990\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.57777778]\n",
      " [0.04635762]\n",
      " [0.20512821]]\n",
      "re: 0.2073\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.05660377]\n",
      " [0.33333333]\n",
      " [0.09933775]\n",
      " [0.34615385]]\n",
      "re: 0.2089\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.01886792]\n",
      " [0.6       ]\n",
      " [0.02649007]\n",
      " [0.26923077]]\n",
      "re: 0.2286\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.05660377]\n",
      " [0.23333333]\n",
      " [0.31125828]\n",
      " [0.15384615]]\n",
      "re: 0.1888\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2265\n",
      "===================================================================\n",
      "K=16, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.04716981]\n",
      " [0.61111111]\n",
      " [0.31125828]\n",
      " [0.16666667]]\n",
      "re: 0.2841\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.22641509]\n",
      " [0.43333333]\n",
      " [0.29801325]\n",
      " [0.30769231]]\n",
      "re: 0.3164\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.06603774]\n",
      " [0.42222222]\n",
      " [0.32450331]\n",
      " [0.12820513]]\n",
      "re: 0.2352\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.48113208]\n",
      " [0.        ]\n",
      " [0.33774834]\n",
      " [0.32051282]]\n",
      "re: 0.2848\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.31132075]\n",
      " [0.54444444]\n",
      " [0.26490066]\n",
      " [0.16666667]]\n",
      "re: 0.3218\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2885\n",
      "===================================================================\n",
      "K=16, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.03773585]\n",
      " [0.61111111]\n",
      " [0.31125828]\n",
      " [0.16666667]]\n",
      "re: 0.2817\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.22641509]\n",
      " [0.43333333]\n",
      " [0.29801325]\n",
      " [0.30769231]]\n",
      "re: 0.3164\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.06603774]\n",
      " [0.44444444]\n",
      " [0.11258278]\n",
      " [0.        ]]\n",
      "re: 0.1558\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.48113208]\n",
      " [0.        ]\n",
      " [0.33774834]\n",
      " [0.30769231]]\n",
      "re: 0.2816\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.01886792]\n",
      " [0.01111111]\n",
      " [0.0397351 ]\n",
      " [0.15384615]]\n",
      "re: 0.0559\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2183\n",
      "===================================================================\n",
      "K=16, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.01886792]\n",
      " [0.61111111]\n",
      " [0.31125828]\n",
      " [0.05128205]]\n",
      "re: 0.2481\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.08490566]\n",
      " [0.4       ]\n",
      " [0.30463576]\n",
      " [0.30769231]]\n",
      "re: 0.2743\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.0754717 ]\n",
      " [0.44444444]\n",
      " [0.05960265]\n",
      " [0.17948718]]\n",
      "re: 0.1898\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.35849057]\n",
      " [0.61111111]\n",
      " [0.22516556]\n",
      " [0.06410256]]\n",
      "re: 0.3147\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.17924528]\n",
      " [0.52222222]\n",
      " [0.09271523]\n",
      " [0.21794872]]\n",
      "re: 0.2530\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2560\n",
      "===================================================================\n",
      "K=16, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.02830189]\n",
      " [0.55555556]\n",
      " [0.31788079]\n",
      " [0.17948718]]\n",
      "re: 0.2703\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.61111111]\n",
      " [0.24503311]\n",
      " [0.43589744]]\n",
      "re: 0.3230\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.09433962]\n",
      " [0.33333333]\n",
      " [0.30463576]\n",
      " [0.32051282]]\n",
      "re: 0.2632\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.18867925]\n",
      " [0.46666667]\n",
      " [0.05960265]\n",
      " [0.38461538]]\n",
      "re: 0.2749\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.43333333]\n",
      " [0.10596026]\n",
      " [0.21794872]]\n",
      "re: 0.1893\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2641\n",
      "===================================================================\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b66f574c3bf240f09ba16b1294ae24e7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "K=16, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.03773585]\n",
      " [0.62222222]\n",
      " [0.1986755 ]\n",
      " [0.41025641]]\n",
      "re: 0.3172\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.03773585]\n",
      " [0.31111111]\n",
      " [0.47682119]\n",
      " [1.        ]]\n",
      "re: 0.4564\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.05660377]\n",
      " [0.05555556]\n",
      " [0.25827815]\n",
      " [0.34615385]]\n",
      "re: 0.1791\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.06603774]\n",
      " [0.63333333]\n",
      " [0.21854305]\n",
      " [0.37179487]]\n",
      "re: 0.3224\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.3490566 ]\n",
      " [0.55555556]\n",
      " [0.1589404 ]\n",
      " [0.61538462]]\n",
      "re: 0.4197\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3390\n",
      "===================================================================\n",
      "K=16, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.03773585]\n",
      " [0.63333333]\n",
      " [0.1986755 ]\n",
      " [0.41025641]]\n",
      "re: 0.3200\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.56603774]\n",
      " [0.62222222]\n",
      " [0.24503311]\n",
      " [0.3974359 ]]\n",
      "re: 0.4577\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.0754717 ]\n",
      " [0.11111111]\n",
      " [0.10596026]\n",
      " [0.87179487]]\n",
      "re: 0.2911\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.06603774]\n",
      " [0.63333333]\n",
      " [0.08609272]\n",
      " [0.88461538]]\n",
      "re: 0.4175\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.3490566 ]\n",
      " [0.21111111]\n",
      " [0.43708609]\n",
      " [0.38461538]]\n",
      "re: 0.3455\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3664\n",
      "===================================================================\n",
      "K=16, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.04716981]\n",
      " [0.62222222]\n",
      " [0.19205298]\n",
      " [0.42307692]]\n",
      "re: 0.3211\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.38679245]\n",
      " [0.63333333]\n",
      " [0.19205298]\n",
      " [0.37179487]]\n",
      "re: 0.3960\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.10377358]\n",
      " [0.02222222]\n",
      " [0.19205298]\n",
      " [0.3974359 ]]\n",
      "re: 0.1789\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.06603774]\n",
      " [0.06666667]\n",
      " [0.77483444]\n",
      " [0.42307692]]\n",
      "re: 0.3327\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.3490566 ]\n",
      " [0.51111111]\n",
      " [0.13245033]\n",
      " [0.5       ]]\n",
      "re: 0.3732\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3204\n",
      "===================================================================\n",
      "K=16, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.04716981]\n",
      " [0.21111111]\n",
      " [0.13907285]\n",
      " [0.17948718]]\n",
      "re: 0.1442\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.53333333]\n",
      " [0.26490066]\n",
      " [0.03846154]]\n",
      "re: 0.2092\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.20754717]\n",
      " [0.35555556]\n",
      " [0.08609272]\n",
      " [0.32051282]]\n",
      "re: 0.2424\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.05660377]\n",
      " [0.64444444]\n",
      " [0.27152318]\n",
      " [0.05128205]]\n",
      "re: 0.2560\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.3490566 ]\n",
      " [0.08888889]\n",
      " [0.06622517]\n",
      " [0.05128205]]\n",
      "re: 0.1389\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.1981\n",
      "===================================================================\n",
      "K=16, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.11320755]\n",
      " [0.42222222]\n",
      " [0.64238411]\n",
      " [0.06410256]]\n",
      "re: 0.3105\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.10377358]\n",
      " [0.61111111]\n",
      " [0.22516556]\n",
      " [0.3974359 ]]\n",
      "re: 0.3344\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.16037736]\n",
      " [0.4       ]\n",
      " [0.7615894 ]\n",
      " [0.33333333]]\n",
      "re: 0.4138\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.05660377]\n",
      " [0.33333333]\n",
      " [0.13907285]\n",
      " [0.85897436]]\n",
      "re: 0.3470\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.26415094]\n",
      " [0.61111111]\n",
      " [0.0397351 ]\n",
      " [0.19230769]]\n",
      "re: 0.2768\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3365\n",
      "===================================================================\n",
      "K=16, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.11320755]\n",
      " [0.34444444]\n",
      " [0.64238411]\n",
      " [0.        ]]\n",
      "re: 0.2750\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.10377358]\n",
      " [0.6       ]\n",
      " [0.22516556]\n",
      " [0.3974359 ]]\n",
      "re: 0.3316\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.16037736]\n",
      " [0.06666667]\n",
      " [0.05960265]\n",
      " [0.16666667]]\n",
      "re: 0.1133\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.05660377]\n",
      " [0.08888889]\n",
      " [0.12582781]\n",
      " [0.78205128]]\n",
      "re: 0.2633\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.26415094]\n",
      " [0.27777778]\n",
      " [0.11258278]\n",
      " [0.42307692]]\n",
      "re: 0.2694\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2505\n",
      "===================================================================\n",
      "K=16, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.11320755]\n",
      " [0.17777778]\n",
      " [0.64238411]\n",
      " [0.12820513]]\n",
      "re: 0.2654\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.10377358]\n",
      " [0.42222222]\n",
      " [0.23178808]\n",
      " [0.3974359 ]]\n",
      "re: 0.2888\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.16037736]\n",
      " [0.24444444]\n",
      " [0.02649007]\n",
      " [0.38461538]]\n",
      "re: 0.2040\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.05660377]\n",
      " [0.11111111]\n",
      " [0.05298013]\n",
      " [0.38461538]]\n",
      "re: 0.1513\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.26415094]\n",
      " [0.65555556]\n",
      " [0.14569536]\n",
      " [0.21794872]]\n",
      "re: 0.3208\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2461\n",
      "===================================================================\n",
      "K=16, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.11320755]\n",
      " [0.6       ]\n",
      " [0.1589404 ]\n",
      " [0.61538462]]\n",
      "re: 0.3719\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.0754717 ]\n",
      " [0.41111111]\n",
      " [0.08609272]\n",
      " [0.42307692]]\n",
      "re: 0.2489\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.1509434 ]\n",
      " [1.        ]\n",
      " [0.25827815]\n",
      " [0.24358974]]\n",
      "re: 0.4132\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.0754717 ]\n",
      " [0.3       ]\n",
      " [0.30463576]\n",
      " [0.12820513]]\n",
      "re: 0.2021\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.38679245]\n",
      " [0.31111111]\n",
      " [0.13907285]\n",
      " [0.70512821]]\n",
      "re: 0.3855\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3243\n",
      "===================================================================\n",
      "K=16, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.02830189]\n",
      " [0.53333333]\n",
      " [0.18543046]\n",
      " [0.03846154]]\n",
      "re: 0.1964\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.08490566]\n",
      " [0.62222222]\n",
      " [0.19205298]\n",
      " [0.25641026]]\n",
      "re: 0.2889\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.16981132]\n",
      " [1.        ]\n",
      " [0.20529801]\n",
      " [0.23076923]]\n",
      "re: 0.4015\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.09433962]\n",
      " [1.        ]\n",
      " [0.2781457 ]\n",
      " [0.57692308]]\n",
      "re: 0.4874\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.03773585]\n",
      " [0.28888889]\n",
      " [0.0794702 ]\n",
      " [0.19230769]]\n",
      "re: 0.1496\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3047\n",
      "===================================================================\n",
      "K=16, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.20754717]\n",
      " [0.52222222]\n",
      " [0.10596026]\n",
      " [0.41025641]]\n",
      "re: 0.3115\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.08490566]\n",
      " [0.52222222]\n",
      " [0.04635762]\n",
      " [0.42307692]]\n",
      "re: 0.2691\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.14150943]\n",
      " [0.14444444]\n",
      " [0.24503311]\n",
      " [0.11538462]]\n",
      "re: 0.1616\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.01886792]\n",
      " [0.58888889]\n",
      " [0.13907285]\n",
      " [0.26923077]]\n",
      "re: 0.2540\n",
      "------------------------------------------------------------------\n",
      "seed:4\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.5754717 ]\n",
      " [0.31111111]\n",
      " [0.31788079]\n",
      " [0.76923077]]\n",
      "re: 0.4934\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2979\n",
      "===================================================================\n",
      "K=16, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.37735849]\n",
      " [0.1       ]\n",
      " [0.22516556]\n",
      " [0.46153846]]\n",
      "re: 0.2910\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.08490566]\n",
      " [0.58888889]\n",
      " [0.18543046]\n",
      " [0.19230769]]\n",
      "re: 0.2629\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.69811321]\n",
      " [0.6       ]\n",
      " [0.30463576]\n",
      " [0.3974359 ]]\n",
      "re: 0.5000\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.17924528]\n",
      " [0.01111111]\n",
      " [0.25827815]\n",
      " [1.        ]]\n",
      "re: 0.3622\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.63207547]\n",
      " [0.54444444]\n",
      " [0.09271523]\n",
      " [0.38461538]]\n",
      "re: 0.4135\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3659\n",
      "===================================================================\n",
      "K=16, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.16037736]\n",
      " [0.61111111]\n",
      " [0.02649007]\n",
      " [0.25641026]]\n",
      "re: 0.2636\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.08490566]\n",
      " [0.47777778]\n",
      " [0.25165563]\n",
      " [0.29487179]]\n",
      "re: 0.2773\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.16981132]\n",
      " [0.2       ]\n",
      " [0.12582781]\n",
      " [0.33333333]]\n",
      "re: 0.2072\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.12264151]\n",
      " [0.61111111]\n",
      " [0.18543046]\n",
      " [0.14102564]]\n",
      "re: 0.2651\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.03773585]\n",
      " [0.62222222]\n",
      " [0.05298013]\n",
      " [0.37179487]]\n",
      "re: 0.2712\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2569\n",
      "===================================================================\n",
      "K=16, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.04716981]\n",
      " [0.57777778]\n",
      " [0.17218543]\n",
      " [0.20512821]]\n",
      "re: 0.2506\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.11320755]\n",
      " [0.58888889]\n",
      " [0.04635762]\n",
      " [0.33333333]]\n",
      "re: 0.2704\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.02830189]\n",
      " [0.14444444]\n",
      " [0.56291391]\n",
      " [0.25641026]]\n",
      "re: 0.2480\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [0.43708609]\n",
      " [0.15384615]]\n",
      "re: 0.6477\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.16981132]\n",
      " [0.16666667]\n",
      " [0.0397351 ]\n",
      " [0.33333333]]\n",
      "re: 0.1774\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3188\n",
      "===================================================================\n",
      "K=16, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.03773585]\n",
      " [0.51111111]\n",
      " [0.05960265]\n",
      " [0.23076923]]\n",
      "re: 0.2098\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.67924528]\n",
      " [0.56666667]\n",
      " [0.54966887]\n",
      " [0.11538462]]\n",
      "re: 0.4777\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.31132075]\n",
      " [0.51111111]\n",
      " [0.17880795]\n",
      " [0.28205128]]\n",
      "re: 0.3208\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [0.17880795]\n",
      " [0.83333333]]\n",
      "re: 0.7530\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.06603774]\n",
      " [0.15555556]\n",
      " [0.21192053]\n",
      " [0.42307692]]\n",
      "re: 0.2141\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3951\n",
      "===================================================================\n",
      "K=16, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.08490566]\n",
      " [0.6       ]\n",
      " [0.8013245 ]\n",
      " [0.28205128]]\n",
      "re: 0.4421\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.38679245]\n",
      " [0.56666667]\n",
      " [0.29801325]\n",
      " [0.41025641]]\n",
      "re: 0.4154\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.3490566 ]\n",
      " [0.56666667]\n",
      " [0.25827815]\n",
      " [0.37179487]]\n",
      "re: 0.3864\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [0.27152318]\n",
      " [1.        ]]\n",
      "re: 0.8179\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.05660377]\n",
      " [0.43333333]\n",
      " [0.59602649]\n",
      " [0.61538462]]\n",
      "re: 0.4253\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.4974\n",
      "===================================================================\n",
      "K=16, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.68867925]\n",
      " [0.61111111]\n",
      " [0.1192053 ]\n",
      " [0.23076923]]\n",
      "re: 0.4124\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.77358491]\n",
      " [0.32222222]\n",
      " [0.66225166]\n",
      " [0.20512821]]\n",
      "re: 0.4908\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.41509434]\n",
      " [0.5       ]\n",
      " [1.        ]\n",
      " [0.65384615]]\n",
      "re: 0.6422\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [0.63333333]\n",
      " [0.05298013]\n",
      " [0.20512821]]\n",
      "re: 0.4729\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.20754717]\n",
      " [1.        ]\n",
      " [0.24503311]\n",
      " [0.30769231]]\n",
      "re: 0.4401\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.4917\n",
      "===================================================================\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "af70d6c68d5e4210b5617b500157e542",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "af8d5d45e9f8444f895b7a3aa935af26",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "K=32, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.03773585]\n",
      " [0.62222222]\n",
      " [0.21854305]\n",
      " [0.26923077]]\n",
      "re: 0.2869\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.16666667]\n",
      " [1.        ]\n",
      " [0.43589744]]\n",
      "re: 0.4006\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.6       ]\n",
      " [0.03311258]\n",
      " [0.06410256]]\n",
      "re: 0.1743\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.53773585]\n",
      " [0.58888889]\n",
      " [0.71523179]\n",
      " [1.        ]]\n",
      "re: 0.7105\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.60377358]\n",
      " [0.57777778]\n",
      " [0.30463576]\n",
      " [0.23076923]]\n",
      "re: 0.4292\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.4003\n",
      "===================================================================\n",
      "K=32, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.03773585]\n",
      " [0.62222222]\n",
      " [0.21854305]\n",
      " [0.26923077]]\n",
      "re: 0.2869\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.23333333]\n",
      " [1.        ]\n",
      " [0.43589744]]\n",
      "re: 0.4173\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.00943396]\n",
      " [0.31111111]\n",
      " [0.22516556]\n",
      " [0.25641026]]\n",
      "re: 0.2005\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.53773585]\n",
      " [0.58888889]\n",
      " [0.71523179]\n",
      " [1.        ]]\n",
      "re: 0.7105\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.60377358]\n",
      " [0.52222222]\n",
      " [0.01986755]\n",
      " [0.01282051]]\n",
      "re: 0.2897\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3810\n",
      "===================================================================\n",
      "K=32, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.03773585]\n",
      " [0.46666667]\n",
      " [0.31125828]\n",
      " [0.05128205]]\n",
      "re: 0.2167\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.23333333]\n",
      " [0.23178808]\n",
      " [0.29487179]]\n",
      "re: 0.1900\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.00943396]\n",
      " [0.03333333]\n",
      " [0.37086093]\n",
      " [0.33333333]]\n",
      "re: 0.1867\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.53773585]\n",
      " [0.58888889]\n",
      " [0.1589404 ]\n",
      " [0.29487179]]\n",
      "re: 0.3951\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [0.18543046]\n",
      " [0.42307692]]\n",
      "re: 0.6521\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3281\n",
      "===================================================================\n",
      "K=32, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.03773585]\n",
      " [0.31111111]\n",
      " [0.26490066]\n",
      " [1.        ]]\n",
      "re: 0.4034\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.02830189]\n",
      " [0.55555556]\n",
      " [0.23178808]\n",
      " [0.23076923]]\n",
      "re: 0.2616\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.02830189]\n",
      " [0.61111111]\n",
      " [0.23178808]\n",
      " [1.        ]]\n",
      "re: 0.4678\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.        ]\n",
      " [0.64444444]\n",
      " [0.20529801]\n",
      " [1.        ]]\n",
      "re: 0.4624\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [0.33774834]\n",
      " [0.30769231]]\n",
      "re: 0.6614\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.4513\n",
      "===================================================================\n",
      "K=32, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.00943396]\n",
      " [0.64444444]\n",
      " [0.21854305]\n",
      " [0.29487179]]\n",
      "re: 0.2918\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.0754717 ]\n",
      " [0.64444444]\n",
      " [0.51655629]\n",
      " [0.67948718]]\n",
      "re: 0.4790\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.03773585]\n",
      " [0.62222222]\n",
      " [0.07284768]\n",
      " [0.01282051]]\n",
      "re: 0.1864\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [0.62222222]\n",
      " [0.0794702 ]\n",
      " [0.25641026]]\n",
      "re: 0.2419\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.61111111]\n",
      " [0.03311258]\n",
      " [0.14102564]]\n",
      "re: 0.1963\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2791\n",
      "===================================================================\n",
      "K=32, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.00943396]\n",
      " [0.64444444]\n",
      " [0.21854305]\n",
      " [0.29487179]]\n",
      "re: 0.2918\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.02830189]\n",
      " [0.64444444]\n",
      " [0.51655629]\n",
      " [0.02564103]]\n",
      "re: 0.3037\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.03773585]\n",
      " [0.62222222]\n",
      " [0.07284768]\n",
      " [0.01282051]]\n",
      "re: 0.1864\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [0.62222222]\n",
      " [0.07284768]\n",
      " [0.24358974]]\n",
      "re: 0.2370\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.61111111]\n",
      " [0.0397351 ]\n",
      " [0.14102564]]\n",
      "re: 0.1980\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2434\n",
      "===================================================================\n",
      "K=32, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.00943396]\n",
      " [0.64444444]\n",
      " [0.21854305]\n",
      " [0.01282051]]\n",
      "re: 0.2213\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.06603774]\n",
      " [0.64444444]\n",
      " [0.10596026]\n",
      " [0.23076923]]\n",
      "re: 0.2618\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.03773585]\n",
      " [0.62222222]\n",
      " [0.0794702 ]\n",
      " [0.01282051]]\n",
      "re: 0.1881\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.02830189]\n",
      " [0.63333333]\n",
      " [0.26490066]\n",
      " [0.41025641]]\n",
      "re: 0.3342\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.61111111]\n",
      " [0.34437086]\n",
      " [0.03846154]]\n",
      "re: 0.2485\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2508\n",
      "===================================================================\n",
      "K=32, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.01886792]\n",
      " [0.42222222]\n",
      " [0.25165563]\n",
      " [0.15384615]]\n",
      "re: 0.2116\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.08490566]\n",
      " [0.64444444]\n",
      " [0.27152318]\n",
      " [0.26923077]]\n",
      "re: 0.3175\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.64444444]\n",
      " [0.05298013]\n",
      " [0.80769231]]\n",
      "re: 0.3763\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.        ]\n",
      " [0.6       ]\n",
      " [0.21854305]\n",
      " [0.43589744]]\n",
      "re: 0.3136\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.55555556]\n",
      " [0.30463576]\n",
      " [1.        ]]\n",
      "re: 0.4650\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3368\n",
      "===================================================================\n",
      "K=32, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.04716981]\n",
      " [0.3       ]\n",
      " [0.23178808]\n",
      " [0.17948718]]\n",
      "re: 0.1896\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.04716981]\n",
      " [0.56666667]\n",
      " [0.18543046]\n",
      " [0.15384615]]\n",
      "re: 0.2383\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.01886792]\n",
      " [0.56666667]\n",
      " [0.25827815]\n",
      " [0.17948718]]\n",
      "re: 0.2558\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.17924528]\n",
      " [0.61111111]\n",
      " [0.24503311]\n",
      " [0.02564103]]\n",
      "re: 0.2653\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.        ]\n",
      " [0.62222222]\n",
      " [0.17880795]\n",
      " [0.41025641]]\n",
      "re: 0.5528\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3004\n",
      "===================================================================\n",
      "K=32, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.03773585]\n",
      " [0.64444444]\n",
      " [0.2384106 ]\n",
      " [0.        ]]\n",
      "re: 0.2301\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.04716981]\n",
      " [0.64444444]\n",
      " [0.1589404 ]\n",
      " [0.23076923]]\n",
      "re: 0.2703\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.01886792]\n",
      " [0.56666667]\n",
      " [0.25827815]\n",
      " [0.19230769]]\n",
      "re: 0.2590\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.17924528]\n",
      " [0.61111111]\n",
      " [0.24503311]\n",
      " [0.02564103]]\n",
      "re: 0.2653\n",
      "------------------------------------------------------------------\n",
      "seed:4\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[1.        ]\n",
      " [0.62222222]\n",
      " [0.17880795]\n",
      " [0.42307692]]\n",
      "re: 0.5560\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3162\n",
      "===================================================================\n",
      "K=32, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.04716981]\n",
      " [0.62222222]\n",
      " [0.10596026]\n",
      " [0.30769231]]\n",
      "re: 0.2708\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.04716981]\n",
      " [0.64444444]\n",
      " [0.1589404 ]\n",
      " [0.21794872]]\n",
      "re: 0.2671\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.00943396]\n",
      " [0.57777778]\n",
      " [0.25827815]\n",
      " [0.16666667]]\n",
      "re: 0.2530\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.16981132]\n",
      " [0.58888889]\n",
      " [0.24503311]\n",
      " [0.02564103]]\n",
      "re: 0.2573\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.        ]\n",
      " [0.62222222]\n",
      " [0.17218543]\n",
      " [0.42307692]]\n",
      "re: 0.5544\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3205\n",
      "===================================================================\n",
      "K=32, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.05660377]\n",
      " [0.61111111]\n",
      " [0.10596026]\n",
      " [0.01282051]]\n",
      "re: 0.1966\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.63333333]\n",
      " [0.31788079]\n",
      " [0.17948718]]\n",
      "re: 0.2850\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.0754717 ]\n",
      " [0.52222222]\n",
      " [0.2384106 ]\n",
      " [0.20512821]]\n",
      "re: 0.2603\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.        ]\n",
      " [0.56666667]\n",
      " [0.04635762]\n",
      " [0.21794872]]\n",
      "re: 0.2077\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.        ]\n",
      " [0.62222222]\n",
      " [0.31788079]\n",
      " [0.11538462]]\n",
      "re: 0.5139\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2927\n",
      "===================================================================\n",
      "K=32, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.43396226]\n",
      " [0.62222222]\n",
      " [0.1589404 ]\n",
      " [0.21794872]]\n",
      "re: 0.3583\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.28888889]\n",
      " [0.81456954]\n",
      " [0.08974359]]\n",
      "re: 0.3007\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.00943396]\n",
      " [0.58888889]\n",
      " [0.49668874]\n",
      " [0.12820513]]\n",
      "re: 0.3058\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.5       ]\n",
      " [0.6       ]\n",
      " [0.17880795]\n",
      " [0.12820513]]\n",
      "re: 0.3518\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.63333333]\n",
      " [0.11258278]\n",
      " [0.41025641]]\n",
      "re: 0.2914\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3216\n",
      "===================================================================\n",
      "K=32, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.43396226]\n",
      " [0.62222222]\n",
      " [0.1589404 ]\n",
      " [0.21794872]]\n",
      "re: 0.3583\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.56666667]\n",
      " [0.25827815]\n",
      " [0.32051282]]\n",
      "re: 0.2887\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.00943396]\n",
      " [0.58888889]\n",
      " [0.50331126]\n",
      " [0.12820513]]\n",
      "re: 0.3075\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.5       ]\n",
      " [0.6       ]\n",
      " [0.17880795]\n",
      " [0.12820513]]\n",
      "re: 0.3518\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.63333333]\n",
      " [0.11258278]\n",
      " [0.41025641]]\n",
      "re: 0.2914\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3195\n",
      "===================================================================\n",
      "K=32, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.43396226]\n",
      " [0.63333333]\n",
      " [0.1589404 ]\n",
      " [0.25641026]]\n",
      "re: 0.3707\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.63333333]\n",
      " [0.30463576]\n",
      " [0.16666667]]\n",
      "re: 0.2762\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.00943396]\n",
      " [0.58888889]\n",
      " [0.50331126]\n",
      " [0.12820513]]\n",
      "re: 0.3075\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.58490566]\n",
      " [0.6       ]\n",
      " [0.64900662]\n",
      " [0.        ]]\n",
      "re: 0.4585\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.61111111]\n",
      " [0.50331126]\n",
      " [0.06410256]]\n",
      "re: 0.2970\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3419\n",
      "===================================================================\n",
      "K=32, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.06603774]\n",
      " [0.56666667]\n",
      " [0.1986755 ]\n",
      " [0.23076923]]\n",
      "re: 0.2655\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.48888889]\n",
      " [0.31125828]\n",
      " [0.        ]]\n",
      "re: 0.2000\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.6       ]\n",
      " [0.24503311]\n",
      " [0.29487179]]\n",
      "re: 0.2850\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.02830189]\n",
      " [0.54444444]\n",
      " [0.25165563]\n",
      " [0.12820513]]\n",
      "re: 0.2382\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.58888889]\n",
      " [0.1589404 ]\n",
      " [0.12820513]]\n",
      "re: 0.2190\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2415\n",
      "===================================================================\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b0a610221f68451881e3c174d8784d63",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "K=32, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.38679245]\n",
      " [0.62222222]\n",
      " [0.88741722]\n",
      " [0.43589744]]\n",
      "re: 0.5831\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.54444444]\n",
      " [0.17218543]\n",
      " [0.62820513]]\n",
      "re: 0.3362\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.12264151]\n",
      " [0.82222222]\n",
      " [0.18543046]\n",
      " [0.28205128]]\n",
      "re: 0.3531\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.28301887]\n",
      " [0.33333333]\n",
      " [0.18543046]\n",
      " [0.88461538]]\n",
      "re: 0.4216\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.        ]\n",
      " [0.36666667]\n",
      " [0.1589404 ]\n",
      " [0.38461538]]\n",
      "re: 0.4776\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.4343\n",
      "===================================================================\n",
      "K=32, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.38679245]\n",
      " [0.5       ]\n",
      " [0.15231788]\n",
      " [1.        ]]\n",
      "re: 0.5098\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.87777778]\n",
      " [0.12582781]\n",
      " [0.17948718]]\n",
      "re: 0.2958\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.12264151]\n",
      " [0.04444444]\n",
      " [0.13245033]\n",
      " [0.01282051]]\n",
      "re: 0.0781\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.28301887]\n",
      " [0.66666667]\n",
      " [0.22516556]\n",
      " [0.74358974]]\n",
      "re: 0.4796\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.        ]\n",
      " [0.14444444]\n",
      " [0.11258278]\n",
      " [0.30769231]]\n",
      "re: 0.3912\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3509\n",
      "===================================================================\n",
      "K=32, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.38679245]\n",
      " [0.55555556]\n",
      " [0.13245033]\n",
      " [0.28205128]]\n",
      "re: 0.3392\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.08888889]\n",
      " [0.27152318]\n",
      " [0.34615385]]\n",
      "re: 0.1790\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.12264151]\n",
      " [0.58888889]\n",
      " [0.14569536]\n",
      " [0.32051282]]\n",
      "re: 0.2944\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.28301887]\n",
      " [0.28888889]\n",
      " [0.16556291]\n",
      " [0.02564103]]\n",
      "re: 0.1908\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.        ]\n",
      " [0.14444444]\n",
      " [0.17880795]\n",
      " [0.41025641]]\n",
      "re: 0.4334\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2874\n",
      "===================================================================\n",
      "K=32, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.38679245]\n",
      " [0.01111111]\n",
      " [0.19205298]\n",
      " [0.08974359]]\n",
      "re: 0.1699\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.28888889]\n",
      " [1.        ]\n",
      " [0.53846154]]\n",
      "re: 0.4592\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.12264151]\n",
      " [0.84444444]\n",
      " [0.18543046]\n",
      " [0.05128205]]\n",
      "re: 0.3009\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.28301887]\n",
      " [0.7       ]\n",
      " [0.09933775]\n",
      " [0.78205128]]\n",
      "re: 0.4661\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.        ]\n",
      " [0.04444444]\n",
      " [0.01986755]\n",
      " [0.03846154]]\n",
      "re: 0.2757\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3344\n",
      "===================================================================\n",
      "K=32, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.03773585]\n",
      " [0.68888889]\n",
      " [0.23178808]\n",
      " [0.35897436]]\n",
      "re: 0.3293\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.14150943]\n",
      " [0.62222222]\n",
      " [0.50993377]\n",
      " [0.38461538]]\n",
      "re: 0.4146\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.10377358]\n",
      " [0.31111111]\n",
      " [0.10596026]\n",
      " [0.42307692]]\n",
      "re: 0.2360\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [0.4       ]\n",
      " [0.05960265]\n",
      " [0.32051282]]\n",
      "re: 0.4450\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.02830189]\n",
      " [0.18888889]\n",
      " [0.1192053 ]\n",
      " [0.01282051]]\n",
      "re: 0.0873\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3024\n",
      "===================================================================\n",
      "K=32, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.03773585]\n",
      " [0.7       ]\n",
      " [0.2384106 ]\n",
      " [0.08974359]]\n",
      "re: 0.2665\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.14150943]\n",
      " [0.62222222]\n",
      " [0.44370861]\n",
      " [0.08974359]]\n",
      "re: 0.3243\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.10377358]\n",
      " [0.62222222]\n",
      " [0.16556291]\n",
      " [0.3974359 ]]\n",
      "re: 0.3222\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [0.33333333]\n",
      " [0.15231788]\n",
      " [0.12820513]]\n",
      "re: 0.4035\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.02830189]\n",
      " [0.06666667]\n",
      " [0.04635762]\n",
      " [0.20512821]]\n",
      "re: 0.0866\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2806\n",
      "===================================================================\n",
      "K=32, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.03773585]\n",
      " [0.2       ]\n",
      " [0.28476821]\n",
      " [0.55128205]]\n",
      "re: 0.2684\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.14150943]\n",
      " [0.62222222]\n",
      " [0.44370861]\n",
      " [0.64102564]]\n",
      "re: 0.4621\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.10377358]\n",
      " [0.68888889]\n",
      " [0.16556291]\n",
      " [0.43589744]]\n",
      "re: 0.3485\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [0.48888889]\n",
      " [0.15231788]\n",
      " [0.20512821]]\n",
      "re: 0.4616\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.02830189]\n",
      " [0.41111111]\n",
      " [0.11258278]\n",
      " [0.62820513]]\n",
      "re: 0.2951\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3671\n",
      "===================================================================\n",
      "K=32, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.02830189]\n",
      " [0.33333333]\n",
      " [0.1192053 ]\n",
      " [0.23076923]]\n",
      "re: 0.1779\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.14150943]\n",
      " [0.62222222]\n",
      " [0.16556291]\n",
      " [0.08974359]]\n",
      "re: 0.2548\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.10377358]\n",
      " [0.66666667]\n",
      " [0.16556291]\n",
      " [0.3974359 ]]\n",
      "re: 0.3334\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [0.53333333]\n",
      " [0.18543046]\n",
      " [0.43589744]]\n",
      "re: 0.5387\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.02830189]\n",
      " [0.03333333]\n",
      " [0.25165563]\n",
      " [0.57692308]]\n",
      "re: 0.2226\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3054\n",
      "===================================================================\n",
      "K=32, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [0.06622517]\n",
      " [0.24358974]]\n",
      "re: 0.5775\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [0.13907285]\n",
      " [1.        ]]\n",
      "re: 0.7848\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.00943396]\n",
      " [0.01111111]\n",
      " [0.05960265]\n",
      " [0.88461538]]\n",
      "re: 0.2412\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [0.16556291]\n",
      " [0.34615385]]\n",
      "re: 0.6279\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [0.73509934]\n",
      " [0.52564103]]\n",
      "re: 0.8152\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.6093\n",
      "===================================================================\n",
      "K=32, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[1.        ]\n",
      " [0.57777778]\n",
      " [1.        ]\n",
      " [1.        ]]\n",
      "re: 0.8944\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [0.04635762]\n",
      " [1.        ]]\n",
      "re: 0.7616\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.00943396]\n",
      " [1.        ]\n",
      " [0.24503311]\n",
      " [0.23076923]]\n",
      "re: 0.3713\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.46226415]\n",
      " [1.        ]\n",
      " [0.02649007]\n",
      " [0.21794872]]\n",
      "re: 0.4267\n",
      "------------------------------------------------------------------\n",
      "seed:4\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[1.       ]\n",
      " [0.6      ]\n",
      " [0.0794702]\n",
      " [1.       ]]\n",
      "re: 0.6699\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.6248\n",
      "===================================================================\n",
      "K=32, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [0.07284768]\n",
      " [0.24358974]]\n",
      "re: 0.5791\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [0.02649007]\n",
      " [0.1025641 ]]\n",
      "re: 0.5323\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.63333333]\n",
      " [1.        ]\n",
      " [0.29487179]]\n",
      "re: 0.4821\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.13207547]\n",
      " [1.        ]\n",
      " [0.21854305]\n",
      " [1.        ]]\n",
      "re: 0.5877\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [0.31788079]\n",
      " [0.12820513]]\n",
      "re: 0.6115\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.5585\n",
      "===================================================================\n",
      "K=32, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [0.1192053 ]\n",
      " [0.42307692]]\n",
      "re: 0.6356\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [0.13907285]\n",
      " [0.26923077]]\n",
      "re: 0.6021\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.17924528]\n",
      " [0.63333333]\n",
      " [1.        ]\n",
      " [1.        ]]\n",
      "re: 0.7031\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [0.25165563]\n",
      " [1.        ]]\n",
      "re: 0.8129\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.        ]\n",
      " [0.64444444]\n",
      " [0.75496689]\n",
      " [1.        ]]\n",
      "re: 0.8499\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.7207\n",
      "===================================================================\n",
      "K=32, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.50943396]\n",
      " [0.05555556]\n",
      " [0.76821192]\n",
      " [0.21794872]]\n",
      "re: 0.3878\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.0754717 ]\n",
      " [1.        ]\n",
      " [0.45695364]\n",
      " [0.30769231]]\n",
      "re: 0.4600\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.       ]\n",
      " [1.       ]\n",
      " [1.       ]\n",
      " [0.3974359]]\n",
      "re: 0.8494\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [0.29139073]\n",
      " [1.        ]]\n",
      "re: 0.8228\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.20754717]\n",
      " [0.48888889]\n",
      " [0.19205298]\n",
      " [1.        ]]\n",
      "re: 0.4721\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.5984\n",
      "===================================================================\n",
      "K=32, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.01886792]\n",
      " [1.        ]\n",
      " [0.05960265]\n",
      " [0.20512821]]\n",
      "re: 0.3209\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.09433962]\n",
      " [1.        ]\n",
      " [0.27152318]\n",
      " [1.        ]]\n",
      "re: 0.5915\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [1.        ]\n",
      " [0.66666667]]\n",
      "re: 0.9167\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [0.24444444]\n",
      " [1.        ]\n",
      " [0.06410256]]\n",
      "re: 0.5771\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.16981132]\n",
      " [0.63333333]\n",
      " [0.1986755 ]\n",
      " [1.        ]]\n",
      "re: 0.5005\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.5813\n",
      "===================================================================\n",
      "K=32, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.73584906]\n",
      " [0.22222222]\n",
      " [1.        ]\n",
      " [1.        ]]\n",
      "re: 0.7395\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [1.        ]\n",
      " [0.08609272]\n",
      " [0.08974359]]\n",
      "re: 0.2940\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [1.        ]\n",
      " [0.88461538]]\n",
      "re: 0.9712\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [0.3       ]\n",
      " [0.17880795]\n",
      " [0.42307692]]\n",
      "re: 0.4755\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.17924528]\n",
      " [0.53333333]\n",
      " [0.1589404 ]\n",
      " [0.38461538]]\n",
      "re: 0.3140\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.5588\n",
      "===================================================================\n",
      "K=32, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.25471698]\n",
      " [0.23333333]\n",
      " [1.        ]\n",
      " [1.        ]]\n",
      "re: 0.6220\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.06603774]\n",
      " [1.        ]\n",
      " [0.31788079]\n",
      " [0.26923077]]\n",
      "re: 0.4133\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [1.        ]\n",
      " [0.41025641]]\n",
      "re: 0.8526\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [0.64444444]\n",
      " [1.        ]\n",
      " [0.        ]]\n",
      "re: 0.6611\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.18867925]\n",
      " [0.6       ]\n",
      " [1.        ]\n",
      " [0.65384615]]\n",
      "re: 0.6106\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.6319\n",
      "===================================================================\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "fbe47b79a3fb4418a97bcb10797db4da",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "159c32b49fae46f49141caa5c2f371d1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "K=64, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.24528302]\n",
      " [0.51111111]\n",
      " [0.17880795]\n",
      " [0.01282051]]\n",
      "re: 0.2370\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.13207547]\n",
      " [0.07777778]\n",
      " [0.30463576]\n",
      " [1.        ]]\n",
      "re: 0.3786\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.56603774]\n",
      " [0.57777778]\n",
      " [0.01986755]\n",
      " [0.33333333]]\n",
      "re: 0.3743\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.28301887]\n",
      " [0.62222222]\n",
      " [0.8807947 ]\n",
      " [0.37179487]]\n",
      "re: 0.5395\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.02830189]\n",
      " [0.64444444]\n",
      " [0.33774834]\n",
      " [0.38461538]]\n",
      "re: 0.3488\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3756\n",
      "===================================================================\n",
      "K=64, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.24528302]\n",
      " [0.51111111]\n",
      " [0.77483444]\n",
      " [0.41025641]]\n",
      "re: 0.4854\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.13207547]\n",
      " [0.62222222]\n",
      " [0.69536424]\n",
      " [0.41025641]]\n",
      "re: 0.4650\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.56603774]\n",
      " [0.57777778]\n",
      " [0.01986755]\n",
      " [0.33333333]]\n",
      "re: 0.3743\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.28301887]\n",
      " [0.62222222]\n",
      " [0.8807947 ]\n",
      " [0.37179487]]\n",
      "re: 0.5395\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.02830189]\n",
      " [0.64444444]\n",
      " [0.33774834]\n",
      " [0.03846154]]\n",
      "re: 0.2622\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.4253\n",
      "===================================================================\n",
      "K=64, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.24528302]\n",
      " [0.52222222]\n",
      " [0.14569536]\n",
      " [0.30769231]]\n",
      "re: 0.3052\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.27358491]\n",
      " [0.57777778]\n",
      " [0.08609272]\n",
      " [0.26923077]]\n",
      "re: 0.3017\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.40566038]\n",
      " [0.52222222]\n",
      " [0.51655629]\n",
      " [0.42307692]]\n",
      "re: 0.4669\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.28301887]\n",
      " [0.62222222]\n",
      " [0.51655629]\n",
      " [0.33333333]]\n",
      "re: 0.4388\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.02830189]\n",
      " [0.64444444]\n",
      " [0.33774834]\n",
      " [0.35897436]]\n",
      "re: 0.3424\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3710\n",
      "===================================================================\n",
      "K=64, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.       ]\n",
      " [1.       ]\n",
      " [0.0397351]\n",
      " [0.       ]]\n",
      "re: 0.2599\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.01886792]\n",
      " [0.48888889]\n",
      " [0.22516556]\n",
      " [0.17948718]]\n",
      "re: 0.2281\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.26415094]\n",
      " [0.34444444]\n",
      " [0.27152318]\n",
      " [0.34615385]]\n",
      "re: 0.3066\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.0754717 ]\n",
      " [0.62222222]\n",
      " [0.11258278]\n",
      " [0.        ]]\n",
      "re: 0.2026\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.37777778]\n",
      " [0.14569536]\n",
      " [0.33333333]]\n",
      "re: 0.2142\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2423\n",
      "===================================================================\n",
      "K=64, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.02830189]\n",
      " [0.45555556]\n",
      " [0.21192053]\n",
      " [0.35897436]]\n",
      "re: 0.2637\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.05660377]\n",
      " [0.43333333]\n",
      " [0.01986755]\n",
      " [0.24358974]]\n",
      "re: 0.1883\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.55555556]\n",
      " [0.17880795]\n",
      " [0.24358974]]\n",
      "re: 0.2445\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.06603774]\n",
      " [0.56666667]\n",
      " [0.12582781]\n",
      " [0.05128205]]\n",
      "re: 0.2025\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.08888889]\n",
      " [0.62251656]\n",
      " [0.33333333]]\n",
      "re: 0.2635\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2325\n",
      "===================================================================\n",
      "K=64, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.02830189]\n",
      " [0.45555556]\n",
      " [0.21192053]\n",
      " [0.35897436]]\n",
      "re: 0.2637\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.05660377]\n",
      " [0.43333333]\n",
      " [0.01986755]\n",
      " [0.24358974]]\n",
      "re: 0.1883\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.55555556]\n",
      " [0.17880795]\n",
      " [0.24358974]]\n",
      "re: 0.2445\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.03773585]\n",
      " [0.56666667]\n",
      " [0.12582781]\n",
      " [0.11538462]]\n",
      "re: 0.2114\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.15555556]\n",
      " [0.07284768]\n",
      " [0.33333333]]\n",
      "re: 0.1428\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2101\n",
      "===================================================================\n",
      "K=64, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.00943396]\n",
      " [0.45555556]\n",
      " [0.21192053]\n",
      " [0.41025641]]\n",
      "re: 0.2718\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.05660377]\n",
      " [0.43333333]\n",
      " [0.01986755]\n",
      " [0.24358974]]\n",
      "re: 0.1883\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.55555556]\n",
      " [0.18543046]\n",
      " [0.23076923]]\n",
      "re: 0.2429\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.03773585]\n",
      " [0.56666667]\n",
      " [0.15231788]\n",
      " [0.16666667]]\n",
      "re: 0.2308\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.35555556]\n",
      " [0.19205298]\n",
      " [0.02564103]]\n",
      "re: 0.1457\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2159\n",
      "===================================================================\n",
      "K=64, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.10377358]\n",
      " [0.6       ]\n",
      " [0.18543046]\n",
      " [0.01282051]]\n",
      "re: 0.2255\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.0754717 ]\n",
      " [0.06666667]\n",
      " [0.21854305]\n",
      " [0.34615385]]\n",
      "re: 0.1767\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.        ]\n",
      " [0.55555556]\n",
      " [0.22516556]\n",
      " [0.29487179]]\n",
      "re: 0.2689\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [0.36666667]\n",
      " [0.26490066]\n",
      " [0.12820513]]\n",
      "re: 0.1923\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.01886792]\n",
      " [0.58888889]\n",
      " [0.25165563]\n",
      " [0.43589744]]\n",
      "re: 0.3238\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2374\n",
      "===================================================================\n",
      "K=64, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.        ]\n",
      " [0.63333333]\n",
      " [0.17880795]\n",
      " [0.1025641 ]]\n",
      "re: 0.2287\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.12264151]\n",
      " [0.38888889]\n",
      " [0.26490066]\n",
      " [0.15384615]]\n",
      "re: 0.2326\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.52830189]\n",
      " [0.15555556]\n",
      " [0.29139073]\n",
      " [0.07692308]]\n",
      "re: 0.2630\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.06603774]\n",
      " [0.52222222]\n",
      " [0.67549669]\n",
      " [0.07692308]]\n",
      "re: 0.3352\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.62222222]\n",
      " [0.15231788]\n",
      " [0.01282051]]\n",
      "re: 0.1992\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2517\n",
      "===================================================================\n",
      "K=64, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.        ]\n",
      " [0.63333333]\n",
      " [0.17880795]\n",
      " [0.05128205]]\n",
      "re: 0.2159\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.12264151]\n",
      " [0.35555556]\n",
      " [0.18543046]\n",
      " [0.38461538]]\n",
      "re: 0.2621\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.52830189]\n",
      " [0.45555556]\n",
      " [0.21192053]\n",
      " [0.32051282]]\n",
      "re: 0.3791\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.06603774]\n",
      " [0.52222222]\n",
      " [0.67549669]\n",
      " [0.25641026]]\n",
      "re: 0.3800\n",
      "------------------------------------------------------------------\n",
      "seed:4\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.00943396]\n",
      " [0.62222222]\n",
      " [0.1589404 ]\n",
      " [0.01282051]]\n",
      "re: 0.2009\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2876\n",
      "===================================================================\n",
      "K=64, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.        ]\n",
      " [0.63333333]\n",
      " [0.17880795]\n",
      " [0.14102564]]\n",
      "re: 0.2383\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.05660377]\n",
      " [0.08888889]\n",
      " [0.33112583]\n",
      " [0.        ]]\n",
      "re: 0.1192\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.32075472]\n",
      " [0.47777778]\n",
      " [0.0397351 ]\n",
      " [0.17948718]]\n",
      "re: 0.2544\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.06603774]\n",
      " [0.52222222]\n",
      " [0.62251656]\n",
      " [0.42307692]]\n",
      "re: 0.4085\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.62222222]\n",
      " [0.15231788]\n",
      " [0.01282051]]\n",
      "re: 0.1992\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2439\n",
      "===================================================================\n",
      "K=64, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.        ]\n",
      " [0.61111111]\n",
      " [0.2781457 ]\n",
      " [0.37179487]]\n",
      "re: 0.3153\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.01886792]\n",
      " [0.36666667]\n",
      " [0.22516556]\n",
      " [0.35897436]]\n",
      "re: 0.2424\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.02830189]\n",
      " [0.01111111]\n",
      " [0.21192053]\n",
      " [0.32051282]]\n",
      "re: 0.1430\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.01886792]\n",
      " [0.63333333]\n",
      " [0.23178808]\n",
      " [0.15384615]]\n",
      "re: 0.2595\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.55555556]\n",
      " [0.30463576]\n",
      " [0.38461538]]\n",
      "re: 0.3136\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2547\n",
      "===================================================================\n",
      "K=64, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.12264151]\n",
      " [0.53333333]\n",
      " [0.30463576]\n",
      " [0.1025641 ]]\n",
      "re: 0.2658\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.05660377]\n",
      " [0.38888889]\n",
      " [0.74834437]\n",
      " [0.05128205]]\n",
      "re: 0.3113\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.04716981]\n",
      " [0.4       ]\n",
      " [0.32450331]\n",
      " [0.19230769]]\n",
      "re: 0.2410\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.27358491]\n",
      " [0.47777778]\n",
      " [0.2384106 ]\n",
      " [0.01282051]]\n",
      "re: 0.2506\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.02830189]\n",
      " [0.64444444]\n",
      " [0.22516556]\n",
      " [0.38461538]]\n",
      "re: 0.3206\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2779\n",
      "===================================================================\n",
      "K=64, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.12264151]\n",
      " [0.53333333]\n",
      " [0.30463576]\n",
      " [0.1025641 ]]\n",
      "re: 0.2658\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.05660377]\n",
      " [0.38888889]\n",
      " [0.74834437]\n",
      " [0.05128205]]\n",
      "re: 0.3113\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.01886792]\n",
      " [0.07777778]\n",
      " [0.25827815]\n",
      " [0.26923077]]\n",
      "re: 0.1560\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.27358491]\n",
      " [0.07777778]\n",
      " [0.05960265]\n",
      " [0.33333333]]\n",
      "re: 0.1861\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.02830189]\n",
      " [0.64444444]\n",
      " [0.22516556]\n",
      " [0.38461538]]\n",
      "re: 0.3206\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2480\n",
      "===================================================================\n",
      "K=64, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.12264151]\n",
      " [0.53333333]\n",
      " [0.31125828]\n",
      " [0.1025641 ]]\n",
      "re: 0.2674\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.05660377]\n",
      " [0.38888889]\n",
      " [0.74834437]\n",
      " [0.05128205]]\n",
      "re: 0.3113\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.05660377]\n",
      " [0.61111111]\n",
      " [0.42384106]\n",
      " [0.12820513]]\n",
      "re: 0.3049\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.27358491]\n",
      " [0.54444444]\n",
      " [0.2384106 ]\n",
      " [0.        ]]\n",
      "re: 0.2641\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.02830189]\n",
      " [0.64444444]\n",
      " [0.22516556]\n",
      " [0.38461538]]\n",
      "re: 0.3206\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2937\n",
      "===================================================================\n",
      "K=64, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.00943396]\n",
      " [0.43333333]\n",
      " [0.1986755 ]\n",
      " [0.24358974]]\n",
      "re: 0.2213\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.03773585]\n",
      " [0.63333333]\n",
      " [0.27152318]\n",
      " [0.42307692]]\n",
      "re: 0.3414\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.03773585]\n",
      " [0.58888889]\n",
      " [0.21192053]\n",
      " [0.21794872]]\n",
      "re: 0.2641\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.01886792]\n",
      " [0.58888889]\n",
      " [0.13245033]\n",
      " [0.30769231]]\n",
      "re: 0.2620\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.01886792]\n",
      " [0.55555556]\n",
      " [0.2781457 ]\n",
      " [0.37179487]]\n",
      "re: 0.3061\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2790\n",
      "===================================================================\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "706d203aa21946cdadfc5a4dfe14d606",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "K=64, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.00943396]\n",
      " [0.48888889]\n",
      " [0.21192053]\n",
      " [0.05128205]]\n",
      "re: 0.1904\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.63333333]\n",
      " [0.0794702 ]\n",
      " [0.38461538]]\n",
      "re: 0.2767\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.38679245]\n",
      " [0.55555556]\n",
      " [0.14569536]\n",
      " [0.37179487]]\n",
      "re: 0.3650\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.14150943]\n",
      " [1.        ]\n",
      " [0.30463576]\n",
      " [0.30769231]]\n",
      "re: 0.4385\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.0754717 ]\n",
      " [0.16666667]\n",
      " [0.8013245 ]\n",
      " [0.05128205]]\n",
      "re: 0.2737\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3088\n",
      "===================================================================\n",
      "K=64, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.01886792]\n",
      " [0.03333333]\n",
      " [0.17880795]\n",
      " [0.24358974]]\n",
      "re: 0.1186\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.63333333]\n",
      " [0.0794702 ]\n",
      " [0.12820513]]\n",
      "re: 0.2126\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.11320755]\n",
      " [0.55555556]\n",
      " [0.14569536]\n",
      " [0.37179487]]\n",
      "re: 0.2966\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.14150943]\n",
      " [1.        ]\n",
      " [0.30463576]\n",
      " [0.33333333]]\n",
      "re: 0.4449\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.0754717 ]\n",
      " [0.1       ]\n",
      " [0.04635762]\n",
      " [0.33333333]]\n",
      "re: 0.1388\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2423\n",
      "===================================================================\n",
      "K=64, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.02830189]\n",
      " [0.5       ]\n",
      " [0.1192053 ]\n",
      " [0.03846154]]\n",
      "re: 0.1715\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.01886792]\n",
      " [0.63333333]\n",
      " [0.0794702 ]\n",
      " [0.11538462]]\n",
      "re: 0.2118\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.23584906]\n",
      " [0.54444444]\n",
      " [0.40397351]\n",
      " [0.41025641]]\n",
      "re: 0.3986\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.03773585]\n",
      " [0.61111111]\n",
      " [0.1589404 ]\n",
      " [0.20512821]]\n",
      "re: 0.2532\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.0754717 ]\n",
      " [0.04444444]\n",
      " [0.02649007]\n",
      " [0.35897436]]\n",
      "re: 0.1263\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2323\n",
      "===================================================================\n",
      "K=64, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.02830189]\n",
      " [0.32222222]\n",
      " [0.09933775]\n",
      " [0.38461538]]\n",
      "re: 0.2086\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.5       ]\n",
      " [0.2384106 ]\n",
      " [0.35897436]]\n",
      "re: 0.2767\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.02830189]\n",
      " [0.52222222]\n",
      " [0.0397351 ]\n",
      " [0.03846154]]\n",
      "re: 0.1572\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.0754717 ]\n",
      " [0.64444444]\n",
      " [0.1589404 ]\n",
      " [0.3974359 ]]\n",
      "re: 0.3191\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.0754717 ]\n",
      " [0.66666667]\n",
      " [0.32450331]\n",
      " [0.20512821]]\n",
      "re: 0.3179\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2559\n",
      "===================================================================\n",
      "K=64, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.03773585]\n",
      " [0.03333333]\n",
      " [0.25165563]\n",
      " [0.17948718]]\n",
      "re: 0.1256\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.13207547]\n",
      " [0.5       ]\n",
      " [0.29139073]\n",
      " [0.6025641 ]]\n",
      "re: 0.3815\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.16037736]\n",
      " [0.52222222]\n",
      " [0.22516556]\n",
      " [0.28205128]]\n",
      "re: 0.2975\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [0.44444444]\n",
      " [0.17218543]\n",
      " [0.58974359]]\n",
      "re: 0.3040\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.16037736]\n",
      " [0.5       ]\n",
      " [0.30463576]\n",
      " [0.30769231]]\n",
      "re: 0.3182\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2853\n",
      "===================================================================\n",
      "K=64, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.03773585]\n",
      " [0.03333333]\n",
      " [0.25827815]\n",
      " [0.17948718]]\n",
      "re: 0.1272\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.13207547]\n",
      " [0.13333333]\n",
      " [0.01986755]\n",
      " [0.61538462]]\n",
      "re: 0.2252\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.16037736]\n",
      " [0.51111111]\n",
      " [0.2384106 ]\n",
      " [0.42307692]]\n",
      "re: 0.3332\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.05660377]\n",
      " [0.58888889]\n",
      " [0.0397351 ]\n",
      " [0.03846154]]\n",
      "re: 0.1809\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.16037736]\n",
      " [0.5       ]\n",
      " [0.56291391]\n",
      " [0.02564103]]\n",
      "re: 0.3122\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2358\n",
      "===================================================================\n",
      "K=64, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.03773585]\n",
      " [0.03333333]\n",
      " [0.25827815]\n",
      " [0.35897436]]\n",
      "re: 0.1721\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.13207547]\n",
      " [0.5       ]\n",
      " [0.06622517]\n",
      " [0.1025641 ]]\n",
      "re: 0.2002\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.16037736]\n",
      " [0.52222222]\n",
      " [0.22516556]\n",
      " [0.03846154]]\n",
      "re: 0.2366\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.05660377]\n",
      " [0.58888889]\n",
      " [0.0397351 ]\n",
      " [0.78205128]]\n",
      "re: 0.3668\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.16037736]\n",
      " [0.51111111]\n",
      " [0.14569536]\n",
      " [0.3974359 ]]\n",
      "re: 0.3037\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2559\n",
      "===================================================================\n",
      "K=64, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.03773585]\n",
      " [0.46666667]\n",
      " [0.1589404 ]\n",
      " [0.57692308]]\n",
      "re: 0.3101\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.13207547]\n",
      " [0.57777778]\n",
      " [0.14569536]\n",
      " [0.02564103]]\n",
      "re: 0.2203\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.16037736]\n",
      " [0.51111111]\n",
      " [0.49668874]\n",
      " [0.28205128]]\n",
      "re: 0.3626\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.03773585]\n",
      " [0.57777778]\n",
      " [0.17218543]\n",
      " [0.43589744]]\n",
      "re: 0.3059\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.16037736]\n",
      " [0.5       ]\n",
      " [0.10596026]\n",
      " [0.26923077]]\n",
      "re: 0.2589\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2915\n",
      "===================================================================\n",
      "K=64, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[1.       ]\n",
      " [1.       ]\n",
      " [1.       ]\n",
      " [0.3974359]]\n",
      "re: 0.8494\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.25471698]\n",
      " [1.        ]\n",
      " [1.        ]\n",
      " [1.        ]]\n",
      "re: 0.8137\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.9326\n",
      "===================================================================\n",
      "K=64, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [1.        ]\n",
      " [0.01282051]]\n",
      "re: 0.7532\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.89622642]\n",
      " [1.        ]\n",
      " [1.        ]\n",
      " [1.        ]]\n",
      "re: 0.9741\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.9455\n",
      "===================================================================\n",
      "K=64, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.001\n",
      "seed:0\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[1.        ]\n",
      " [1.        ]\n",
      " [1.        ]\n",
      " [0.33333333]]\n",
      "re: 0.8333\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.19811321]\n",
      " [1.        ]\n",
      " [1.        ]\n",
      " [1.        ]]\n",
      "re: 0.7995\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.9266\n",
      "===================================================================\n",
      "K=64, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [1.        ]\n",
      " [0.23076923]]\n",
      "re: 0.8077\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [1.        ]\n",
      " [1.        ]\n",
      " [1.        ]]\n",
      "re: 0.7524\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.9120\n",
      "===================================================================\n",
      "K=64, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "re mean: 1.0000\n",
      "===================================================================\n",
      "K=64, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "re mean: 1.0000\n",
      "===================================================================\n",
      "K=64, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "re mean: 1.0000\n",
      "===================================================================\n",
      "K=64, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "re mean: 1.0000\n",
      "===================================================================\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8b064c2bcc1947cbb80784e1e9e5c2a7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "36410bd569044f0286726a868791284b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "K=128, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.        ]\n",
      " [0.64444444]\n",
      " [0.89403974]\n",
      " [0.07692308]]\n",
      "re: 0.4039\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.11320755]\n",
      " [0.63333333]\n",
      " [0.05298013]\n",
      " [0.43589744]]\n",
      "re: 0.3089\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.16037736]\n",
      " [0.63333333]\n",
      " [0.21192053]\n",
      " [0.06410256]]\n",
      "re: 0.2674\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [0.54444444]\n",
      " [0.30463576]\n",
      " [0.43589744]]\n",
      "re: 0.5712\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.6       ]\n",
      " [0.11258278]\n",
      " [0.38461538]]\n",
      "re: 0.2743\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3651\n",
      "===================================================================\n",
      "K=128, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.        ]\n",
      " [0.03333333]\n",
      " [0.35099338]\n",
      " [0.41025641]]\n",
      "re: 0.1986\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.11320755]\n",
      " [0.63333333]\n",
      " [0.05298013]\n",
      " [0.43589744]]\n",
      "re: 0.3089\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.16037736]\n",
      " [0.63333333]\n",
      " [0.21192053]\n",
      " [0.06410256]]\n",
      "re: 0.2674\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [0.8       ]\n",
      " [0.20529801]\n",
      " [0.43589744]]\n",
      "re: 0.6103\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.6       ]\n",
      " [0.11258278]\n",
      " [0.38461538]]\n",
      "re: 0.2743\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3319\n",
      "===================================================================\n",
      "K=128, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.        ]\n",
      " [0.34444444]\n",
      " [0.35099338]\n",
      " [0.41025641]]\n",
      "re: 0.2764\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.11320755]\n",
      " [0.63333333]\n",
      " [0.05298013]\n",
      " [0.43589744]]\n",
      "re: 0.3089\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.16037736]\n",
      " [0.63333333]\n",
      " [0.21192053]\n",
      " [0.06410256]]\n",
      "re: 0.2674\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [0.83333333]\n",
      " [0.20529801]\n",
      " [0.25641026]]\n",
      "re: 0.5738\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.6       ]\n",
      " [0.1192053 ]\n",
      " [0.34615385]]\n",
      "re: 0.2663\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3386\n",
      "===================================================================\n",
      "K=128, lr=0.001, num_layers=1, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.00943396]\n",
      " [0.63333333]\n",
      " [0.15231788]\n",
      " [0.24358974]]\n",
      "re: 0.2597\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.13207547]\n",
      " [0.04444444]\n",
      " [0.20529801]\n",
      " [0.25641026]]\n",
      "re: 0.1596\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.16981132]\n",
      " [0.31111111]\n",
      " [0.07284768]\n",
      " [0.41025641]]\n",
      "re: 0.2410\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [0.2       ]\n",
      " [0.25827815]\n",
      " [1.        ]]\n",
      "re: 0.6146\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.06603774]\n",
      " [0.58888889]\n",
      " [0.2781457 ]\n",
      " [1.        ]]\n",
      "re: 0.4833\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3516\n",
      "===================================================================\n",
      "K=128, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.0754717 ]\n",
      " [0.36666667]\n",
      " [0.50331126]\n",
      " [0.35897436]]\n",
      "re: 0.3261\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.58888889]\n",
      " [0.1589404 ]\n",
      " [0.33333333]]\n",
      "re: 0.2726\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.11320755]\n",
      " [1.        ]\n",
      " [0.0794702 ]\n",
      " [0.42307692]]\n",
      "re: 0.4039\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.03773585]\n",
      " [0.54444444]\n",
      " [0.09933775]\n",
      " [0.43589744]]\n",
      "re: 0.2794\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.62222222]\n",
      " [0.17218543]\n",
      " [0.30769231]]\n",
      "re: 0.2779\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3120\n",
      "===================================================================\n",
      "K=128, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.0754717 ]\n",
      " [0.36666667]\n",
      " [0.25165563]\n",
      " [0.43589744]]\n",
      "re: 0.2824\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.58888889]\n",
      " [0.1589404 ]\n",
      " [0.33333333]]\n",
      "re: 0.2726\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.11320755]\n",
      " [1.        ]\n",
      " [0.0794702 ]\n",
      " [0.5       ]]\n",
      "re: 0.4232\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.03773585]\n",
      " [0.54444444]\n",
      " [0.09933775]\n",
      " [0.8974359 ]]\n",
      "re: 0.3947\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.62222222]\n",
      " [0.17218543]\n",
      " [0.30769231]]\n",
      "re: 0.2779\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3302\n",
      "===================================================================\n",
      "K=128, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.0754717 ]\n",
      " [0.64444444]\n",
      " [0.13245033]\n",
      " [0.73076923]]\n",
      "re: 0.3958\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.17924528]\n",
      " [0.04444444]\n",
      " [0.0794702 ]\n",
      " [0.25641026]]\n",
      "re: 0.1399\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.11320755]\n",
      " [1.        ]\n",
      " [0.09933775]\n",
      " [0.78205128]]\n",
      "re: 0.4986\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.03773585]\n",
      " [0.54444444]\n",
      " [0.09271523]\n",
      " [0.8974359 ]]\n",
      "re: 0.3931\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.62222222]\n",
      " [0.17880795]\n",
      " [0.30769231]]\n",
      "re: 0.2795\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3414\n",
      "===================================================================\n",
      "K=128, lr=0.001, num_layers=1, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.02830189]\n",
      " [0.64444444]\n",
      " [0.25165563]\n",
      " [0.23076923]]\n",
      "re: 0.2888\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.64444444]\n",
      " [0.27152318]\n",
      " [0.43589744]]\n",
      "re: 0.3403\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.11320755]\n",
      " [0.63333333]\n",
      " [0.18543046]\n",
      " [0.43589744]]\n",
      "re: 0.3420\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.04716981]\n",
      " [0.62222222]\n",
      " [0.29139073]\n",
      " [0.07692308]]\n",
      "re: 0.2594\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.62222222]\n",
      " [0.2781457 ]\n",
      " [0.35897436]]\n",
      "re: 0.3172\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3095\n",
      "===================================================================\n",
      "K=128, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.01886792]\n",
      " [0.76666667]\n",
      " [0.2781457 ]\n",
      " [0.21794872]]\n",
      "re: 0.3204\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.0754717 ]\n",
      " [0.28888889]\n",
      " [0.34437086]\n",
      " [0.35897436]]\n",
      "re: 0.2669\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.12264151]\n",
      " [0.43333333]\n",
      " [0.07284768]\n",
      " [0.30769231]]\n",
      "re: 0.2341\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [0.27777778]\n",
      " [0.21854305]\n",
      " [0.37179487]]\n",
      "re: 0.4670\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.63333333]\n",
      " [0.18543046]\n",
      " [0.53846154]]\n",
      "re: 0.3393\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3256\n",
      "===================================================================\n",
      "K=128, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.01886792]\n",
      " [0.8       ]\n",
      " [0.25827815]\n",
      " [0.32051282]]\n",
      "re: 0.3494\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.0754717 ]\n",
      " [0.07777778]\n",
      " [0.2384106 ]\n",
      " [0.05128205]]\n",
      "re: 0.1107\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.12264151]\n",
      " [0.43333333]\n",
      " [0.07284768]\n",
      " [0.30769231]]\n",
      "re: 0.2341\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [0.31111111]\n",
      " [0.2384106 ]\n",
      " [0.42307692]]\n",
      "re: 0.4931\n",
      "------------------------------------------------------------------\n",
      "seed:4\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.        ]\n",
      " [0.63333333]\n",
      " [0.18543046]\n",
      " [0.30769231]]\n",
      "re: 0.2816\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2938\n",
      "===================================================================\n",
      "K=128, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.02830189]\n",
      " [0.71111111]\n",
      " [0.2781457 ]\n",
      " [0.41025641]]\n",
      "re: 0.3570\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.0754717 ]\n",
      " [0.35555556]\n",
      " [0.21192053]\n",
      " [0.43589744]]\n",
      "re: 0.2697\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.06603774]\n",
      " [1.        ]\n",
      " [0.27152318]\n",
      " [0.21794872]]\n",
      "re: 0.3889\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [0.52222222]\n",
      " [0.2384106 ]\n",
      " [0.24358974]]\n",
      "re: 0.5011\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.61111111]\n",
      " [0.2384106 ]\n",
      " [0.34615385]]\n",
      "re: 0.2989\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3631\n",
      "===================================================================\n",
      "K=128, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.00943396]\n",
      " [0.64444444]\n",
      " [0.25165563]\n",
      " [0.34615385]]\n",
      "re: 0.3129\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.10377358]\n",
      " [0.34444444]\n",
      " [0.18543046]\n",
      " [0.37179487]]\n",
      "re: 0.2514\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.0754717 ]\n",
      " [0.13333333]\n",
      " [0.27152318]\n",
      " [0.42307692]]\n",
      "re: 0.2259\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.        ]\n",
      " [0.71111111]\n",
      " [0.31788079]\n",
      " [0.47435897]]\n",
      "re: 0.6258\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.        ]\n",
      " [0.34444444]\n",
      " [0.16556291]\n",
      " [0.05128205]]\n",
      "re: 0.1403\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3113\n",
      "===================================================================\n",
      "K=128, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.44339623]\n",
      " [0.3       ]\n",
      " [0.24503311]\n",
      " [0.07692308]]\n",
      "re: 0.2663\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.52222222]\n",
      " [0.25827815]\n",
      " [0.11538462]]\n",
      "re: 0.2263\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.1509434 ]\n",
      " [0.17777778]\n",
      " [0.07284768]\n",
      " [0.26923077]]\n",
      "re: 0.1677\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [0.63333333]\n",
      " [0.12582781]\n",
      " [0.30769231]]\n",
      "re: 0.2691\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.61111111]\n",
      " [0.17880795]\n",
      " [0.20512821]]\n",
      "re: 0.2511\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2361\n",
      "===================================================================\n",
      "K=128, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.44339623]\n",
      " [0.3       ]\n",
      " [0.24503311]\n",
      " [0.3974359 ]]\n",
      "re: 0.3465\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.02222222]\n",
      " [0.06622517]\n",
      " [0.33333333]]\n",
      "re: 0.1078\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.1509434 ]\n",
      " [0.62222222]\n",
      " [0.2781457 ]\n",
      " [0.33333333]]\n",
      "re: 0.3462\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [0.63333333]\n",
      " [0.12582781]\n",
      " [0.41025641]]\n",
      "re: 0.2947\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.61111111]\n",
      " [0.17880795]\n",
      " [0.85897436]]\n",
      "re: 0.4146\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3019\n",
      "===================================================================\n",
      "K=128, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.44339623]\n",
      " [0.1       ]\n",
      " [0.09933775]\n",
      " [0.05128205]]\n",
      "re: 0.1735\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.00943396]\n",
      " [0.37777778]\n",
      " [0.01986755]\n",
      " [0.34615385]]\n",
      "re: 0.1883\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.14150943]\n",
      " [0.05555556]\n",
      " [0.07284768]\n",
      " [0.30769231]]\n",
      "re: 0.1444\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [0.63333333]\n",
      " [0.12582781]\n",
      " [0.01282051]]\n",
      "re: 0.1954\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.00943396]\n",
      " [0.62222222]\n",
      " [0.20529801]\n",
      " [0.42307692]]\n",
      "re: 0.3150\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2033\n",
      "===================================================================\n",
      "K=128, lr=0.001, num_layers=2, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.10377358]\n",
      " [0.31111111]\n",
      " [0.17218543]\n",
      " [0.35897436]]\n",
      "re: 0.2365\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.        ]\n",
      " [0.74444444]\n",
      " [0.1589404 ]\n",
      " [0.15384615]]\n",
      "re: 0.2643\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.08490566]\n",
      " [0.62222222]\n",
      " [0.25165563]\n",
      " [0.37179487]]\n",
      "re: 0.3326\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [0.45555556]\n",
      " [0.25165563]\n",
      " [0.        ]]\n",
      "re: 0.1792\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.01886792]\n",
      " [0.62222222]\n",
      " [0.2781457 ]\n",
      " [0.30769231]]\n",
      "re: 0.3067\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.2639\n",
      "===================================================================\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bf1e8e3ce7f247a5ac6fda9f893e76b6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "K=128, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.12264151]\n",
      " [0.9       ]\n",
      " [0.15231788]\n",
      " [1.        ]]\n",
      "re: 0.5437\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.01886792]\n",
      " [1.        ]\n",
      " [0.09933775]\n",
      " [1.        ]]\n",
      "re: 0.5296\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.10377358]\n",
      " [1.        ]\n",
      " [0.01986755]\n",
      " [1.        ]]\n",
      "re: 0.5309\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.03773585]\n",
      " [1.        ]\n",
      " [0.0794702 ]\n",
      " [0.23076923]]\n",
      "re: 0.3370\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.02830189]\n",
      " [1.        ]\n",
      " [0.02649007]\n",
      " [1.        ]]\n",
      "re: 0.5137\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.4910\n",
      "===================================================================\n",
      "K=128, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[0.12264151]\n",
      " [0.87777778]\n",
      " [0.15231788]\n",
      " [1.        ]]\n",
      "re: 0.5382\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.1509434 ]\n",
      " [1.        ]\n",
      " [0.13907285]\n",
      " [0.3974359 ]]\n",
      "re: 0.4219\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.09433962]\n",
      " [0.51111111]\n",
      " [0.06622517]\n",
      " [1.        ]]\n",
      "re: 0.4179\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.00943396]\n",
      " [0.44444444]\n",
      " [0.0397351 ]\n",
      " [1.        ]]\n",
      "re: 0.3734\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.02830189]\n",
      " [1.        ]\n",
      " [0.12582781]\n",
      " [1.        ]]\n",
      "re: 0.5385\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.4580\n",
      "===================================================================\n",
      "K=128, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.001\n",
      "seed:0\n",
      "[[0.05660377]\n",
      " [0.61111111]\n",
      " [0.0397351 ]\n",
      " [1.        ]]\n",
      "re: 0.4269\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.1509434]\n",
      " [1.       ]\n",
      " [0.0397351]\n",
      " [1.       ]]\n",
      "re: 0.5477\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.13207547]\n",
      " [1.        ]\n",
      " [0.06622517]\n",
      " [0.43589744]]\n",
      "re: 0.4085\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.02830189]\n",
      " [0.64444444]\n",
      " [0.09933775]\n",
      " [1.        ]]\n",
      "re: 0.4430\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.02830189]\n",
      " [1.        ]\n",
      " [0.04635762]\n",
      " [1.        ]]\n",
      "re: 0.5187\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.4690\n",
      "===================================================================\n",
      "K=128, lr=0.01, num_layers=1, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[0.05660377]\n",
      " [1.        ]\n",
      " [0.14569536]\n",
      " [1.        ]]\n",
      "re: 0.5506\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.12264151]\n",
      " [0.56666667]\n",
      " [0.10596026]\n",
      " [1.        ]]\n",
      "re: 0.4488\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.10377358]\n",
      " [0.92222222]\n",
      " [0.09271523]\n",
      " [1.        ]]\n",
      "re: 0.5297\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.03773585]\n",
      " [0.93333333]\n",
      " [0.15231788]\n",
      " [1.        ]]\n",
      "re: 0.5308\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.02830189]\n",
      " [1.        ]\n",
      " [0.16556291]\n",
      " [1.        ]]\n",
      "re: 0.5485\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.5217\n",
      "===================================================================\n",
      "K=128, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[0.0754717 ]\n",
      " [1.        ]\n",
      " [0.03311258]\n",
      " [0.15384615]]\n",
      "re: 0.3156\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.06603774]\n",
      " [1.        ]\n",
      " [0.14569536]\n",
      " [1.        ]]\n",
      "re: 0.5529\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.26415094]\n",
      " [1.        ]\n",
      " [0.09933775]\n",
      " [0.42307692]]\n",
      "re: 0.4466\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.18867925]\n",
      " [0.46666667]\n",
      " [0.12582781]\n",
      " [1.        ]]\n",
      "re: 0.4453\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.19811321]\n",
      " [0.58888889]\n",
      " [0.18543046]\n",
      " [1.        ]]\n",
      "re: 0.4931\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.4507\n",
      "===================================================================\n",
      "K=128, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[0.0754717 ]\n",
      " [1.        ]\n",
      " [0.0397351 ]\n",
      " [0.20512821]]\n",
      "re: 0.3301\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.08490566]\n",
      " [0.64444444]\n",
      " [0.02649007]\n",
      " [1.        ]]\n",
      "re: 0.4390\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.27358491]\n",
      " [0.12222222]\n",
      " [0.06622517]\n",
      " [0.05128205]]\n",
      "re: 0.1283\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.22641509]\n",
      " [0.55555556]\n",
      " [0.09933775]\n",
      " [1.        ]]\n",
      "re: 0.4703\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.19811321]\n",
      " [1.        ]\n",
      " [0.13245033]\n",
      " [1.        ]]\n",
      "re: 0.5826\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.3901\n",
      "===================================================================\n",
      "K=128, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[0.06603774]\n",
      " [1.        ]\n",
      " [0.0397351 ]\n",
      " [1.        ]]\n",
      "re: 0.5264\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.09433962]\n",
      " [0.64444444]\n",
      " [0.12582781]\n",
      " [0.02564103]]\n",
      "re: 0.2226\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.17924528]\n",
      " [1.        ]\n",
      " [0.09933775]\n",
      " [0.37179487]]\n",
      "re: 0.4126\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.05660377]\n",
      " [1.        ]\n",
      " [0.14569536]\n",
      " [0.1025641 ]]\n",
      "re: 0.3262\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.19811321]\n",
      " [1.        ]\n",
      " [0.01986755]\n",
      " [1.        ]]\n",
      "re: 0.5545\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.4085\n",
      "===================================================================\n",
      "K=128, lr=0.01, num_layers=1, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[0.00943396]\n",
      " [1.        ]\n",
      " [0.03311258]\n",
      " [0.42307692]]\n",
      "re: 0.3664\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[0.03773585]\n",
      " [0.07777778]\n",
      " [0.05960265]\n",
      " [1.        ]]\n",
      "re: 0.2938\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[0.16981132]\n",
      " [1.        ]\n",
      " [0.04635762]\n",
      " [1.        ]]\n",
      "re: 0.5540\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[0.18867925]\n",
      " [1.        ]\n",
      " [0.07284768]\n",
      " [0.61538462]]\n",
      "re: 0.4692\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[0.19811321]\n",
      " [1.        ]\n",
      " [0.13907285]\n",
      " [1.        ]]\n",
      "re: 0.5843\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.4536\n",
      "===================================================================\n",
      "K=128, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.0\n",
      "seed:0\n",
      "[[0.0754717]\n",
      " [1.       ]\n",
      " [1.       ]\n",
      " [1.       ]]\n",
      "re: 0.7689\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.9538\n",
      "===================================================================\n",
      "K=128, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.0001\n",
      "seed:0\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "re mean: 1.0000\n",
      "===================================================================\n",
      "K=128, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.001\n",
      "seed:0\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "re mean: 1.0000\n",
      "===================================================================\n",
      "K=128, lr=0.01, num_layers=2, hidden_dim=16, alpha=0.01\n",
      "seed:0\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "re mean: 1.0000\n",
      "===================================================================\n",
      "K=128, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.0\n",
      "seed:0\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [1.        ]\n",
      " [0.44871795]]\n",
      "re: 0.8622\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.9724\n",
      "===================================================================\n",
      "K=128, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.0001\n",
      "seed:0\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "re mean: 1.0000\n",
      "===================================================================\n",
      "K=128, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.001\n",
      "seed:0\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "re mean: 1.0000\n",
      "===================================================================\n",
      "K=128, lr=0.01, num_layers=2, hidden_dim=32, alpha=0.01\n",
      "seed:0\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:1\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:2\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:3\n",
      "[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "re: 1.0000\n",
      "------------------------------------------------------------------\n",
      "seed:4\n",
      "[[1.        ]\n",
      " [1.        ]\n",
      " [1.        ]\n",
      " [0.64102564]]\n",
      "re: 0.9103\n",
      "------------------------------------------------------------------\n",
      "re mean: 0.9821\n",
      "===================================================================\n",
      "optimal parameters: K=16, lr=0.001, num_layers=2, hidden_dim=16, alpha=0.001, result: 0.141478825782212\n"
     ]
    }
   ],
   "source": [
    "Rated_Capacity = 2.0\n",
    "feature_size = 16\n",
    "feature_num = 1\n",
    "dropout = 0.0\n",
    "epochs = 2000\n",
    "nhead = 1\n",
    "\n",
    "weight_decay = 0.0\n",
    "noise_level = 0.0\n",
    "alpha = 0.0\n",
    "metric = 're'\n",
    "\n",
    "states = {}\n",
    "for K in tqdm([1, 4, 8, 16, 32, 64, 128], position=0):\n",
    "    for lr in tqdm([0.001, 0.01], position=1):\n",
    "        for num_layers in tqdm([1, 2], position=2):\n",
    "            for hidden_dim in [16, 32]:\n",
    "                for alpha in [0.0, 1e-4, 1e-3, 1e-2]:\n",
    "                    show_str = 'K={}, lr={}, num_layers={}, hidden_dim={}, alpha={}'.format(K, lr, num_layers, hidden_dim, alpha)\n",
    "                    print(show_str)\n",
    "                    SCORE = []\n",
    "                    for seed in range(5):\n",
    "                        print('seed:{}'.format(seed))\n",
    "                        score_list, _, _ = train(lr=lr, feature_size=feature_size, feature_num=feature_num, hidden_dim=hidden_dim, \n",
    "                                                 num_layers=num_layers, nhead=nhead, weight_decay=weight_decay, epochs=epochs, \n",
    "                                                 seed=seed, dropout=dropout, alpha=alpha, noise_level=noise_level, metric=metric)\n",
    "                        print(np.array(score_list))\n",
    "                        print(metric + ': {:<6.4f}'.format(np.mean(np.array(score_list))))\n",
    "                        print('------------------------------------------------------------------')\n",
    "                        for s in score_list:\n",
    "                            SCORE.append(s)\n",
    "\n",
    "                    print(metric + ' mean: {:<6.4f}'.format(np.mean(np.array(SCORE))))\n",
    "                    states[show_str] = np.mean(np.array(SCORE))\n",
    "                    print('===================================================================')\n",
    "\n",
    "min_key = min(states, key = states.get)\n",
    "print('optimal parameters: {}, result: {}'.format(min_key, states[min_key]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "torch_11",
   "language": "python",
   "name": "torch_11"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}