{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Demo: Fitting a sine curve"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"scrolled": true
},
"outputs": [],
"source": [
"%load_ext autoreload\n",
"%autoreload 2\n",
"import warnings\n",
"warnings.filterwarnings(\"ignore\", category=UserWarning, module=\"multiprocessing.resource_tracker\")\n",
"\n",
"import numpy as np\n",
"import matplotlib.pyplot as plt\n",
"\n",
"import nbi"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"0.4.1\n"
]
}
],
"source": [
"print(nbi.__version__)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"This notebook provides a minimal example that could be used as a template to adapt to your specific tasks.\n",
"\n",
"Let us consider a very simple sine curve model, where $t$ is time, $\\omega$ is angular frequency, $A$ is amplitude, and $\\phi_0$ is initial phase:
\n",
"$f(t) = A\\cdot\\sin(\\omega\\cdot t+\\phi_0$)\n",
"\n",
"Assume parameters from the following uniform distribution:
\n",
"$\\phi_0\\in[0,2\\pi]$, $A\\in[-4,4]$, $\\omega\\in[2\\pi,12\\pi]$\n",
"\n",
"Note that the amplitude is allowed to be negative. Thus expect bi-modal posteriors in the amplitude.\n",
"\n",
"Also assume Gaussian measurement noise fixed to be $\\sigma_0$:
\n",
"$x_{\\rm obs}\\sim N(f(t), \\sigma=\\sigma_0)$"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
"t = np.linspace(0,1,50)\n",
"def sine(param):\n",
" phi0, A, omega = param\n",
" return np.sin(omega * t + phi0) * A"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
"# nbi requires prior to be defined with scipy functions\n",
"# alternatively, you may supply pre-generated parameters with numpy arrays\n",
"\n",
"from scipy.stats import uniform, truncnorm\n",
"prior = {\n",
" 'phi0': uniform(loc=0, scale=np.pi*2),\n",
" 'A': uniform(loc=-4, scale=8),\n",
" 'omega': uniform(loc=2*np.pi, scale=10*np.pi)\n",
"}\n",
"labels = list(prior.keys())\n",
"priors = [prior[k] for k in labels]"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAbUAAAESCAYAAAB6s0uLAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAAAa40lEQVR4nO3df2xV5eHH8c9ttRf42l4kFGhHFQthTmXa4WTAFFjIQA3ObFETFwLG4aZlyajZVsakU6cljphljOHGpphv3HDbV3QRokMUDQ7jV6SJyo+lA0YFy2A/7q0gLaXP9w+/vVLovbf39j7nx3Per+Qm9va09+Hx9HyeX+c5MWOMEQAADijxuwAAABQLoQYAcAahBgBwBqEGAHAGoQYAcAahBgBwBqEGAHDGeX4XIJuenh4dPnxY5eXlisVifhcHAOATY4w6OjpUXV2tkpLM/bFAh9rhw4dVU1PjdzEAAAHR1tamsWPHZvx+oEOtvLxc0sf/iIqKCp9LAwDwSyqVUk1NTToXMgl0qPUOOVZUVBBqAICcU1EsFAEAOINQAwA4g1ADADiDUAMAOINQAwA4g1ADADiDUAMAOINQAwA4g1ADCnSiq1vjGjdqXONGnejq9rs4AESoAQAcQqgBAJxBqAEAnEGoAQCcQagBAJxhNdSam5v1+c9/XuXl5Ro1apRuvvlm7d271+ZHAr5oT570uwgAZDnUXn31VdXX1+uNN97Q5s2bderUKX35y1/W8ePHbX4s4In/2fF++r9nP/qqnv7fgz6WBoAkxYwxxqsPO3r0qEaNGqVXX31V1113Xc7jU6mUEomEkskkDwlFoHyQ/EjTV7ysnjP+ekpjMW1rnKWqxFD/CgY4aqB54OmTr5PJpCRpxIgR/X6/s7NTnZ2d6a9TqZQn5QLytf/Y8T6BJkmnjdGBYycINcBHni0U6enp0Xe+8x1Nnz5dV1xxRb/HNDc3K5FIpF81NTVeFc9p7HxRfJeM/C+VnPVU+dJYTONGDvOnQAAkeRhq9fX1evfdd7V+/fqMxyxdulTJZDL9amtr86p4QF6qEkN1/02Xp78uiUkPf/UKemmAzzwJtcWLF+v555/XK6+8orFjx2Y8Lh6Pq6Kios8LudET88fXJn9yLr/UMEO3ff4iH0sDQLI8p2aM0be//W1t2LBBW7du1SWXXGLz4wDfjEkM8bsIAGQ51Orr6/Xb3/5Wzz33nMrLy9Xe3i5JSiQSGjqUYRoAQHFZHX5cs2aNksmkZs6cqaqqqvTr6aeftvmxAICIsj78iGBpT55UbeUFfhcDAKxg78cIYOcLAFHh6c3XsO/sntgHyY/U9Kf30l/3GOkHz7yr6yZWsvx8kIaVnacDK270uxjOOdHVrcuWvyhJ2vXAHA0r4zKFgaOn5oBsPbFsO18AgGsItZDL1BP7IPmRJHa+ABAthFrI5eqJsfMFgCgh1EJuID0xdr4oHLu12EG9Bo8r/08ItZDLtyfGzhdANLkSWrkQag6gJwZX8UTx/EQluLIh1BxDTwxhx32VGAxCDUBg5FrNC+RCqAEIDO6rxGARaiHBWLn/mN+x48x65b5KDBahFgG92zkdWHEjWw7lifkdOzLVK/dVeiNXAy3MDThCDciA+R07ctUrq3mL4+xgytVAc6UBR6g5gJ6YHczv2JFPvbKaNz+ZgilXQ8KlBhyhBmQwmPkd5kAzY97MjmzBlKsh4VIDjlADMmB+xw7q1Y5swZSrIeFSQ4NQC6EwT+KGDfM7dhRar/SAM8sWTLkaEi41NAi1kHBlEjfMmN+xg3otjlzBlKsh4UoDjlALAZuTuLR8ETQsfCrcQIMpV0MizA0NQi0EXJrEBeCNMAfTYBBqIeDSJG4UMQcKeIdQCwGXJnGjgjlQwB8MWIfE1yaP1X3PfTyv9lLDDNVWXuBziaKhd34nH5nmQK+bWElD5P8VUq/AQBBqIRTVsfKwyDYHSqgVT3vyJI27PORqSLjS0GD4EWnM/RQHc6D2MKybHStHCbXI4yJRfMyB2uHS/oSwh1CLMC4S9rhyI2uQcGsLBoJQizAuEt5gDrQ4GNbFQBBqIWFjrJyLBMKEYV0MBKEWYVwkEDbZhnXZ8g0SoRZ5zP0grBjWRX8INaRF8SJB6x5wC6EWIFxgAWBwonl3HmCZK7szAGFjtaf22muvad68eaqurlYsFtOzzz5r8+MAABFnNdSOHz+uK6+8UqtXr7b5MQAiYqC3trDlW3RZHX68/vrrdf3119v8CACQdO6Wb81fncRq3ggK1JxaZ2enOjs701+nUikfSxMN2eZ+TnR167LlL0qSdj0wJ7IbpCL4eNwPegVq9WNzc7MSiUT6VVNT43eRfMPwifeo8+LycjUvW76hV6BCbenSpUomk+lXW1ub30XyFDvme486dwNbvqFXoEItHo+roqKizysq2DHfe9S5O9jyDb0CFWpRxvCJ96hzt7DlGyTLC0U+/PBDtba2pr/ev3+/WlpaNGLECF10ESfcmXqHT868yDJ8Ypdfdc4CHPtc3vKN8yc7qz21t956S3V1daqrq5MkNTQ0qK6uTsuXL7f5saHE8In3qHNgYMK0hZ/VUJs5c6aMMee81q1bZ/NjQyvIwyeurgwMcp0DyB9zagEVhOGTqK0MDEKdu8rVRhGCh1BDv1gZiMGKWqMIwUCooV+sDMRg0CiCX1g2g36xGtMb7cmTqq28wO9iFF22RpGthTg87gcSPTVkwMpAe6IwLMcOH97wY64y6POjhFqADPSxGl6JwspAr+s8KsNyNIrs8aNRFKaGGKGGAWFlYHFEaa4yCo0ir/nRKApbQ4xQAzwU1WE5GkXF4UejKGwNMUIN8BDDchgMPxpFYWuIEWqAxxiWQ6H8aBSFrSFGqAE+YlgO+fKjURSmhpj/S+wQWNz3g1zYMd5ffjSKgt4Q4wwEYAWNIviB4UcAgDMINQCAMwg1AIAzmFMDPMZcEwaD8yc7Qg1AUbj6xAGEK0gZfgRQsDBtdItoINQAFCRsG90iGgg1AAUJ20a3iAZCDUBBwrbRLaKBUANQkLBtdItoINQAFCxMG90iGgg1IEBOdHVrXONGjWvcqBNd3X4XJy9B3+gW0UCoAQCcQagBAJxBqAEACha0IXNCDQDgDPZ+BFCwMO0JiGigpwYEVHvypN9FAEKHUAMChA2CgcEh1ICAYINgYPAINTgvaKuzMmGDYLvCch6EWRCGzAk1ICDYIBhhFLQhc0INCAg2CEbYBHHI3JNQW716tcaNG6chQ4ZoypQpevPNN734WCB02CAYYRLEIXProfb000+roaFBTU1Nevvtt3XllVdqzpw5+sc//mH7o4FQY4Pg6ArL/F8Qh8yth9qjjz6qRYsW6Y477tBll12mxx57TMOGDdPjjz9u+6MBABYFccjcaqh1dXVpx44dmj179icfWFKi2bNna/v27ecc39nZqVQq1ecFFFMQVmfBf5wHxRO0IXOroXbs2DGdPn1ao0eP7vP+6NGj1d7efs7xzc3NSiQS6VdNTY3N4iEigrY6C/7gPLAvCEPmgVr9uHTpUiWTyfSrra3N7yIh5IK4Ogve4zyIDqsbGo8cOVKlpaU6cuRIn/ePHDmiMWPGnHN8PB5XPB63WSRETLbVWSyVjw7Og+iw2lMrKyvT5MmTtWXLlvR7PT092rJli6ZOnWrzowFJwVydlU3vrvcHVtyoYWU8RKNYwnYenIn5v/xYH35saGjQ2rVr9eSTT2r37t26++67dfz4cd1xxx22PxoI5OoseC9s5wHzf4WLGWNM7sMG5+c//7l+8pOfqL29XVdddZV+9rOfacqUKTl/LpVKKZFIKJlMqqKiwnYx4agTXd26bPmLkqSX752h2soLfC4R/BCW8+CD5EeavuLlPsOlpbGYtjXOCmwIe2GgeeDJ+MbixYu1ePFiLz4KyCoIq7PgvyCfB8z/DU6gVj8CQNSFef4vCAg1AAiQsM3/BQ2hBgABE7RdOsKEUENBwrLhKhB2QZ7/CyJuhAEQCb33AMJthBqcx8UMiA6GH+EEhkMBSIQagCxoLCBsCDUAgDOYU/PYmVv17HpgjhOb1rYnTwZ2yyEgjJgHLhw9NRSEDVcBBBGhhrzxwEUAQUWoIW/ZNlwNAp4/BUQXoYa8BXHDVYZD7aOxgDAg1JC3oG24ynCoPTQWEDaEmo/C3PIN0oarQR8OLSYv7xujsYAwItQ85mLL1+8NV4M4HOqCKDUW4A5CzUO0fO0I2nCoK2gsIIwINQ/R8rUnSMOhrqCxgMHyY5s1Qs1DtHy94fdwqEtoLCBsCDUP0fJFmNFYQBiEf+PBkPna5LG677mP59VeapgR2j0T2ZvOf+y5CZyLnpqPaPkiXy6ungWKiVADQoLVs0BuDD/CCVEYDs22epZ5WeBj9NSAkGD1LMLMqx2UCDUgJPxYPdvbAz6w4kYnHmgLb/kxBxwzxpjch/kjlUopkUgomUyqoqLC7+IAvjvzyekv3xve1bNw3wfJjzR9xct9hsxLYzFta5xVUENsoHlAT80CP+6iR/SwejbcXL9O+LWDkvOhlu3Ecf2kAjAwXAuKz685YOdDDQDgPb92UCLUAABW+LF3KKEGALDOqzlg1uii6M5cobfrgTksBS+iKNxkDgwGPTXLvLrhEEB4cZ0oHmuh9tBDD2natGkaNmyYhg8fbutj8uLVicOms0B4cZ0IN2uh1tXVpVtuuUV33323rY8YkIGeOMU6kdl0FggfrwOG64Q91kLt/vvv15IlSzRp0iRbH5FTrhPHxons1w2HAArjR8BE5TrhxzZrgZpT6+zsVCqV6vMajGwnjq0TmU1ngXDxI2C4TtgTqFBrbm5WIpFIv2pqagb1+7KdOLZOZL9uOAwqJsARdH4EDNcJe/IKtcbGRsVisayvPXv2FFyYpUuXKplMpl9tbW0F/y4p+4lj80T244bDIGECHGHiV8BE/TphS16DnPfee68WLlyY9Zja2tqCCxOPxxWPxwv++f58bfJY3ffcx8OMLzV8sqt574nc+z1bJ3LUNp3NNKx73cRKWqEIrEzXCa9E7TphU16hVllZqcrKSltlse7sE6fQE5mbizPj6cwIu7OvE/y9h4u1/zsHDx7Uv/71Lx08eFCnT59WS0uLJGnChAm64ILgPQOKllJx9A7rnv0MJSbAAXjB2kKR5cuXq66uTk1NTfrwww9VV1enuro6vfXWW7Y+EgHABDgAP1kLtXXr1skYc85r5syZtj4SAWFrApxnXgHIhcFhC9h09hMM6wL94zphh/OhxokDIBeuE+5wPtSyKcaJ3J486fnyXwD+4O89+AK1o0hYcHMxEB38vYcLoZYndtdGULGQpvj4ew8fQi1PUdldO+jYUxJe4O89fAi1PLG7dm62HjfBMBC8Npi/d3rO/iDU8sTNxf5gGAh+4O89fAi1ArC7tvcYBoJf+HsPF0JtkLi52BsM+yII+HsPPkINocAwUH5YSIOoItQQGgwDZcdCGoBQQ0gxDNQXC2mCjZ6zdwg1BArLoAvDQprgoefsj0jv/VgoNj9F0Azm4aw82bn4MvWcr5tYyTywZfTUAAewkMaeQjYToOfsH0INcES2hTQM63qLW1D8Q6ghNGxtv+UiFtL4i56zfwg1ALCAW1D8QaghsFgGDVfQc/YOoYZAYRm0v2hIIOwINQQGNxD7g4YEXEKowVPZVuGxDNobZ/bGaEjANYRaBiyB9h7LoAcn2+rQTL0xGhJwDaGGwGAZtB3ZemM0JOAaQg2BwjLo4svWG6MhYQ/3VfqDUENgsQy6OHL1xmhIwCWEGuC4fHpjNCQQdoQafMM9Ud6hN4aoINQGgItv8XBPlP/ojcFlhFoGXHyLj3uiANjGkpx+8IA/O3KtwpN4ACuAwaGn1g9uSLWDe6IA2Eao9YOLrx3cExVM3E8FlxBq/eDiaw+r8PxBcCEqCLUMuPjaxyo8hB17xAaPtVA7cOCA7rzzTl1yySUaOnSoxo8fr6amJnV1ddn6SGu4+AJAOFgbh9izZ496enr0y1/+UhMmTNC7776rRYsW6fjx41q5cqWtjwUARJi1UJs7d67mzp2b/rq2tlZ79+7VmjVrCDUAgBWezhgnk0mNGDEi4/c7OzvV2dmZ/jqVSnlRLACAIzwLtdbWVq1atSprL625uVn333+/V0WCD7i5GoBNeS8UaWxsVCwWy/ras2dPn585dOiQ5s6dq1tuuUWLFi3K+LuXLl2qZDKZfrW1teX/LwIAH7BHbDDEjDEm92GfOHr0qP75z39mPaa2tlZlZWWSpMOHD2vmzJn6whe+oHXr1qmkZOA5mkqllEgklEwmVVFRkU8xAcC6/95+QPc99/GWeiUxqfmrk7j9x5KB5kHew4+VlZWqrKwc0LGHDh3SrFmzNHnyZD3xxBN5BRoABBl7xAaTtTm1Q4cOaebMmbr44ou1cuVKHT16NP29MWPG2PpYAPDEQDbohveshdrmzZvV2tqq1tZWjR07ts/38hzxBIDA6d0j9sxgY49Y/1kbD1y4cKGMMf2+ACDs2CM2mJjkAoACsUds8BBqAFAE7BEbDIQaAMAZhBoAwBmEGgDAGYQaAMAZhBoAwBmePnoGAFzCUyeCh54aAMAZhBoAwBmEGgDAGYQaAMAZhBoAwBmEGgDAGYQaAMAZhBoAwBmBvvm694GiqVTK55IAAPzUmwO5HjQd6FDr6OiQJNXU1PhcEgBAEHR0dCiRSGT8fszkij0f9fT06PDhwyovL1csFiv496RSKdXU1KitrU0VFRVFLGG4US+ZUTf9o14yo276V6x6Mcaoo6ND1dXVKinJPHMW6J5aSUmJxo4dm/vAAaqoqOBk6wf1khl10z/qJTPqpn/FqJdsPbReLBQBADiDUAMAOCMSoRaPx9XU1KR4PO53UQKFesmMuukf9ZIZddM/r+sl0AtFAADIRyR6agCAaCDUAADOINQAAM4g1AAAziDUAADOcCbUVq9erXHjxmnIkCGaMmWK3nzzzazH/+EPf9Cll16qIUOGaNKkSdq0aZNHJfVWPvWydu1aXXvttbrwwgt14YUXavbs2TnrMczyPWd6rV+/XrFYTDfffLPdAvok33r5z3/+o/r6elVVVSkej2vixIn8Pf2/n/70p/r0pz+toUOHqqamRkuWLNHJkyc9Kq03XnvtNc2bN0/V1dWKxWJ69tlnc/7M1q1b9bnPfU7xeFwTJkzQunXrilcg44D169ebsrIy8/jjj5v33nvPLFq0yAwfPtwcOXKk3+Nff/11U1paah555BGza9cu88Mf/tCcf/755p133vG45HblWy+33367Wb16tdm5c6fZvXu3WbhwoUkkEub999/3uOT25Vs3vfbv328+9alPmWuvvdZ85Stf8aawHsq3Xjo7O83VV19tbrjhBrNt2zazf/9+s3XrVtPS0uJxye3Lt26eeuopE4/HzVNPPWX2799vXnzxRVNVVWWWLFniccnt2rRpk1m2bJl55plnjCSzYcOGrMfv27fPDBs2zDQ0NJhdu3aZVatWmdLSUvPCCy8UpTxOhNo111xj6uvr01+fPn3aVFdXm+bm5n6Pv/XWW82NN97Y570pU6aYb37zm1bL6bV86+Vs3d3dpry83Dz55JO2iuibQuqmu7vbTJs2zfz61782CxYscDLU8q2XNWvWmNraWtPV1eVVEX2Tb93U19ebL33pS33ea2hoMNOnT7daTj8NJNS+973vmcsvv7zPe7fddpuZM2dOUcoQ+uHHrq4u7dixQ7Nnz06/V1JSotmzZ2v79u39/sz27dv7HC9Jc+bMyXh8GBVSL2c7ceKETp06pREjRtgqpi8KrZsHHnhAo0aN0p133ulFMT1XSL386U9/0tSpU1VfX6/Ro0friiuu0MMPP6zTp097VWxPFFI306ZN044dO9JDlPv27dOmTZt0ww03eFLmoLJ9/Q30Lv0DcezYMZ0+fVqjR4/u8/7o0aO1Z8+efn+mvb293+Pb29utldNrhdTL2b7//e+rurr6nBMw7Aqpm23btuk3v/mNWlpaPCihPwqpl3379unll1/W17/+dW3atEmtra265557dOrUKTU1NXlRbE8UUje33367jh07pi9+8Ysyxqi7u1vf+ta39IMf/MCLIgdWputvKpXSRx99pKFDhw7q94e+pwY7VqxYofXr12vDhg0aMmSI38XxVUdHh+bPn6+1a9dq5MiRfhcnUHp6ejRq1Cj96le/0uTJk3Xbbbdp2bJleuyxx/wumu+2bt2qhx9+WL/4xS/09ttv65lnntHGjRv14IMP+l00p4W+pzZy5EiVlpbqyJEjfd4/cuSIxowZ0+/PjBkzJq/jw6iQeum1cuVKrVixQi+99JI++9nP2iymL/Ktm7/97W86cOCA5s2bl36vp6dHknTeeedp7969Gj9+vN1Ce6CQc6aqqkrnn3++SktL0+995jOfUXt7u7q6ulRWVma1zF4ppG7uu+8+zZ8/X9/4xjckSZMmTdLx48d11113admyZVkfdOmyTNffioqKQffSJAd6amVlZZo8ebK2bNmSfq+np0dbtmzR1KlT+/2ZqVOn9jlekjZv3pzx+DAqpF4k6ZFHHtGDDz6oF154QVdffbUXRfVcvnVz6aWX6p133lFLS0v6ddNNN2nWrFlqaWlRTU2Nl8W3ppBzZvr06WptbU2HvCT99a9/VVVVlTOBJhVWNydOnDgnuHrD30R4H3nr19+iLDfx2fr16008Hjfr1q0zu3btMnfddZcZPny4aW9vN8YYM3/+fNPY2Jg+/vXXXzfnnXeeWblypdm9e7dpampydkl/PvWyYsUKU1ZWZv74xz+aDz74IP3q6Ojw659gTb51czZXVz/mWy8HDx405eXlZvHixWbv3r3m+eefN6NGjTI//vGP/fonWJNv3TQ1NZny8nLzu9/9zuzbt8/8+c9/NuPHjze33nqrX/8EKzo6OszOnTvNzp07jSTz6KOPmp07d5q///3vxhhjGhsbzfz589PH9y7p/+53v2t2795tVq9ezZL+/qxatcpcdNFFpqyszFxzzTXmjTfeSH9vxowZZsGCBX2O//3vf28mTpxoysrKzOWXX242btzocYm9kU+9XHzxxUbSOa+mpibvC+6BfM+ZM7kaasbkXy9/+ctfzJQpU0w8Hje1tbXmoYceMt3d3R6X2hv51M2pU6fMj370IzN+/HgzZMgQU1NTY+655x7z73//2/uCW/TKK6/0e93orYsFCxaYGTNmnPMzV111lSkrKzO1tbXmiSeeKFp5eJ4aAMAZoZ9TAwCgF6EGAHAGoQYAcAahBgBwBqEGAHAGoQYAcAahBgBwBqEGAHAGoQYAcAahBgBwBqEGAHDG/wEJhQ4y7fcVhgAAAABJRU5ErkJggg==",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# a randomly drawn sine curve for illustration\n",
"\n",
"plt.figure(figsize=(5,3))\n",
"np.random.seed(0)\n",
"\n",
"# draw random parameter from prior\n",
"y_true = [var.rvs(1)[0] for var in priors]\n",
"\n",
"# add fixed gaussian noise of 1\n",
"x_err = 0.2\n",
"x_obs = sine(y_true) + np.random.normal(size=50) * x_err\n",
"\n",
"plt.errorbar(t, x_obs, yerr=x_err, fmt='.')\n",
"plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Use case 1: Amortized Neural Posterior Estimation\n",
"if we want to derive the posterior for a large number of different x_obs, then train nbi for 1 round on a large number of samples generated from the prior."
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [],
"source": [
"# the noise function can include any component of the forward model that is cheap\n",
"# therefore it takes in both data (x) and parameter (y) as input\n",
"# Here let us only consider iid Gaussian noise\n",
"\n",
"def noise(x, y):\n",
" rand = np.random.normal(0, 1, size=x.shape[0])\n",
" \n",
" # let's say x_err is drawn from [0.05, 0.5]\n",
" x_err = np.random.uniform() * 0.45 + 0.05\n",
" x_noise = x + rand * x_err\n",
" return x_noise, y"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {
"scrolled": true
},
"outputs": [],
"source": [
"# hyperparameters for the normalizing flow\n",
"flow = {\n",
" 'n_dims': 3, # dimension of parameter space\n",
" 'flow_hidden': 128,\n",
" 'num_blocks': 5,\n",
" 'n_mog':4 # Number of Mixture of Gaussian as base density\n",
"}\n",
"\n",
"# the NBI package provides the \"ResNet-GRU\" network as the default\n",
"# featurizer network for sequential data\n",
"featurizer = {\n",
" 'type': 'resnet-gru',\n",
" 'norm': 'weight_norm',\n",
" 'dim_in': 1, # Number of channels for input data\n",
" 'dim_out': 128, # Output feature vector dimension\n",
" 'dim_conv_max': 256, # Maximum hidden dimension for CNN\n",
" 'depth': 3 # Number of 1D ResNet layers\n",
"}\n",
"\n",
"# initialize NBI engine\n",
"engine = nbi.NBI(\n",
" flow=flow,\n",
" featurizer=featurizer,\n",
" simulator=sine,\n",
" priors=priors,\n",
" labels=labels,\n",
" device='cpu',\n",
" path='test2',\n",
" n_jobs=10\n",
")"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {
"scrolled": true
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Auto learning rate to min_lr = 1.3333333333333334e-06\n",
"Generating simulations\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"1280it [00:00, 2693.36it/s]\n",
"/Users/Keming/anaconda3/envs/py311-arm/lib/python3.11/multiprocessing/resource_tracker.py:224: UserWarning: resource_tracker: There appear to be 1 leaked semaphore objects to clean up at shutdown\n",
" warnings.warn('resource_tracker: There appear to be %d '\n",
"/Users/Keming/anaconda3/envs/py311-arm/lib/python3.11/multiprocessing/resource_tracker.py:224: UserWarning: resource_tracker: There appear to be 1 leaked semaphore objects to clean up at shutdown\n",
" warnings.warn('resource_tracker: There appear to be %d '\n",
"/Users/Keming/anaconda3/envs/py311-arm/lib/python3.11/multiprocessing/resource_tracker.py:224: UserWarning: resource_tracker: There appear to be 1 leaked semaphore objects to clean up at shutdown\n",
" warnings.warn('resource_tracker: There appear to be %d '\n",
"/Users/Keming/anaconda3/envs/py311-arm/lib/python3.11/multiprocessing/resource_tracker.py:224: UserWarning: resource_tracker: There appear to be 1 leaked semaphore objects to clean up at shutdown\n",
" warnings.warn('resource_tracker: There appear to be %d '\n",
"/Users/Keming/anaconda3/envs/py311-arm/lib/python3.11/multiprocessing/resource_tracker.py:224: UserWarning: resource_tracker: There appear to be 1 leaked semaphore objects to clean up at shutdown\n",
" warnings.warn('resource_tracker: There appear to be %d '\n",
"/Users/Keming/anaconda3/envs/py311-arm/lib/python3.11/multiprocessing/resource_tracker.py:224: UserWarning: resource_tracker: There appear to be 1 leaked semaphore objects to clean up at shutdown\n",
" warnings.warn('resource_tracker: There appear to be %d '\n",
"/Users/Keming/anaconda3/envs/py311-arm/lib/python3.11/multiprocessing/resource_tracker.py:224: UserWarning: resource_tracker: There appear to be 1 leaked semaphore objects to clean up at shutdown\n",
" warnings.warn('resource_tracker: There appear to be %d '\n",
"/Users/Keming/anaconda3/envs/py311-arm/lib/python3.11/multiprocessing/resource_tracker.py:224: UserWarning: resource_tracker: There appear to be 1 leaked semaphore objects to clean up at shutdown\n",
" warnings.warn('resource_tracker: There appear to be %d '\n",
"/Users/Keming/anaconda3/envs/py311-arm/lib/python3.11/multiprocessing/resource_tracker.py:224: UserWarning: resource_tracker: There appear to be 1 leaked semaphore objects to clean up at shutdown\n",
" warnings.warn('resource_tracker: There appear to be %d '\n",
"/Users/Keming/anaconda3/envs/py311-arm/lib/python3.11/multiprocessing/resource_tracker.py:224: UserWarning: resource_tracker: There appear to be 1 leaked semaphore objects to clean up at shutdown\n",
" warnings.warn('resource_tracker: There appear to be %d '\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"---------------------- Round: 0 ----------------------\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 0: Train, Loglike in nats: -3.851653: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: -2.984723: 80%|▊| 1024/1280 [00:00<00:00, 2999.58it\n",
"Epoch 1: Train, Loglike in nats: -2.498076: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: -2.240993: 80%|▊| 1024/1280 [00:00<00:00, 3671.05it\n",
"Epoch 2: Train, Loglike in nats: -2.853344: 98%|▉| 11264/11520 [00:04<00:00\n",
"- Val, Loglike in nats: -2.363469: 80%|▊| 1024/1280 [00:00<00:00, 8830.57it\n",
"Epoch 3: Train, Loglike in nats: -2.240685: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: -1.919772: 80%|▊| 1024/1280 [00:00<00:00, 8462.06it\n",
"Epoch 4: Train, Loglike in nats: -1.620814: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: -1.600617: 80%|▊| 1024/1280 [00:00<00:00, 9509.40it\n",
"Epoch 5: Train, Loglike in nats: -1.647011: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: -1.350998: 80%|▊| 1024/1280 [00:00<00:00, 9575.18it\n",
"Epoch 6: Train, Loglike in nats: -1.284166: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: -1.201143: 80%|▊| 1024/1280 [00:00<00:00, 10342.09i\n",
"Epoch 7: Train, Loglike in nats: -1.126907: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: -0.892067: 80%|▊| 1024/1280 [00:00<00:00, 9703.31it\n",
"Epoch 8: Train, Loglike in nats: -0.873063: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: -0.581992: 80%|▊| 1024/1280 [00:00<00:00, 9966.42it\n",
"Epoch 9: Train, Loglike in nats: -0.389662: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: -0.010337: 80%|▊| 1024/1280 [00:00<00:00, 9833.00it\n",
"Epoch 10: Train, Loglike in nats: 0.248763: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 1.072841: 80%|▊| 1024/1280 [00:00<00:00, 9800.09it/\n",
"Epoch 11: Train, Loglike in nats: -0.398143: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 0.714003: 80%|▊| 1024/1280 [00:00<00:00, 9971.46it/\n",
"Epoch 12: Train, Loglike in nats: 1.219615: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 1.337432: 80%|▊| 1024/1280 [00:00<00:00, 9759.74it/\n",
"Epoch 13: Train, Loglike in nats: 1.615358: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: -0.473811: 80%|▊| 1024/1280 [00:00<00:00, 9705.31it\n",
"Epoch 14: Train, Loglike in nats: 1.239683: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 1.872741: 80%|▊| 1024/1280 [00:00<00:00, 9798.57it/\n",
"Epoch 15: Train, Loglike in nats: 2.123171: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 2.436503: 80%|▊| 1024/1280 [00:00<00:00, 9982.35it/\n",
"Epoch 16: Train, Loglike in nats: 1.984375: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 2.195008: 80%|▊| 1024/1280 [00:00<00:00, 10073.95it\n",
"Epoch 17: Train, Loglike in nats: 2.481320: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: 2.621574: 80%|▊| 1024/1280 [00:00<00:00, 3150.82it/\n",
"Epoch 18: Train, Loglike in nats: 1.901767: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 2.066959: 80%|▊| 1024/1280 [00:00<00:00, 9610.88it/\n",
"Epoch 19: Train, Loglike in nats: 2.438454: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 2.733421: 80%|▊| 1024/1280 [00:00<00:00, 10007.22it\n",
"Epoch 20: Train, Loglike in nats: 2.840736: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 2.403073: 80%|▊| 1024/1280 [00:00<00:00, 9782.39it/\n",
"Epoch 21: Train, Loglike in nats: 2.468852: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 3.099226: 80%|▊| 1024/1280 [00:00<00:00, 4324.23it/\n",
"Epoch 22: Train, Loglike in nats: 2.001492: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 1.776518: 80%|▊| 1024/1280 [00:00<00:00, 9607.55it/\n",
"Epoch 23: Train, Loglike in nats: 1.780422: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 2.564617: 80%|▊| 1024/1280 [00:00<00:00, 9735.31it/\n",
"Epoch 24: Train, Loglike in nats: 2.985596: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 3.067122: 80%|▊| 1024/1280 [00:00<00:00, 9614.73it/\n",
"Epoch 25: Train, Loglike in nats: 3.349774: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 3.073960: 80%|▊| 1024/1280 [00:00<00:00, 10122.69it\n",
"Epoch 26: Train, Loglike in nats: 3.362078: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: 3.403142: 80%|▊| 1024/1280 [00:00<00:00, 9832.35it/\n",
"Epoch 27: Train, Loglike in nats: 2.625667: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 2.227534: 80%|▊| 1024/1280 [00:00<00:00, 9904.25it/\n",
"Epoch 28: Train, Loglike in nats: 3.055156: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 3.482126: 80%|▊| 1024/1280 [00:00<00:00, 9950.35it/\n",
"Epoch 29: Train, Loglike in nats: 3.519974: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 3.781979: 80%|▊| 1024/1280 [00:00<00:00, 10020.15it\n",
"Epoch 30: Train, Loglike in nats: 3.705919: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: 3.777465: 80%|▊| 1024/1280 [00:00<00:00, 9517.87it/\n",
"Epoch 31: Train, Loglike in nats: 3.792369: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 3.554900: 80%|▊| 1024/1280 [00:00<00:00, 9846.30it/\n",
"Epoch 32: Train, Loglike in nats: 3.811675: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.087920: 80%|▊| 1024/1280 [00:00<00:00, 9896.33it/\n",
"Epoch 33: Train, Loglike in nats: 3.368757: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 3.579859: 80%|▊| 1024/1280 [00:00<00:00, 9677.36it/\n",
"Epoch 34: Train, Loglike in nats: 3.884577: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 3.275751: 80%|▊| 1024/1280 [00:00<00:00, 4258.89it/\n",
"Epoch 35: Train, Loglike in nats: 3.244787: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: 2.795378: 80%|▊| 1024/1280 [00:00<00:00, 9711.39it/\n",
"Epoch 36: Train, Loglike in nats: 3.677319: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 3.880242: 80%|▊| 1024/1280 [00:00<00:00, 9689.26it/\n",
"Epoch 37: Train, Loglike in nats: 3.790177: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 3.772333: 80%|▊| 1024/1280 [00:00<00:00, 9088.26it/\n",
"Epoch 38: Train, Loglike in nats: 2.899288: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 2.833591: 80%|▊| 1024/1280 [00:00<00:00, 9678.93it/\n",
"Epoch 39: Train, Loglike in nats: 3.666038: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: 3.546713: 80%|▊| 1024/1280 [00:00<00:00, 4539.23it/\n",
"Epoch 40: Train, Loglike in nats: 3.843576: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 3.989748: 80%|▊| 1024/1280 [00:00<00:00, 9549.85it/\n",
"Epoch 41: Train, Loglike in nats: 4.191560: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.034100: 80%|▊| 1024/1280 [00:00<00:00, 10026.72it\n",
"Epoch 42: Train, Loglike in nats: 4.227902: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 3.976105: 80%|▊| 1024/1280 [00:00<00:00, 9976.23it/\n",
"Epoch 43: Train, Loglike in nats: 4.240566: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.241155: 80%|▊| 1024/1280 [00:00<00:00, 3762.98it/\n",
"Epoch 44: Train, Loglike in nats: 4.260778: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: 4.282600: 80%|▊| 1024/1280 [00:00<00:00, 9537.10it/\n",
"Epoch 45: Train, Loglike in nats: 4.029519: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 3.595071: 80%|▊| 1024/1280 [00:00<00:00, 9886.37it/\n",
"Epoch 46: Train, Loglike in nats: 3.826512: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.256767: 80%|▊| 1024/1280 [00:00<00:00, 9860.84it/\n",
"Epoch 47: Train, Loglike in nats: 4.325788: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: 2.841599: 80%|▊| 1024/1280 [00:00<00:00, 9614.02it/\n",
"Epoch 48: Train, Loglike in nats: 3.838478: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.202137: 80%|▊| 1024/1280 [00:00<00:00, 8691.08it/\n",
"Epoch 49: Train, Loglike in nats: 4.236606: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 2.871080: 80%|▊| 1024/1280 [00:00<00:00, 9852.79it/\n",
"Epoch 50: Train, Loglike in nats: 3.959212: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.223527: 80%|▊| 1024/1280 [00:00<00:00, 9810.30it/\n",
"Epoch 51: Train, Loglike in nats: 4.446829: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: 4.428692: 80%|▊| 1024/1280 [00:00<00:00, 9974.86it/\n",
"Epoch 52: Train, Loglike in nats: 4.377276: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.429189: 80%|▊| 1024/1280 [00:00<00:00, 9921.32it/\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 53: Train, Loglike in nats: 4.444990: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.276281: 80%|▊| 1024/1280 [00:00<00:00, 9887.60it/\n",
"Epoch 54: Train, Loglike in nats: 3.256611: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 3.587253: 80%|▊| 1024/1280 [00:00<00:00, 9367.15it/\n",
"Epoch 55: Train, Loglike in nats: 4.213576: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: 4.282850: 80%|▊| 1024/1280 [00:00<00:00, 9826.30it/\n",
"Epoch 56: Train, Loglike in nats: 3.593654: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 3.533817: 80%|▊| 1024/1280 [00:00<00:00, 8401.16it/\n",
"Epoch 57: Train, Loglike in nats: 4.071354: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.336431: 80%|▊| 1024/1280 [00:00<00:00, 9924.11it/\n",
"Epoch 58: Train, Loglike in nats: 4.523057: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.639695: 80%|▊| 1024/1280 [00:00<00:00, 9383.29it/\n",
"Epoch 59: Train, Loglike in nats: 4.584000: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: 4.797822: 80%|▊| 1024/1280 [00:00<00:00, 9560.60it/\n",
"Epoch 60: Train, Loglike in nats: 4.451227: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.397796: 80%|▊| 1024/1280 [00:00<00:00, 9511.95it/\n",
"Epoch 61: Train, Loglike in nats: 4.597460: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.642589: 80%|▊| 1024/1280 [00:00<00:00, 9789.95it/\n",
"Epoch 62: Train, Loglike in nats: 4.608378: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.271307: 80%|▊| 1024/1280 [00:00<00:00, 9862.38it/\n",
"Epoch 63: Train, Loglike in nats: 4.524945: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: 4.542982: 80%|▊| 1024/1280 [00:00<00:00, 9657.99it/\n",
"Epoch 64: Train, Loglike in nats: 4.180767: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 3.689496: 80%|▊| 1024/1280 [00:00<00:00, 9750.23it/\n",
"Epoch 65: Train, Loglike in nats: 4.373944: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.675422: 80%|▊| 1024/1280 [00:00<00:00, 9866.28it/\n",
"Epoch 66: Train, Loglike in nats: 4.692604: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.092716: 80%|▊| 1024/1280 [00:00<00:00, 10000.88it\n",
"Epoch 67: Train, Loglike in nats: 4.693291: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.835148: 80%|▊| 1024/1280 [00:00<00:00, 4596.81it/\n",
"Epoch 68: Train, Loglike in nats: 4.714162: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.709657: 80%|▊| 1024/1280 [00:00<00:00, 9798.12it/\n",
"Epoch 69: Train, Loglike in nats: 3.702480: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.176120: 80%|▊| 1024/1280 [00:00<00:00, 9810.68it/\n",
"Epoch 70: Train, Loglike in nats: 4.596511: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.895304: 80%|▊| 1024/1280 [00:00<00:00, 9745.98it/\n",
"Epoch 71: Train, Loglike in nats: 4.744214: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.635634: 80%|▊| 1024/1280 [00:00<00:00, 3372.82it/\n",
"Epoch 72: Train, Loglike in nats: 4.747933: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.404232: 80%|▊| 1024/1280 [00:00<00:00, 9886.58it/\n",
"Epoch 73: Train, Loglike in nats: 4.754324: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.708319: 80%|▊| 1024/1280 [00:00<00:00, 9816.69it/\n",
"Epoch 74: Train, Loglike in nats: 4.764262: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.911592: 80%|▊| 1024/1280 [00:00<00:00, 9560.09it/\n",
"Epoch 75: Train, Loglike in nats: 4.751526: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: 4.937023: 80%|▊| 1024/1280 [00:00<00:00, 9830.17it/\n",
"Epoch 76: Train, Loglike in nats: 4.845479: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.812368: 80%|▊| 1024/1280 [00:00<00:00, 9621.54it/\n",
"Epoch 77: Train, Loglike in nats: 4.756401: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.579947: 80%|▊| 1024/1280 [00:00<00:00, 9523.02it/\n",
"Epoch 78: Train, Loglike in nats: 4.751733: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.718035: 80%|▊| 1024/1280 [00:00<00:00, 9685.57it/\n",
"Epoch 79: Train, Loglike in nats: 4.659982: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: 4.533284: 80%|▊| 1024/1280 [00:00<00:00, 9395.59it/\n",
"Epoch 80: Train, Loglike in nats: 4.834358: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.939500: 80%|▊| 1024/1280 [00:00<00:00, 9704.50it/\n",
"Epoch 81: Train, Loglike in nats: 4.158889: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.597799: 80%|▊| 1024/1280 [00:00<00:00, 9613.78it/\n",
"Epoch 82: Train, Loglike in nats: 4.799724: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: 4.801846: 80%|▊| 1024/1280 [00:00<00:00, 3328.98it/\n",
"Epoch 83: Train, Loglike in nats: 4.938389: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.943264: 80%|▊| 1024/1280 [00:00<00:00, 8806.11it/\n",
"Epoch 84: Train, Loglike in nats: 4.793490: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.866713: 80%|▊| 1024/1280 [00:00<00:00, 8939.41it/\n",
"Epoch 85: Train, Loglike in nats: 4.981260: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: 4.867730: 80%|▊| 1024/1280 [00:00<00:00, 8745.48it/\n",
"Epoch 86: Train, Loglike in nats: 4.620130: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.890939: 80%|▊| 1024/1280 [00:00<00:00, 8800.33it/\n",
"Epoch 87: Train, Loglike in nats: 4.935327: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: 5.109634: 80%|▊| 1024/1280 [00:00<00:00, 3369.02it/\n",
"Epoch 88: Train, Loglike in nats: 5.010200: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: 4.557707: 80%|▊| 1024/1280 [00:00<00:00, 9224.90it/\n",
"Epoch 89: Train, Loglike in nats: 4.949794: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: 5.041967: 80%|▊| 1024/1280 [00:00<00:00, 8892.27it/\n",
"Epoch 90: Train, Loglike in nats: 5.044551: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: 4.961290: 80%|▊| 1024/1280 [00:00<00:00, 8944.16it/\n",
"Epoch 91: Train, Loglike in nats: 5.020473: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.783405: 80%|▊| 1024/1280 [00:00<00:00, 8682.10it/\n",
"Epoch 92: Train, Loglike in nats: 5.048864: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: 4.879153: 80%|▊| 1024/1280 [00:00<00:00, 4046.55it/\n",
"Epoch 93: Train, Loglike in nats: 4.972085: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: 5.095383: 80%|▊| 1024/1280 [00:00<00:00, 9264.28it/\n",
"Epoch 94: Train, Loglike in nats: 5.011032: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.682728: 80%|▊| 1024/1280 [00:00<00:00, 9485.21it/\n",
"Epoch 95: Train, Loglike in nats: 5.034967: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: 4.982537: 80%|▊| 1024/1280 [00:00<00:00, 2814.13it/\n",
"Epoch 96: Train, Loglike in nats: 4.959682: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 5.074445: 80%|▊| 1024/1280 [00:00<00:00, 9143.17it/\n",
"Epoch 97: Train, Loglike in nats: 5.148802: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 5.195337: 80%|▊| 1024/1280 [00:00<00:00, 9187.15it/\n",
"Epoch 98: Train, Loglike in nats: 5.189330: 98%|▉| 11264/11520 [00:03<00:00\n",
"- Val, Loglike in nats: 5.026703: 80%|▊| 1024/1280 [00:00<00:00, 9295.02it/\n",
"Epoch 99: Train, Loglike in nats: 5.186990: 98%|▉| 11264/11520 [00:02<00:00\n",
"- Val, Loglike in nats: 4.806613: 80%|▊| 1024/1280 [00:00<00:00, 8731.31it/\n",
"Epoch 100: Train, Loglike in nats: 4.410088: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 4.800928: 80%|▊| 1024/1280 [00:00<00:00, 9437.48it/\n",
"Epoch 101: Train, Loglike in nats: 5.033157: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 4.942412: 80%|▊| 1024/1280 [00:00<00:00, 8906.82it/\n",
"Epoch 102: Train, Loglike in nats: 5.235607: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.188683: 80%|▊| 1024/1280 [00:00<00:00, 8963.29it/\n",
"Epoch 103: Train, Loglike in nats: 5.226388: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.216347: 80%|▊| 1024/1280 [00:00<00:00, 5639.42it/\n",
"Epoch 104: Train, Loglike in nats: 5.231412: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.187630: 80%|▊| 1024/1280 [00:00<00:00, 9436.56it/\n",
"Epoch 105: Train, Loglike in nats: 4.746549: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 3.862481: 80%|▊| 1024/1280 [00:00<00:00, 9128.75it/\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 106: Train, Loglike in nats: 4.637647: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.023439: 80%|▊| 1024/1280 [00:00<00:00, 3347.82it/\n",
"Epoch 107: Train, Loglike in nats: 5.183452: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 4.972174: 80%|▊| 1024/1280 [00:00<00:00, 9162.91it/\n",
"Epoch 108: Train, Loglike in nats: 5.198407: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.045882: 80%|▊| 1024/1280 [00:00<00:00, 9146.85it/\n",
"Epoch 109: Train, Loglike in nats: 5.246219: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.143646: 80%|▊| 1024/1280 [00:00<00:00, 8695.50it/\n",
"Epoch 110: Train, Loglike in nats: 5.315694: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.255485: 80%|▊| 1024/1280 [00:00<00:00, 9071.33it/\n",
"Epoch 111: Train, Loglike in nats: 5.331820: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 4.875888: 80%|▊| 1024/1280 [00:00<00:00, 8933.55it/\n",
"Epoch 112: Train, Loglike in nats: 5.283599: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.465637: 80%|▊| 1024/1280 [00:00<00:00, 9557.75it/\n",
"Epoch 113: Train, Loglike in nats: 5.402383: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.397207: 80%|▊| 1024/1280 [00:00<00:00, 9627.51it/\n",
"Epoch 114: Train, Loglike in nats: 5.212236: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 4.660688: 80%|▊| 1024/1280 [00:00<00:00, 8635.08it/\n",
"Epoch 115: Train, Loglike in nats: 3.903254: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 4.963943: 80%|▊| 1024/1280 [00:00<00:00, 9256.10it/\n",
"Epoch 116: Train, Loglike in nats: 5.131039: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.248237: 80%|▊| 1024/1280 [00:00<00:00, 4741.29it/\n",
"Epoch 117: Train, Loglike in nats: 5.332259: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.280462: 80%|▊| 1024/1280 [00:00<00:00, 9437.54it/\n",
"Epoch 118: Train, Loglike in nats: 5.310867: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.365846: 80%|▊| 1024/1280 [00:00<00:00, 8837.58it/\n",
"Epoch 119: Train, Loglike in nats: 5.397320: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.448100: 80%|▊| 1024/1280 [00:00<00:00, 9051.22it/\n",
"Epoch 120: Train, Loglike in nats: 5.294904: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.092909: 80%|▊| 1024/1280 [00:00<00:00, 9444.22it/\n",
"Epoch 121: Train, Loglike in nats: 5.403662: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.428323: 80%|▊| 1024/1280 [00:00<00:00, 9349.89it/\n",
"Epoch 122: Train, Loglike in nats: 4.936172: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 4.632895: 80%|▊| 1024/1280 [00:00<00:00, 9073.29it/\n",
"Epoch 123: Train, Loglike in nats: 4.890215: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.241644: 80%|▊| 1024/1280 [00:00<00:00, 8893.91it/\n",
"Epoch 124: Train, Loglike in nats: 5.355583: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.400315: 80%|▊| 1024/1280 [00:00<00:00, 8969.02it/\n",
"Epoch 125: Train, Loglike in nats: 5.431152: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.438829: 80%|▊| 1024/1280 [00:00<00:00, 9106.68it/\n",
"Epoch 126: Train, Loglike in nats: 5.402216: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.395393: 80%|▊| 1024/1280 [00:00<00:00, 8714.45it/\n",
"Epoch 127: Train, Loglike in nats: 5.436257: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.518850: 80%|▊| 1024/1280 [00:00<00:00, 8462.24it/\n",
"Epoch 128: Train, Loglike in nats: 5.470064: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.469751: 80%|▊| 1024/1280 [00:00<00:00, 8691.68it/\n",
"Epoch 129: Train, Loglike in nats: 5.401027: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.540334: 80%|▊| 1024/1280 [00:00<00:00, 9273.42it/\n",
"Epoch 130: Train, Loglike in nats: 5.290651: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 4.525930: 80%|▊| 1024/1280 [00:00<00:00, 3103.10it/\n",
"Epoch 131: Train, Loglike in nats: 5.193254: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.429794: 80%|▊| 1024/1280 [00:00<00:00, 9134.71it/\n",
"Epoch 132: Train, Loglike in nats: 5.392039: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.312921: 80%|▊| 1024/1280 [00:00<00:00, 9025.37it/\n",
"Epoch 133: Train, Loglike in nats: 5.464343: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.525427: 80%|▊| 1024/1280 [00:00<00:00, 9260.35it/\n",
"Epoch 134: Train, Loglike in nats: 5.527062: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.470475: 80%|▊| 1024/1280 [00:00<00:00, 8726.77it/\n",
"Epoch 135: Train, Loglike in nats: 5.161821: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.504158: 80%|▊| 1024/1280 [00:00<00:00, 8916.51it/\n",
"Epoch 136: Train, Loglike in nats: 5.545118: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.499725: 80%|▊| 1024/1280 [00:00<00:00, 9071.98it/\n",
"Epoch 137: Train, Loglike in nats: 5.584017: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.566711: 80%|▊| 1024/1280 [00:00<00:00, 8864.12it/\n",
"Epoch 138: Train, Loglike in nats: 5.560418: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.479045: 80%|▊| 1024/1280 [00:00<00:00, 4167.45it/\n",
"Epoch 139: Train, Loglike in nats: 5.583067: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.404629: 80%|▊| 1024/1280 [00:00<00:00, 8566.67it/\n",
"Epoch 140: Train, Loglike in nats: 5.499278: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.194180: 80%|▊| 1024/1280 [00:00<00:00, 9151.35it/\n",
"Epoch 141: Train, Loglike in nats: 5.478994: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.141514: 80%|▊| 1024/1280 [00:00<00:00, 9006.94it/\n",
"Epoch 142: Train, Loglike in nats: 5.513113: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.509935: 80%|▊| 1024/1280 [00:00<00:00, 8835.91it/\n",
"Epoch 143: Train, Loglike in nats: 5.606976: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.353518: 80%|▊| 1024/1280 [00:00<00:00, 8432.86it/\n",
"Epoch 144: Train, Loglike in nats: 5.556012: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.529669: 80%|▊| 1024/1280 [00:00<00:00, 9035.70it/\n",
"Epoch 145: Train, Loglike in nats: 5.557709: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.637734: 80%|▊| 1024/1280 [00:00<00:00, 9365.55it/\n",
"Epoch 146: Train, Loglike in nats: 5.603655: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.427556: 80%|▊| 1024/1280 [00:00<00:00, 9342.80it/\n",
"Epoch 147: Train, Loglike in nats: 5.493373: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 4.985003: 80%|▊| 1024/1280 [00:00<00:00, 9056.11it/\n",
"Epoch 148: Train, Loglike in nats: 5.163940: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.225825: 80%|▊| 1024/1280 [00:00<00:00, 8907.52it/\n",
"Epoch 149: Train, Loglike in nats: 5.390919: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.512096: 80%|▊| 1024/1280 [00:00<00:00, 3878.39it/\n",
"Epoch 150: Train, Loglike in nats: 5.563761: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.757576: 80%|▊| 1024/1280 [00:00<00:00, 9103.60it/\n",
"Epoch 151: Train, Loglike in nats: 5.659038: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.611406: 80%|▊| 1024/1280 [00:00<00:00, 9320.96it/\n",
"Epoch 152: Train, Loglike in nats: 5.726348: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.555246: 80%|▊| 1024/1280 [00:00<00:00, 9146.52it/\n",
"Epoch 153: Train, Loglike in nats: 5.600896: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.613087: 80%|▊| 1024/1280 [00:00<00:00, 9125.41it/\n",
"Epoch 154: Train, Loglike in nats: 5.136520: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 4.812824: 80%|▊| 1024/1280 [00:00<00:00, 7323.13it/\n",
"Epoch 155: Train, Loglike in nats: 5.125684: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.097010: 80%|▊| 1024/1280 [00:00<00:00, 3785.95it/\n",
"Epoch 156: Train, Loglike in nats: 5.510932: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.518552: 80%|▊| 1024/1280 [00:00<00:00, 9420.17it/\n",
"Epoch 157: Train, Loglike in nats: 5.649294: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.620698: 80%|▊| 1024/1280 [00:00<00:00, 9148.64it/\n",
"Epoch 158: Train, Loglike in nats: 5.640782: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.626383: 80%|▊| 1024/1280 [00:00<00:00, 9523.88it/\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 159: Train, Loglike in nats: 5.649284: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.628625: 80%|▊| 1024/1280 [00:00<00:00, 9511.76it/\n",
"Epoch 160: Train, Loglike in nats: 5.646411: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.577348: 80%|▊| 1024/1280 [00:00<00:00, 9199.28it/\n",
"Epoch 161: Train, Loglike in nats: 5.624748: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.624774: 80%|▊| 1024/1280 [00:00<00:00, 4714.98it/\n",
"Epoch 162: Train, Loglike in nats: 5.749968: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.792416: 80%|▊| 1024/1280 [00:00<00:00, 9005.11it/\n",
"Epoch 163: Train, Loglike in nats: 5.768485: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.791962: 80%|▊| 1024/1280 [00:00<00:00, 9167.66it/\n",
"Epoch 164: Train, Loglike in nats: 5.813198: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.494507: 80%|▊| 1024/1280 [00:00<00:00, 9091.55it/\n",
"Epoch 165: Train, Loglike in nats: 5.487433: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.293005: 80%|▊| 1024/1280 [00:00<00:00, 9121.44it/\n",
"Epoch 166: Train, Loglike in nats: 5.668660: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.703479: 80%|▊| 1024/1280 [00:00<00:00, 9244.12it/\n",
"Epoch 167: Train, Loglike in nats: 5.764008: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.684761: 80%|▊| 1024/1280 [00:00<00:00, 2754.00it/\n",
"Epoch 168: Train, Loglike in nats: 5.502714: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.555971: 80%|▊| 1024/1280 [00:00<00:00, 8929.99it/\n",
"Epoch 169: Train, Loglike in nats: 5.602808: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.726156: 80%|▊| 1024/1280 [00:00<00:00, 8835.36it/\n",
"Epoch 170: Train, Loglike in nats: 5.743564: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.610476: 80%|▊| 1024/1280 [00:00<00:00, 8582.04it/\n",
"Epoch 171: Train, Loglike in nats: 5.600174: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 4.870948: 80%|▊| 1024/1280 [00:00<00:00, 8936.60it/\n",
"Epoch 172: Train, Loglike in nats: 5.443271: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.456948: 80%|▊| 1024/1280 [00:00<00:00, 4551.76it/\n",
"Epoch 173: Train, Loglike in nats: 5.688820: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.759488: 80%|▊| 1024/1280 [00:00<00:00, 6678.28it/\n",
"Epoch 174: Train, Loglike in nats: 5.863158: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.839815: 80%|▊| 1024/1280 [00:00<00:00, 8779.68it/\n",
"Epoch 175: Train, Loglike in nats: 5.826236: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.789112: 80%|▊| 1024/1280 [00:00<00:00, 8776.58it/\n",
"Epoch 176: Train, Loglike in nats: 5.864328: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.841046: 80%|▊| 1024/1280 [00:00<00:00, 9252.31it/\n",
"Epoch 177: Train, Loglike in nats: 5.817114: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.772132: 80%|▊| 1024/1280 [00:00<00:00, 8582.61it/\n",
"Epoch 178: Train, Loglike in nats: 5.864300: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.811681: 80%|▊| 1024/1280 [00:00<00:00, 8744.68it/\n",
"Epoch 179: Train, Loglike in nats: 5.878690: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.612947: 80%|▊| 1024/1280 [00:00<00:00, 8825.52it/\n",
"Epoch 180: Train, Loglike in nats: 5.813038: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.817533: 80%|▊| 1024/1280 [00:00<00:00, 4110.87it/\n",
"Epoch 181: Train, Loglike in nats: 5.788204: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.734297: 80%|▊| 1024/1280 [00:00<00:00, 8817.26it/\n",
"Epoch 182: Train, Loglike in nats: 5.840208: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.852358: 80%|▊| 1024/1280 [00:00<00:00, 9053.87it/\n",
"Epoch 183: Train, Loglike in nats: 5.907214: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.763256: 80%|▊| 1024/1280 [00:00<00:00, 9166.51it/\n",
"Epoch 184: Train, Loglike in nats: 5.852499: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.862482: 80%|▊| 1024/1280 [00:00<00:00, 8775.91it/\n",
"Epoch 185: Train, Loglike in nats: 5.802456: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.770654: 80%|▊| 1024/1280 [00:00<00:00, 4140.67it/\n",
"Epoch 186: Train, Loglike in nats: 5.839282: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.707422: 80%|▊| 1024/1280 [00:00<00:00, 8670.54it/\n",
"Epoch 187: Train, Loglike in nats: 5.907727: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.821457: 80%|▊| 1024/1280 [00:00<00:00, 8987.90it/\n",
"Epoch 188: Train, Loglike in nats: 5.913223: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.795648: 80%|▊| 1024/1280 [00:00<00:00, 8452.53it/\n",
"Epoch 189: Train, Loglike in nats: 5.868620: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.921119: 80%|▊| 1024/1280 [00:00<00:00, 8634.51it/\n",
"Epoch 190: Train, Loglike in nats: 5.916395: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.820592: 80%|▊| 1024/1280 [00:00<00:00, 3827.15it/\n",
"Epoch 191: Train, Loglike in nats: 5.783606: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.655878: 80%|▊| 1024/1280 [00:00<00:00, 8920.25it/\n",
"Epoch 192: Train, Loglike in nats: 5.930478: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.880970: 80%|▊| 1024/1280 [00:00<00:00, 8703.39it/\n",
"Epoch 193: Train, Loglike in nats: 5.958016: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.913445: 80%|▊| 1024/1280 [00:00<00:00, 9190.04it/\n",
"Epoch 194: Train, Loglike in nats: 5.962620: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.843554: 80%|▊| 1024/1280 [00:00<00:00, 8877.55it/\n",
"Epoch 195: Train, Loglike in nats: 5.979379: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.846795: 80%|▊| 1024/1280 [00:00<00:00, 4059.29it/\n",
"Epoch 196: Train, Loglike in nats: 5.970039: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.873135: 80%|▊| 1024/1280 [00:00<00:00, 8630.20it/\n",
"Epoch 197: Train, Loglike in nats: 5.933705: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.847226: 80%|▊| 1024/1280 [00:00<00:00, 8702.09it/\n",
"Epoch 198: Train, Loglike in nats: 6.035864: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.952021: 80%|▊| 1024/1280 [00:00<00:00, 8835.75it/\n",
"Epoch 199: Train, Loglike in nats: 6.031181: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.905683: 80%|▊| 1024/1280 [00:00<00:00, 8928.04it/\n",
"Epoch 200: Train, Loglike in nats: 5.976564: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.767876: 80%|▊| 1024/1280 [00:00<00:00, 4212.59it/\n",
"Epoch 201: Train, Loglike in nats: 5.867866: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.985537: 80%|▊| 1024/1280 [00:00<00:00, 9176.38it/\n",
"Epoch 202: Train, Loglike in nats: 5.999950: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.933154: 80%|▊| 1024/1280 [00:00<00:00, 8880.80it/\n",
"Epoch 203: Train, Loglike in nats: 5.997913: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.934641: 80%|▊| 1024/1280 [00:00<00:00, 3114.24it/\n",
"Epoch 204: Train, Loglike in nats: 6.016090: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.831066: 80%|▊| 1024/1280 [00:00<00:00, 8482.53it/\n",
"Epoch 205: Train, Loglike in nats: 6.028709: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.927909: 80%|▊| 1024/1280 [00:00<00:00, 8480.80it/\n",
"Epoch 206: Train, Loglike in nats: 6.028294: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.908641: 80%|▊| 1024/1280 [00:00<00:00, 8542.52it/\n",
"Epoch 207: Train, Loglike in nats: 6.023085: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.956295: 80%|▊| 1024/1280 [00:00<00:00, 9020.90it/\n",
"Epoch 208: Train, Loglike in nats: 5.969785: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.960278: 80%|▊| 1024/1280 [00:00<00:00, 4130.62it/\n",
"Epoch 209: Train, Loglike in nats: 6.031216: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.982039: 80%|▊| 1024/1280 [00:00<00:00, 9087.26it/\n",
"Epoch 210: Train, Loglike in nats: 6.069467: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.079695: 80%|▊| 1024/1280 [00:00<00:00, 8758.79it/\n",
"Epoch 211: Train, Loglike in nats: 6.058956: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.048627: 80%|▊| 1024/1280 [00:00<00:00, 8766.99it/\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 212: Train, Loglike in nats: 6.098628: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.035763: 80%|▊| 1024/1280 [00:00<00:00, 8962.71it/\n",
"Epoch 213: Train, Loglike in nats: 6.042898: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.003551: 80%|▊| 1024/1280 [00:00<00:00, 6201.92it/\n",
"Epoch 214: Train, Loglike in nats: 6.053481: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.963653: 80%|▊| 1024/1280 [00:00<00:00, 8837.71it/\n",
"Epoch 215: Train, Loglike in nats: 6.071988: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.977085: 80%|▊| 1024/1280 [00:00<00:00, 8339.86it/\n",
"Epoch 216: Train, Loglike in nats: 6.081740: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.977153: 80%|▊| 1024/1280 [00:00<00:00, 9069.59it/\n",
"Epoch 217: Train, Loglike in nats: 6.076434: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.990683: 80%|▊| 1024/1280 [00:00<00:00, 8634.58it/\n",
"Epoch 218: Train, Loglike in nats: 6.065740: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.861808: 80%|▊| 1024/1280 [00:00<00:00, 8997.92it/\n",
"Epoch 219: Train, Loglike in nats: 6.065818: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.942234: 80%|▊| 1024/1280 [00:00<00:00, 8868.66it/\n",
"Epoch 220: Train, Loglike in nats: 6.134555: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.021082: 80%|▊| 1024/1280 [00:00<00:00, 9039.47it/\n",
"Epoch 221: Train, Loglike in nats: 6.082656: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.081383: 80%|▊| 1024/1280 [00:00<00:00, 5831.84it/\n",
"Epoch 222: Train, Loglike in nats: 6.091392: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.966986: 80%|▊| 1024/1280 [00:00<00:00, 5201.16it/\n",
"Epoch 223: Train, Loglike in nats: 6.097937: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.022796: 80%|▊| 1024/1280 [00:00<00:00, 8698.32it/\n",
"Epoch 224: Train, Loglike in nats: 6.104909: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.983981: 80%|▊| 1024/1280 [00:00<00:00, 9118.67it/\n",
"Epoch 225: Train, Loglike in nats: 6.089239: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.978047: 80%|▊| 1024/1280 [00:00<00:00, 8860.06it/\n",
"Epoch 226: Train, Loglike in nats: 6.107898: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.014693: 80%|▊| 1024/1280 [00:00<00:00, 8717.21it/\n",
"Epoch 227: Train, Loglike in nats: 6.120309: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.006057: 80%|▊| 1024/1280 [00:00<00:00, 8375.95it/\n",
"Epoch 228: Train, Loglike in nats: 6.101321: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.013000: 80%|▊| 1024/1280 [00:00<00:00, 9173.99it/\n",
"Epoch 229: Train, Loglike in nats: 6.103277: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.930502: 80%|▊| 1024/1280 [00:00<00:00, 4906.76it/\n",
"Epoch 230: Train, Loglike in nats: 6.103433: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.133472: 80%|▊| 1024/1280 [00:00<00:00, 4004.72it/\n",
"Epoch 231: Train, Loglike in nats: 6.133838: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.137036: 80%|▊| 1024/1280 [00:00<00:00, 8936.83it/\n",
"Epoch 232: Train, Loglike in nats: 6.146494: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 5.921103: 80%|▊| 1024/1280 [00:00<00:00, 9032.38it/\n",
"Epoch 233: Train, Loglike in nats: 6.133113: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.110283: 80%|▊| 1024/1280 [00:00<00:00, 9605.46it/\n",
"Epoch 234: Train, Loglike in nats: 6.145494: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.016614: 80%|▊| 1024/1280 [00:00<00:00, 9014.71it/\n",
"Epoch 235: Train, Loglike in nats: 6.124565: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.170072: 80%|▊| 1024/1280 [00:00<00:00, 9011.79it/\n",
"Epoch 236: Train, Loglike in nats: 6.157085: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.005931: 80%|▊| 1024/1280 [00:00<00:00, 9075.11it/\n",
"Epoch 237: Train, Loglike in nats: 6.126128: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.098151: 80%|▊| 1024/1280 [00:00<00:00, 6846.62it/\n",
"Epoch 238: Train, Loglike in nats: 6.175637: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 5.992302: 80%|▊| 1024/1280 [00:00<00:00, 8850.24it/\n",
"Epoch 239: Train, Loglike in nats: 6.169770: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.095167: 80%|▊| 1024/1280 [00:00<00:00, 8975.37it/\n",
"Epoch 240: Train, Loglike in nats: 6.162028: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.027113: 80%|▊| 1024/1280 [00:00<00:00, 8647.56it/\n",
"Epoch 241: Train, Loglike in nats: 6.150702: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.091898: 80%|▊| 1024/1280 [00:00<00:00, 9325.43it/\n",
"Epoch 242: Train, Loglike in nats: 6.202268: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.074091: 80%|▊| 1024/1280 [00:00<00:00, 9097.14it/\n",
"Epoch 243: Train, Loglike in nats: 6.195301: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.210519: 80%|▊| 1024/1280 [00:00<00:00, 9122.10it/\n",
"Epoch 244: Train, Loglike in nats: 6.201016: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.221901: 80%|▊| 1024/1280 [00:00<00:00, 8993.98it/\n",
"Epoch 245: Train, Loglike in nats: 6.191539: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.183374: 80%|▊| 1024/1280 [00:00<00:00, 9100.37it/\n",
"Epoch 246: Train, Loglike in nats: 6.205428: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.089732: 80%|▊| 1024/1280 [00:00<00:00, 8974.59it/\n",
"Epoch 247: Train, Loglike in nats: 6.155436: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.109187: 80%|▊| 1024/1280 [00:00<00:00, 8889.49it/\n",
"Epoch 248: Train, Loglike in nats: 6.160395: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.166186: 80%|▊| 1024/1280 [00:00<00:00, 4535.63it/\n",
"Epoch 249: Train, Loglike in nats: 6.218191: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.177854: 80%|▊| 1024/1280 [00:00<00:00, 8780.81it/\n",
"Epoch 250: Train, Loglike in nats: 6.203799: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.156746: 80%|▊| 1024/1280 [00:00<00:00, 9049.62it/\n",
"Epoch 251: Train, Loglike in nats: 6.183799: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.134890: 80%|▊| 1024/1280 [00:00<00:00, 8978.81it/\n",
"Epoch 252: Train, Loglike in nats: 6.192711: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.215394: 80%|▊| 1024/1280 [00:00<00:00, 8987.04it/\n",
"Epoch 253: Train, Loglike in nats: 6.202236: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.144240: 80%|▊| 1024/1280 [00:00<00:00, 9499.98it/\n",
"Epoch 254: Train, Loglike in nats: 6.226951: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.201624: 80%|▊| 1024/1280 [00:00<00:00, 6926.77it/\n",
"Epoch 255: Train, Loglike in nats: 6.228964: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.221393: 80%|▊| 1024/1280 [00:00<00:00, 8973.48it/\n",
"Epoch 256: Train, Loglike in nats: 6.216074: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.090775: 80%|▊| 1024/1280 [00:00<00:00, 9150.84it/\n",
"Epoch 257: Train, Loglike in nats: 6.247015: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.176671: 80%|▊| 1024/1280 [00:00<00:00, 9582.45it/\n",
"Epoch 258: Train, Loglike in nats: 6.218173: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.134167: 80%|▊| 1024/1280 [00:00<00:00, 9110.34it/\n",
"Epoch 259: Train, Loglike in nats: 6.243144: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.259633: 80%|▊| 1024/1280 [00:00<00:00, 8868.73it/\n",
"Epoch 260: Train, Loglike in nats: 6.229244: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.070783: 80%|▊| 1024/1280 [00:00<00:00, 5754.37it/\n",
"Epoch 261: Train, Loglike in nats: 6.223459: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.175782: 80%|▊| 1024/1280 [00:00<00:00, 8954.42it/\n",
"Epoch 262: Train, Loglike in nats: 6.238076: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.192896: 80%|▊| 1024/1280 [00:00<00:00, 8817.41it/\n",
"Epoch 263: Train, Loglike in nats: 6.228003: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.087106: 80%|▊| 1024/1280 [00:00<00:00, 9041.96it/\n",
"Epoch 264: Train, Loglike in nats: 6.239620: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.196321: 80%|▊| 1024/1280 [00:00<00:00, 8469.52it/\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 265: Train, Loglike in nats: 6.240774: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.130541: 80%|▊| 1024/1280 [00:00<00:00, 9059.14it/\n",
"Epoch 266: Train, Loglike in nats: 6.258704: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.134007: 80%|▊| 1024/1280 [00:00<00:00, 8928.99it/\n",
"Epoch 267: Train, Loglike in nats: 6.240042: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.201356: 80%|▊| 1024/1280 [00:00<00:00, 3442.43it/\n",
"Epoch 268: Train, Loglike in nats: 6.225747: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.070212: 80%|▊| 1024/1280 [00:00<00:00, 9378.93it/\n",
"Epoch 269: Train, Loglike in nats: 6.202410: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.229268: 80%|▊| 1024/1280 [00:00<00:00, 8724.09it/\n",
"Epoch 270: Train, Loglike in nats: 6.238284: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.154505: 80%|▊| 1024/1280 [00:00<00:00, 8554.72it/\n",
"Epoch 271: Train, Loglike in nats: 6.255405: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.183676: 80%|▊| 1024/1280 [00:00<00:00, 8894.28it/\n",
"Epoch 272: Train, Loglike in nats: 6.251064: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.109962: 80%|▊| 1024/1280 [00:00<00:00, 8777.35it/\n",
"Epoch 273: Train, Loglike in nats: 6.222325: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.216261: 80%|▊| 1024/1280 [00:00<00:00, 4552.89it/\n",
"Epoch 274: Train, Loglike in nats: 6.230293: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.085872: 80%|▊| 1024/1280 [00:00<00:00, 9135.93it/\n",
"Epoch 275: Train, Loglike in nats: 6.239738: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.084197: 80%|▊| 1024/1280 [00:00<00:00, 8925.68it/\n",
"Epoch 276: Train, Loglike in nats: 6.242204: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.218448: 80%|▊| 1024/1280 [00:00<00:00, 8773.06it/\n",
"Epoch 277: Train, Loglike in nats: 6.265805: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.248243: 80%|▊| 1024/1280 [00:00<00:00, 8622.22it/\n",
"Epoch 278: Train, Loglike in nats: 6.281532: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.252408: 80%|▊| 1024/1280 [00:00<00:00, 8575.13it/\n",
"Epoch 279: Train, Loglike in nats: 6.275621: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.258754: 80%|▊| 1024/1280 [00:00<00:00, 5241.85it/\n",
"Epoch 280: Train, Loglike in nats: 6.264493: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.190695: 80%|▊| 1024/1280 [00:00<00:00, 4368.48it/\n",
"Epoch 281: Train, Loglike in nats: 6.252128: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.139920: 80%|▊| 1024/1280 [00:00<00:00, 4147.97it/\n",
"Epoch 282: Train, Loglike in nats: 6.280600: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.094954: 80%|▊| 1024/1280 [00:00<00:00, 9193.49it/\n",
"Epoch 283: Train, Loglike in nats: 6.248595: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.147238: 80%|▊| 1024/1280 [00:00<00:00, 8943.62it/\n",
"Epoch 284: Train, Loglike in nats: 6.253563: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.219102: 80%|▊| 1024/1280 [00:00<00:00, 8768.69it/\n",
"Epoch 285: Train, Loglike in nats: 6.248441: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.253283: 80%|▊| 1024/1280 [00:00<00:00, 8753.32it/\n",
"Epoch 286: Train, Loglike in nats: 6.274932: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.201911: 80%|▊| 1024/1280 [00:00<00:00, 9119.97it/\n",
"Epoch 287: Train, Loglike in nats: 6.282821: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.179048: 80%|▊| 1024/1280 [00:00<00:00, 9206.63it/\n",
"Epoch 288: Train, Loglike in nats: 6.283505: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.161070: 80%|▊| 1024/1280 [00:00<00:00, 8414.15it/\n",
"Epoch 289: Train, Loglike in nats: 6.253220: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.129789: 80%|▊| 1024/1280 [00:00<00:00, 8541.32it/\n",
"Epoch 290: Train, Loglike in nats: 6.281913: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.177587: 80%|▊| 1024/1280 [00:00<00:00, 3787.79it/\n",
"Epoch 291: Train, Loglike in nats: 6.260824: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.169320: 80%|▊| 1024/1280 [00:00<00:00, 3484.71it/\n",
"Epoch 292: Train, Loglike in nats: 6.284425: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.072508: 80%|▊| 1024/1280 [00:00<00:00, 9032.30it/\n",
"Epoch 293: Train, Loglike in nats: 6.253657: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.129054: 80%|▊| 1024/1280 [00:00<00:00, 8892.14it/\n",
"Epoch 294: Train, Loglike in nats: 6.269070: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.288439: 80%|▊| 1024/1280 [00:00<00:00, 9185.50it/\n",
"Epoch 295: Train, Loglike in nats: 6.219231: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.162266: 80%|▊| 1024/1280 [00:00<00:00, 9304.28it/\n",
"Epoch 296: Train, Loglike in nats: 6.251959: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.235776: 80%|▊| 1024/1280 [00:00<00:00, 9260.95it/\n",
"Epoch 297: Train, Loglike in nats: 6.250069: 98%|▉| 11264/11520 [00:02<00:0\n",
"- Val, Loglike in nats: 6.188086: 80%|▊| 1024/1280 [00:00<00:00, 9132.09it/\n",
"Epoch 298: Train, Loglike in nats: 6.266000: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.193928: 80%|▊| 1024/1280 [00:00<00:00, 8817.64it/\n",
"Epoch 299: Train, Loglike in nats: 6.284518: 98%|▉| 11264/11520 [00:03<00:0\n",
"- Val, Loglike in nats: 6.120339: 80%|▊| 1024/1280 [00:00<00:00, 5651.68it/\n"
]
}
],
"source": [
"engine.fit(\n",
" n_sims=12800,\n",
" n_rounds=1,\n",
" n_epochs=300,\n",
" batch_size=512,\n",
" lr=0.001,\n",
" early_stop_patience=250,\n",
" noise=noise # this can also be an array if fixed noise\n",
")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now we can apply the trained model to derive the model posterior for any sine curve. Nbi uses importance sampling to deliver asymptotically exact results if you specify x_err in engine.predict()."
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {
"scrolled": true
},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Samples outside prior N = 1512\n",
"surrogate posterior\n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Generating simulations\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"1129it [00:00, 2181.67it/s]\n",
"WARNING:root:Too few points to create valid contours\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Effective Sample Size = 2656.2\n",
"Sampling efficiency = 20.8%\n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"plt.figure(figsize=(5,3))\n",
"np.random.seed(4)\n",
"\n",
"# draw random parameter from prior\n",
"y_true = [var.rvs(1)[0] for var in priors]\n",
"x_err = 0.1\n",
"x_obs = sine(y_true) + np.random.normal(size=50) * x_err\n",
"plt.errorbar(t, x_obs, yerr=x_err, fmt='.')\n",
"plt.show()\n",
"\n",
"y_pred, weights = engine.predict(x_obs, x_err=np.array([0.1]*50), y_true=y_true, n_samples=12800, corner=True, corner_reweight=True,seed=0)\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Best model checkpoint is saved to: test2/0/299.pth\n"
]
}
],
"source": [
"# You can now easily distribute the trained model by sending the best checkpoint:\n",
"best_model = engine.best_params\n",
"print('Best model checkpoint is saved to:', best_model)"
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Samples outside prior N = 1512\n",
"Generating simulations\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"1129it [00:00, 2437.05it/s]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Effective Sample Size = 2656.2\n",
"Sampling efficiency = 20.8%\n"
]
}
],
"source": [
"# Others can easily load the trained model for inference using the state_dict parameter\n",
"# Since the state_dict contains the network hyperparameters, no need to supply flow and featurizer dictionaries\n",
"engine_loaded = nbi.NBI(\n",
" state_dict=best_model,\n",
" simulator=sine,\n",
" priors=priors,\n",
" labels=labels,\n",
" device='cpu',\n",
" path='test2',\n",
" n_jobs=10\n",
")\n",
"y_pred2, weights2 = engine_loaded.predict(x_obs, x_err=np.array([0.1]*50), y_true=y_true, n_samples=12800, seed=0)"
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"True"
]
},
"execution_count": 19,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# the loaded model produces the identical result\n",
"np.allclose(y_pred2,y_pred)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Use case 2: Sequential Neural Posterior Estimation (SNPE)\n",
"Let's say we only want to fit the specific sine curve above.\n",
"\n",
"In this case, we can specify n_rounds > 1 to train the model over multiple rounds to reduce the number of training samples required. The basic idea is to generate more training samples near the final posterior (as opposed to the broad prior)."
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {
"scrolled": true
},
"outputs": [],
"source": [
"# hyperparameters remain the same as ANPE\n",
"flow = {\n",
" 'n_dims': 3, # dimension of parameter space\n",
" 'flow_hidden': 128,\n",
" 'num_blocks': 5,\n",
" 'n_mog':4\n",
"}\n",
"\n",
"# the NBI package provides the \"ResNet-GRU\" network as the default\n",
"# featurizer network for sequential data\n",
"featurizer = {\n",
" 'type': 'resnet-gru',\n",
" 'norm': 'weight_norm',\n",
" 'dim_in': 1,\n",
" 'dim_out': 128,\n",
" 'dim_conv_max': 256,\n",
" 'depth': 3\n",
"}\n",
"\n",
"# initializtion is teh same with ANPE\n",
"engine = nbi.NBI(\n",
" flow=flow,\n",
" featurizer=featurizer,\n",
" simulator=sine,\n",
" priors=priors,\n",
" labels=labels,\n",
" path='test',\n",
" device='cpu',\n",
" n_jobs=10,\n",
")"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {
"scrolled": true
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Auto learning rate to min_lr = 8.333333333333333e-07\n",
"Generating simulations\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"256it [00:00, 2883.50it/s]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Effective sample size for current/all rounds 0.0/0.0\n",
"\n",
"---------------------- Round: 0 ----------------------\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 0: Train, Loglike in nats: -3.268526: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: -2.459209: 100%|█| 256/256 [00:00<00:00, 1261.47it/s\n",
"Epoch 1: Train, Loglike in nats: -2.159331: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: -1.682208: 100%|█| 256/256 [00:00<00:00, 5980.08it/s\n",
"Epoch 2: Train, Loglike in nats: -1.848469: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: -1.720842: 100%|█| 256/256 [00:00<00:00, 5995.34it/s\n",
"Epoch 3: Train, Loglike in nats: -1.418813: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: -1.714325: 100%|█| 256/256 [00:00<00:00, 2912.96it/s\n",
"Epoch 4: Train, Loglike in nats: -1.574528: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: -1.149077: 100%|█| 256/256 [00:00<00:00, 6075.26it/s\n",
"Epoch 5: Train, Loglike in nats: -1.042438: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: -0.894966: 100%|█| 256/256 [00:00<00:00, 6068.22it/s\n",
"Epoch 6: Train, Loglike in nats: -1.492501: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: -0.985928: 100%|█| 256/256 [00:00<00:00, 6169.26it/s\n",
"Epoch 7: Train, Loglike in nats: -0.870418: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: -1.858480: 100%|█| 256/256 [00:00<00:00, 6307.59it/s\n",
"Epoch 8: Train, Loglike in nats: -1.463720: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: -0.996620: 100%|█| 256/256 [00:00<00:00, 4096.20it/s\n",
"Epoch 9: Train, Loglike in nats: -0.352051: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: -0.155534: 100%|█| 256/256 [00:00<00:00, 6595.59it/s\n",
"Epoch 10: Train, Loglike in nats: 0.064110: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: -0.162349: 100%|█| 256/256 [00:00<00:00, 6389.30it/s\n",
"Epoch 11: Train, Loglike in nats: 0.033803: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 0.437606: 100%|█| 256/256 [00:00<00:00, 6336.63it/s]\n",
"Epoch 12: Train, Loglike in nats: 0.734700: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 1.128706: 100%|█| 256/256 [00:00<00:00, 6170.47it/s]\n",
"Epoch 13: Train, Loglike in nats: 0.713297: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 0.360700: 100%|█| 256/256 [00:00<00:00, 4876.74it/s]\n",
"Epoch 14: Train, Loglike in nats: 0.514453: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 0.425054: 100%|█| 256/256 [00:00<00:00, 6240.54it/s]\n",
"Epoch 15: Train, Loglike in nats: 0.929215: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 1.359069: 100%|█| 256/256 [00:00<00:00, 5791.86it/s]\n",
"Epoch 16: Train, Loglike in nats: 1.220240: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 1.527185: 100%|█| 256/256 [00:00<00:00, 6140.12it/s]\n",
"Epoch 17: Train, Loglike in nats: 0.711048: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 1.133904: 100%|█| 256/256 [00:00<00:00, 6629.71it/s]\n",
"Epoch 18: Train, Loglike in nats: 1.415741: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 1.513332: 100%|█| 256/256 [00:00<00:00, 1634.17it/s]\n",
"Epoch 19: Train, Loglike in nats: 1.074183: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 0.832191: 100%|█| 256/256 [00:00<00:00, 6499.61it/s]\n",
"Epoch 20: Train, Loglike in nats: 1.293864: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 1.259685: 100%|█| 256/256 [00:00<00:00, 6307.11it/s]\n",
"Epoch 21: Train, Loglike in nats: 1.439562: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 1.355938: 100%|█| 256/256 [00:00<00:00, 6386.91it/s]\n",
"Epoch 22: Train, Loglike in nats: 1.374695: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 0.132194: 100%|█| 256/256 [00:00<00:00, 6079.77it/s]\n",
"Epoch 23: Train, Loglike in nats: 0.716420: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 1.554517: 100%|█| 256/256 [00:00<00:00, 4842.89it/s]\n",
"Epoch 24: Train, Loglike in nats: 1.595632: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 1.770648: 100%|█| 256/256 [00:00<00:00, 6380.08it/s]\n",
"Epoch 25: Train, Loglike in nats: 1.967934: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 1.051805: 100%|█| 256/256 [00:00<00:00, 6309.63it/s]\n",
"Epoch 26: Train, Loglike in nats: -0.527291: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 0.457400: 100%|█| 256/256 [00:00<00:00, 6251.52it/s]\n",
"Epoch 27: Train, Loglike in nats: 1.125868: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 1.251953: 100%|█| 256/256 [00:00<00:00, 6726.02it/s]\n",
"Epoch 28: Train, Loglike in nats: 1.543339: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 1.801289: 100%|█| 256/256 [00:00<00:00, 6317.58it/s]\n",
"Epoch 29: Train, Loglike in nats: 1.911407: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 1.962026: 100%|█| 256/256 [00:00<00:00, 6368.61it/s]\n",
"Epoch 30: Train, Loglike in nats: 0.037019: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 0.268609: 100%|█| 256/256 [00:00<00:00, 6307.45it/s]\n",
"Epoch 31: Train, Loglike in nats: 0.250173: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 0.763747: 100%|█| 256/256 [00:00<00:00, 6089.01it/s]\n",
"Epoch 32: Train, Loglike in nats: 1.401325: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 1.665264: 100%|█| 256/256 [00:00<00:00, 6457.98it/s]\n",
"Epoch 33: Train, Loglike in nats: 1.984000: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.180026: 100%|█| 256/256 [00:00<00:00, 4943.61it/s]\n",
"Epoch 34: Train, Loglike in nats: 2.048396: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 1.844047: 100%|█| 256/256 [00:00<00:00, 6295.46it/s]\n",
"Epoch 35: Train, Loglike in nats: 1.620334: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 1.313894: 100%|█| 256/256 [00:00<00:00, 6130.55it/s]\n",
"Epoch 36: Train, Loglike in nats: 1.512009: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 1.536890: 100%|█| 256/256 [00:00<00:00, 6033.04it/s]\n",
"Epoch 37: Train, Loglike in nats: 1.894795: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.025642: 100%|█| 256/256 [00:00<00:00, 5940.91it/s]\n",
"Epoch 38: Train, Loglike in nats: 2.259384: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.483783: 100%|█| 256/256 [00:00<00:00, 1330.48it/s]\n",
"Epoch 39: Train, Loglike in nats: 2.342173: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.026701: 100%|█| 256/256 [00:00<00:00, 6246.50it/s]\n",
"Epoch 40: Train, Loglike in nats: 1.204104: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 0.647072: 100%|█| 256/256 [00:00<00:00, 5923.05it/s]\n",
"Epoch 41: Train, Loglike in nats: 1.778243: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 1.732202: 100%|█| 256/256 [00:00<00:00, 6028.94it/s]\n",
"Epoch 42: Train, Loglike in nats: 2.145499: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.264459: 100%|█| 256/256 [00:00<00:00, 6059.05it/s]\n",
"Epoch 43: Train, Loglike in nats: 2.471862: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 2.704911: 100%|█| 256/256 [00:00<00:00, 5937.75it/s]\n",
"Epoch 44: Train, Loglike in nats: 2.117804: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.275875: 100%|█| 256/256 [00:00<00:00, 6285.00it/s]\n",
"Epoch 45: Train, Loglike in nats: 0.706343: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 1.236873: 100%|█| 256/256 [00:00<00:00, 5896.28it/s]\n",
"Epoch 46: Train, Loglike in nats: 2.225131: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 2.611018: 100%|█| 256/256 [00:00<00:00, 5948.94it/s]\n",
"Epoch 47: Train, Loglike in nats: 2.906169: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.025538: 100%|█| 256/256 [00:00<00:00, 6114.46it/s]\n",
"Epoch 48: Train, Loglike in nats: 3.254640: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 3.117925: 100%|█| 256/256 [00:00<00:00, 6070.52it/s]\n",
"Epoch 49: Train, Loglike in nats: 2.208368: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 1.511522: 100%|█| 256/256 [00:00<00:00, 6092.50it/s]\n",
"Epoch 50: Train, Loglike in nats: 0.267897: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.196132: 100%|█| 256/256 [00:00<00:00, 1537.70it/s]\n",
"Epoch 51: Train, Loglike in nats: 2.917408: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.254388: 100%|█| 256/256 [00:00<00:00, 6633.85it/s]\n",
"Epoch 52: Train, Loglike in nats: 3.546430: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.643027: 100%|█| 256/256 [00:00<00:00, 6455.34it/s]\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 53: Train, Loglike in nats: 3.714629: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 3.687727: 100%|█| 256/256 [00:00<00:00, 6390.10it/s]\n",
"Epoch 54: Train, Loglike in nats: 3.723377: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.727478: 100%|█| 256/256 [00:00<00:00, 6687.23it/s]\n",
"Epoch 55: Train, Loglike in nats: 3.166823: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 1.733759: 100%|█| 256/256 [00:00<00:00, 6281.76it/s]\n",
"Epoch 56: Train, Loglike in nats: 3.368278: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.668584: 100%|█| 256/256 [00:00<00:00, 6511.99it/s]\n",
"Epoch 57: Train, Loglike in nats: 3.880386: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.082314: 100%|█| 256/256 [00:00<00:00, 6652.96it/s]\n",
"Epoch 58: Train, Loglike in nats: 4.114445: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 4.050641: 100%|█| 256/256 [00:00<00:00, 6616.19it/s]\n",
"Epoch 59: Train, Loglike in nats: 3.763349: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.134168: 100%|█| 256/256 [00:00<00:00, 6589.48it/s]\n",
"Epoch 60: Train, Loglike in nats: 3.282981: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 3.616098: 100%|█| 256/256 [00:00<00:00, 6412.85it/s]\n",
"Epoch 61: Train, Loglike in nats: 4.023603: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.176076: 100%|█| 256/256 [00:00<00:00, 6790.68it/s]\n",
"Epoch 62: Train, Loglike in nats: 4.472724: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.489458: 100%|█| 256/256 [00:00<00:00, 6416.53it/s]\n",
"Epoch 63: Train, Loglike in nats: 4.359788: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 4.409226: 100%|█| 256/256 [00:00<00:00, 6229.75it/s]\n",
"Epoch 64: Train, Loglike in nats: 3.739900: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.325349: 100%|█| 256/256 [00:00<00:00, 6471.64it/s]\n",
"Epoch 65: Train, Loglike in nats: 4.625143: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 4.595809: 100%|█| 256/256 [00:00<00:00, 6779.70it/s]\n",
"Epoch 66: Train, Loglike in nats: 4.712687: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.558172: 100%|█| 256/256 [00:00<00:00, 6508.67it/s]\n",
"Epoch 67: Train, Loglike in nats: 4.634473: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.542986: 100%|█| 256/256 [00:00<00:00, 6633.32it/s]\n",
"Epoch 68: Train, Loglike in nats: 4.698652: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 4.720885: 100%|█| 256/256 [00:00<00:00, 6428.78it/s]\n",
"Epoch 69: Train, Loglike in nats: 3.947856: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.944192: 100%|█| 256/256 [00:00<00:00, 6437.80it/s]\n",
"Epoch 70: Train, Loglike in nats: 4.448578: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 4.082244: 100%|█| 256/256 [00:00<00:00, 6629.06it/s]\n",
"Epoch 71: Train, Loglike in nats: 4.788922: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.949507: 100%|█| 256/256 [00:00<00:00, 6749.61it/s]\n",
"Epoch 72: Train, Loglike in nats: 4.978190: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.942903: 100%|█| 256/256 [00:00<00:00, 1460.23it/s]\n",
"Epoch 73: Train, Loglike in nats: 4.892810: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 4.318020: 100%|█| 256/256 [00:00<00:00, 6285.96it/s]\n",
"Epoch 74: Train, Loglike in nats: 2.601273: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.881264: 100%|█| 256/256 [00:00<00:00, 6373.83it/s]\n",
"Epoch 75: Train, Loglike in nats: 4.411237: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 4.428556: 100%|█| 256/256 [00:00<00:00, 6526.00it/s]\n",
"Epoch 76: Train, Loglike in nats: 4.798160: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.791045: 100%|█| 256/256 [00:00<00:00, 6663.53it/s]\n",
"Epoch 77: Train, Loglike in nats: 5.071071: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 4.851388: 100%|█| 256/256 [00:00<00:00, 6373.91it/s]\n",
"Epoch 78: Train, Loglike in nats: 5.101960: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.549327: 100%|█| 256/256 [00:00<00:00, 5811.86it/s]\n",
"Epoch 79: Train, Loglike in nats: 5.037960: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.690504: 100%|█| 256/256 [00:00<00:00, 5669.21it/s]\n",
"Epoch 80: Train, Loglike in nats: 4.213198: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 4.821634: 100%|█| 256/256 [00:00<00:00, 6268.63it/s]\n",
"Epoch 81: Train, Loglike in nats: 4.719924: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 5.144917: 100%|█| 256/256 [00:00<00:00, 6301.52it/s]\n",
"Epoch 82: Train, Loglike in nats: 5.324505: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 5.265930: 100%|█| 256/256 [00:00<00:00, 6158.44it/s]\n",
"Epoch 83: Train, Loglike in nats: 4.917983: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.838337: 100%|█| 256/256 [00:00<00:00, 6288.35it/s]\n",
"Epoch 84: Train, Loglike in nats: 5.263070: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 5.360596: 100%|█| 256/256 [00:00<00:00, 6268.34it/s]\n",
"Epoch 85: Train, Loglike in nats: 5.480107: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 5.295392: 100%|█| 256/256 [00:00<00:00, 5964.60it/s]\n",
"Epoch 86: Train, Loglike in nats: 5.351205: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.789269: 100%|█| 256/256 [00:00<00:00, 5830.90it/s]\n",
"Epoch 87: Train, Loglike in nats: 4.893774: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 4.206510: 100%|█| 256/256 [00:00<00:00, 6414.73it/s]\n",
"Epoch 88: Train, Loglike in nats: 4.854108: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.709023: 100%|█| 256/256 [00:00<00:00, 6088.73it/s]\n",
"Epoch 89: Train, Loglike in nats: 5.183822: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 5.275519: 100%|█| 256/256 [00:00<00:00, 5933.06it/s]\n",
"Epoch 90: Train, Loglike in nats: 5.462302: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 5.404655: 100%|█| 256/256 [00:00<00:00, 6086.83it/s]\n",
"Epoch 91: Train, Loglike in nats: 5.478829: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 5.444710: 100%|█| 256/256 [00:00<00:00, 6074.85it/s]\n",
"Epoch 92: Train, Loglike in nats: 5.184221: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 4.115625: 100%|█| 256/256 [00:00<00:00, 6139.25it/s]\n",
"Epoch 93: Train, Loglike in nats: 5.066727: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 5.575367: 100%|█| 256/256 [00:00<00:00, 6355.83it/s]\n",
"Epoch 94: Train, Loglike in nats: 5.498914: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 5.554475: 100%|█| 256/256 [00:00<00:00, 6413.92it/s]\n",
"Epoch 95: Train, Loglike in nats: 5.521467: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 5.059930: 100%|█| 256/256 [00:00<00:00, 6179.84it/s]\n",
"Epoch 96: Train, Loglike in nats: 4.965427: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.261661: 100%|█| 256/256 [00:00<00:00, 6291.78it/s]\n",
"Epoch 97: Train, Loglike in nats: 4.968622: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 5.169197: 100%|█| 256/256 [00:00<00:00, 6325.69it/s]\n",
"Epoch 98: Train, Loglike in nats: 5.541847: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 5.742117: 100%|█| 256/256 [00:00<00:00, 6243.30it/s]\n",
"Epoch 99: Train, Loglike in nats: 5.661441: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 5.223972: 100%|█| 256/256 [00:00<00:00, 5850.21it/s]\n",
"Epoch 100: Train, Loglike in nats: 4.972398: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.572842: 100%|█| 256/256 [00:00<00:00, 6295.65it/s]\n",
"Epoch 101: Train, Loglike in nats: 5.779009: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 5.781581: 100%|█| 256/256 [00:00<00:00, 6259.79it/s]\n",
"Epoch 102: Train, Loglike in nats: 5.850139: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.807649: 100%|█| 256/256 [00:00<00:00, 6041.42it/s]\n",
"Epoch 103: Train, Loglike in nats: 5.958394: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.970178: 100%|█| 256/256 [00:00<00:00, 6045.71it/s]\n",
"Epoch 104: Train, Loglike in nats: 5.922030: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 5.640943: 100%|█| 256/256 [00:00<00:00, 6191.64it/s]\n",
"Epoch 105: Train, Loglike in nats: 5.830493: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.658373: 100%|█| 256/256 [00:00<00:00, 6261.18it/s]\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 106: Train, Loglike in nats: 5.917760: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 5.658114: 100%|█| 256/256 [00:00<00:00, 5785.68it/s]\n",
"Epoch 107: Train, Loglike in nats: 5.942582: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.720334: 100%|█| 256/256 [00:00<00:00, 6181.20it/s]\n",
"Epoch 108: Train, Loglike in nats: 6.048959: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.026139: 100%|█| 256/256 [00:00<00:00, 5834.19it/s]\n",
"Epoch 109: Train, Loglike in nats: 6.068287: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.803365: 100%|█| 256/256 [00:00<00:00, 6333.15it/s]\n",
"Epoch 110: Train, Loglike in nats: 5.634220: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.819095: 100%|█| 256/256 [00:00<00:00, 6168.09it/s]\n",
"Epoch 111: Train, Loglike in nats: 5.769787: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 5.741638: 100%|█| 256/256 [00:00<00:00, 6521.32it/s]\n",
"Epoch 112: Train, Loglike in nats: 5.882873: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.010417: 100%|█| 256/256 [00:00<00:00, 6583.86it/s]\n",
"Epoch 113: Train, Loglike in nats: 6.009948: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 5.562791: 100%|█| 256/256 [00:00<00:00, 6101.15it/s]\n",
"Epoch 114: Train, Loglike in nats: 6.008356: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.888279: 100%|█| 256/256 [00:00<00:00, 6160.84it/s]\n",
"Epoch 115: Train, Loglike in nats: 6.052705: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.892685: 100%|█| 256/256 [00:00<00:00, 6297.94it/s]\n",
"Epoch 116: Train, Loglike in nats: 6.042388: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 5.901352: 100%|█| 256/256 [00:00<00:00, 6343.74it/s]\n",
"Epoch 117: Train, Loglike in nats: 6.034147: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.474616: 100%|█| 256/256 [00:00<00:00, 6370.50it/s]\n",
"Epoch 118: Train, Loglike in nats: 5.642377: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 5.856933: 100%|█| 256/256 [00:00<00:00, 6289.01it/s]\n",
"Epoch 119: Train, Loglike in nats: 5.894026: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.044425: 100%|█| 256/256 [00:00<00:00, 6137.91it/s]\n",
"Epoch 120: Train, Loglike in nats: 6.172974: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.079485: 100%|█| 256/256 [00:00<00:00, 1686.47it/s]\n",
"Epoch 121: Train, Loglike in nats: 6.026621: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.917090: 100%|█| 256/256 [00:00<00:00, 6058.33it/s]\n",
"Epoch 122: Train, Loglike in nats: 6.219800: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.005662: 100%|█| 256/256 [00:00<00:00, 6077.43it/s]\n",
"Epoch 123: Train, Loglike in nats: 6.309060: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.213336: 100%|█| 256/256 [00:00<00:00, 6058.02it/s]\n",
"Epoch 124: Train, Loglike in nats: 6.197256: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.882795: 100%|█| 256/256 [00:00<00:00, 6226.32it/s]\n",
"Epoch 125: Train, Loglike in nats: 6.096331: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.173139: 100%|█| 256/256 [00:00<00:00, 6019.14it/s]\n",
"Epoch 126: Train, Loglike in nats: 5.885933: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 5.818267: 100%|█| 256/256 [00:00<00:00, 6200.04it/s]\n",
"Epoch 127: Train, Loglike in nats: 5.378657: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.571177: 100%|█| 256/256 [00:00<00:00, 5992.53it/s]\n",
"Epoch 128: Train, Loglike in nats: 5.645857: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 5.298625: 100%|█| 256/256 [00:00<00:00, 6510.21it/s]\n",
"Epoch 129: Train, Loglike in nats: 5.837600: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.992451: 100%|█| 256/256 [00:00<00:00, 6344.00it/s]\n",
"Epoch 130: Train, Loglike in nats: 6.119033: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.149038: 100%|█| 256/256 [00:00<00:00, 6883.53it/s]\n",
"Epoch 131: Train, Loglike in nats: 6.245521: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.930585: 100%|█| 256/256 [00:00<00:00, 6699.16it/s]\n",
"Epoch 132: Train, Loglike in nats: 6.261621: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.998144: 100%|█| 256/256 [00:00<00:00, 6566.02it/s]\n",
"Epoch 133: Train, Loglike in nats: 6.341798: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.346358: 100%|█| 256/256 [00:00<00:00, 6596.60it/s]\n",
"Epoch 134: Train, Loglike in nats: 6.401124: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.237489: 100%|█| 256/256 [00:00<00:00, 6981.37it/s]\n",
"Epoch 135: Train, Loglike in nats: 6.132567: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.047589: 100%|█| 256/256 [00:00<00:00, 6708.96it/s]\n",
"Epoch 136: Train, Loglike in nats: 6.230725: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.994621: 100%|█| 256/256 [00:00<00:00, 6747.62it/s]\n",
"Epoch 137: Train, Loglike in nats: 6.246450: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.284387: 100%|█| 256/256 [00:00<00:00, 6725.18it/s]\n",
"Epoch 138: Train, Loglike in nats: 6.182283: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.257266: 100%|█| 256/256 [00:00<00:00, 6654.86it/s]\n",
"Epoch 139: Train, Loglike in nats: 6.378484: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.369561: 100%|█| 256/256 [00:00<00:00, 6337.90it/s]\n",
"Epoch 140: Train, Loglike in nats: 6.480492: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.383032: 100%|█| 256/256 [00:00<00:00, 6444.64it/s]\n",
"Epoch 141: Train, Loglike in nats: 6.489468: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.293998: 100%|█| 256/256 [00:00<00:00, 6710.22it/s]\n",
"Epoch 142: Train, Loglike in nats: 6.533051: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.425118: 100%|█| 256/256 [00:00<00:00, 6697.87it/s]\n",
"Epoch 143: Train, Loglike in nats: 6.455296: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.181000: 100%|█| 256/256 [00:00<00:00, 6532.11it/s]\n",
"Epoch 144: Train, Loglike in nats: 6.200677: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.747252: 100%|█| 256/256 [00:00<00:00, 6332.56it/s]\n",
"Epoch 145: Train, Loglike in nats: 6.105745: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.781116: 100%|█| 256/256 [00:00<00:00, 6397.76it/s]\n",
"Epoch 146: Train, Loglike in nats: 6.144907: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.042254: 100%|█| 256/256 [00:00<00:00, 6527.27it/s]\n",
"Epoch 147: Train, Loglike in nats: 6.060318: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.090507: 100%|█| 256/256 [00:00<00:00, 6646.91it/s]\n",
"Epoch 148: Train, Loglike in nats: 6.419754: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.284282: 100%|█| 256/256 [00:00<00:00, 6610.04it/s]\n",
"Epoch 149: Train, Loglike in nats: 6.500065: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.493800: 100%|█| 256/256 [00:00<00:00, 6348.92it/s]\n",
"Epoch 150: Train, Loglike in nats: 6.546723: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.087693: 100%|█| 256/256 [00:00<00:00, 6234.89it/s]\n",
"Epoch 151: Train, Loglike in nats: 6.468393: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.122276: 100%|█| 256/256 [00:00<00:00, 6540.94it/s]\n",
"Epoch 152: Train, Loglike in nats: 6.583395: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.481585: 100%|█| 256/256 [00:00<00:00, 6110.63it/s]\n",
"Epoch 153: Train, Loglike in nats: 6.571469: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.435233: 100%|█| 256/256 [00:00<00:00, 6414.73it/s]\n",
"Epoch 154: Train, Loglike in nats: 6.534500: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.312248: 100%|█| 256/256 [00:00<00:00, 6529.25it/s]\n",
"Epoch 155: Train, Loglike in nats: 6.627038: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.548282: 100%|█| 256/256 [00:00<00:00, 6447.08it/s]\n",
"Epoch 156: Train, Loglike in nats: 6.588944: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.499773: 100%|█| 256/256 [00:00<00:00, 6890.43it/s]\n",
"Epoch 157: Train, Loglike in nats: 6.562883: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.519076: 100%|█| 256/256 [00:00<00:00, 6318.99it/s]\n",
"Epoch 158: Train, Loglike in nats: 6.608741: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.425825: 100%|█| 256/256 [00:00<00:00, 6462.37it/s]\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 159: Train, Loglike in nats: 6.558365: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.334709: 100%|█| 256/256 [00:00<00:00, 6423.78it/s]\n",
"Epoch 160: Train, Loglike in nats: 6.552215: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.399486: 100%|█| 256/256 [00:00<00:00, 6294.91it/s]\n",
"Epoch 161: Train, Loglike in nats: 6.655611: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.340299: 100%|█| 256/256 [00:00<00:00, 6637.46it/s]\n",
"Epoch 162: Train, Loglike in nats: 6.651125: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.579459: 100%|█| 256/256 [00:00<00:00, 6802.21it/s]\n",
"Epoch 163: Train, Loglike in nats: 6.688755: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.509596: 100%|█| 256/256 [00:00<00:00, 6130.66it/s]\n",
"Epoch 164: Train, Loglike in nats: 6.641705: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.578617: 100%|█| 256/256 [00:00<00:00, 6181.59it/s]\n",
"Epoch 165: Train, Loglike in nats: 6.652033: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.501852: 100%|█| 256/256 [00:00<00:00, 6098.69it/s]\n",
"Epoch 166: Train, Loglike in nats: 6.695315: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.601800: 100%|█| 256/256 [00:00<00:00, 5699.93it/s]\n",
"Epoch 167: Train, Loglike in nats: 6.740462: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.544092: 100%|█| 256/256 [00:00<00:00, 5968.32it/s]\n",
"Epoch 168: Train, Loglike in nats: 6.769489: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.610115: 100%|█| 256/256 [00:00<00:00, 6250.46it/s]\n",
"Epoch 169: Train, Loglike in nats: 6.754818: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.583900: 100%|█| 256/256 [00:00<00:00, 6161.55it/s]\n",
"Epoch 170: Train, Loglike in nats: 6.728368: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.433847: 100%|█| 256/256 [00:00<00:00, 6122.79it/s]\n",
"Epoch 171: Train, Loglike in nats: 5.814623: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.100624: 100%|█| 256/256 [00:00<00:00, 6103.06it/s]\n",
"Epoch 172: Train, Loglike in nats: 6.252972: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.292130: 100%|█| 256/256 [00:00<00:00, 1814.75it/s]\n",
"Epoch 173: Train, Loglike in nats: 6.469107: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.249124: 100%|█| 256/256 [00:00<00:00, 6305.74it/s]\n",
"Epoch 174: Train, Loglike in nats: 6.617064: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.365734: 100%|█| 256/256 [00:00<00:00, 6289.16it/s]\n",
"Epoch 175: Train, Loglike in nats: 6.779203: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.570490: 100%|█| 256/256 [00:00<00:00, 6326.32it/s]\n",
"Epoch 176: Train, Loglike in nats: 6.786192: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.652312: 100%|█| 256/256 [00:00<00:00, 6574.18it/s]\n",
"Epoch 177: Train, Loglike in nats: 6.838615: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.647413: 100%|█| 256/256 [00:00<00:00, 6475.46it/s]\n",
"Epoch 178: Train, Loglike in nats: 6.867536: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.653259: 100%|█| 256/256 [00:00<00:00, 6538.95it/s]\n",
"Epoch 179: Train, Loglike in nats: 6.817409: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.631561: 100%|█| 256/256 [00:00<00:00, 6505.91it/s]\n",
"Epoch 180: Train, Loglike in nats: 6.831194: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.713398: 100%|█| 256/256 [00:00<00:00, 6680.78it/s]\n",
"Epoch 181: Train, Loglike in nats: 6.880255: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.747829: 100%|█| 256/256 [00:00<00:00, 6326.29it/s]\n",
"Epoch 182: Train, Loglike in nats: 6.865285: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.633148: 100%|█| 256/256 [00:00<00:00, 6515.98it/s]\n",
"Epoch 183: Train, Loglike in nats: 6.836265: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.745353: 100%|█| 256/256 [00:00<00:00, 6167.46it/s]\n",
"Epoch 184: Train, Loglike in nats: 6.809363: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.620671: 100%|█| 256/256 [00:00<00:00, 6354.89it/s]\n",
"Epoch 185: Train, Loglike in nats: 6.810407: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.665987: 100%|█| 256/256 [00:00<00:00, 5770.70it/s]\n",
"Epoch 186: Train, Loglike in nats: 6.903464: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.718751: 100%|█| 256/256 [00:00<00:00, 6524.77it/s]\n",
"Epoch 187: Train, Loglike in nats: 6.888197: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.679201: 100%|█| 256/256 [00:00<00:00, 3655.99it/s]\n",
"Epoch 188: Train, Loglike in nats: 6.892800: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.746929: 100%|█| 256/256 [00:00<00:00, 6605.61it/s]\n",
"Epoch 189: Train, Loglike in nats: 6.929261: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.784347: 100%|█| 256/256 [00:00<00:00, 4970.87it/s]\n",
"Epoch 190: Train, Loglike in nats: 6.902624: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.666506: 100%|█| 256/256 [00:00<00:00, 6124.50it/s]\n",
"Epoch 191: Train, Loglike in nats: 6.884923: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.703870: 100%|█| 256/256 [00:00<00:00, 5761.22it/s]\n",
"Epoch 192: Train, Loglike in nats: 6.912938: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.814867: 100%|█| 256/256 [00:00<00:00, 5764.22it/s]\n",
"Epoch 193: Train, Loglike in nats: 6.906339: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.639987: 100%|█| 256/256 [00:00<00:00, 5835.52it/s]\n",
"Epoch 194: Train, Loglike in nats: 6.959879: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.855103: 100%|█| 256/256 [00:00<00:00, 6027.38it/s]\n",
"Epoch 195: Train, Loglike in nats: 6.985030: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.678816: 100%|█| 256/256 [00:00<00:00, 6013.47it/s]\n",
"Epoch 196: Train, Loglike in nats: 6.999685: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.808109: 100%|█| 256/256 [00:00<00:00, 5925.69it/s]\n",
"Epoch 197: Train, Loglike in nats: 7.021703: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.759712: 100%|█| 256/256 [00:00<00:00, 5961.49it/s]\n",
"Epoch 198: Train, Loglike in nats: 7.042663: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.783630: 100%|█| 256/256 [00:00<00:00, 6524.81it/s]\n",
"Epoch 199: Train, Loglike in nats: 6.993849: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.784029: 100%|█| 256/256 [00:00<00:00, 5300.52it/s]\n",
"Epoch 200: Train, Loglike in nats: 7.002424: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.748533: 100%|█| 256/256 [00:00<00:00, 6204.70it/s]\n",
"Epoch 201: Train, Loglike in nats: 7.012407: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.860180: 100%|█| 256/256 [00:00<00:00, 3633.31it/s]\n",
"Epoch 202: Train, Loglike in nats: 6.953123: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.838411: 100%|█| 256/256 [00:00<00:00, 6253.70it/s]\n",
"Epoch 203: Train, Loglike in nats: 7.038790: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.843570: 100%|█| 256/256 [00:00<00:00, 2620.69it/s]\n",
"Epoch 204: Train, Loglike in nats: 7.013794: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.860623: 100%|█| 256/256 [00:00<00:00, 5836.50it/s]\n",
"Epoch 205: Train, Loglike in nats: 7.036593: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.772624: 100%|█| 256/256 [00:00<00:00, 6114.60it/s]\n",
"Epoch 206: Train, Loglike in nats: 7.053233: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.794318: 100%|█| 256/256 [00:00<00:00, 6186.90it/s]\n",
"Epoch 207: Train, Loglike in nats: 7.078134: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.932597: 100%|█| 256/256 [00:00<00:00, 6422.13it/s]\n",
"Epoch 208: Train, Loglike in nats: 7.057461: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.850798: 100%|█| 256/256 [00:00<00:00, 6722.15it/s]\n",
"Epoch 209: Train, Loglike in nats: 7.077501: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.900552: 100%|█| 256/256 [00:00<00:00, 5401.88it/s]\n",
"Epoch 210: Train, Loglike in nats: 7.080196: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.857109: 100%|█| 256/256 [00:00<00:00, 5633.93it/s]\n",
"Epoch 211: Train, Loglike in nats: 7.114897: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.912014: 100%|█| 256/256 [00:00<00:00, 4374.14it/s]\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 212: Train, Loglike in nats: 7.128048: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.849624: 100%|█| 256/256 [00:00<00:00, 4464.29it/s]\n",
"Epoch 213: Train, Loglike in nats: 7.143510: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.848959: 100%|█| 256/256 [00:00<00:00, 3617.30it/s]\n",
"Epoch 214: Train, Loglike in nats: 7.088161: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.936358: 100%|█| 256/256 [00:00<00:00, 5338.33it/s]\n",
"Epoch 215: Train, Loglike in nats: 7.116961: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.940352: 100%|█| 256/256 [00:00<00:00, 4689.83it/s]\n",
"Epoch 216: Train, Loglike in nats: 7.121807: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.835408: 100%|█| 256/256 [00:00<00:00, 5121.86it/s]\n",
"Epoch 217: Train, Loglike in nats: 7.098231: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.857529: 100%|█| 256/256 [00:00<00:00, 4665.74it/s]\n",
"Epoch 218: Train, Loglike in nats: 7.107583: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.848279: 100%|█| 256/256 [00:00<00:00, 5203.22it/s]\n",
"Epoch 219: Train, Loglike in nats: 7.154936: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.872961: 100%|█| 256/256 [00:00<00:00, 4903.65it/s]\n",
"Epoch 220: Train, Loglike in nats: 7.113741: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.871563: 100%|█| 256/256 [00:00<00:00, 5313.37it/s]\n",
"Epoch 221: Train, Loglike in nats: 7.158687: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 6.966138: 100%|█| 256/256 [00:00<00:00, 4899.51it/s]\n",
"Epoch 222: Train, Loglike in nats: 7.149098: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.841647: 100%|█| 256/256 [00:00<00:00, 5525.21it/s]\n",
"Epoch 223: Train, Loglike in nats: 7.155971: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.948618: 100%|█| 256/256 [00:00<00:00, 3561.99it/s]\n",
"Epoch 224: Train, Loglike in nats: 7.183771: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.890191: 100%|█| 256/256 [00:00<00:00, 5595.38it/s]\n",
"Epoch 225: Train, Loglike in nats: 7.138261: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.933090: 100%|█| 256/256 [00:00<00:00, 2250.73it/s]\n",
"Epoch 226: Train, Loglike in nats: 7.139557: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.847615: 100%|█| 256/256 [00:00<00:00, 5188.89it/s]\n",
"Epoch 227: Train, Loglike in nats: 7.189766: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.976440: 100%|█| 256/256 [00:00<00:00, 2123.72it/s]\n",
"Epoch 228: Train, Loglike in nats: 7.204551: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.932121: 100%|█| 256/256 [00:00<00:00, 5167.54it/s]\n",
"Epoch 229: Train, Loglike in nats: 7.198164: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.939200: 100%|█| 256/256 [00:00<00:00, 5386.97it/s]\n",
"Epoch 230: Train, Loglike in nats: 7.184942: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.868310: 100%|█| 256/256 [00:00<00:00, 2147.27it/s]\n",
"Epoch 231: Train, Loglike in nats: 7.207561: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.894862: 100%|█| 256/256 [00:00<00:00, 2809.85it/s]\n",
"Epoch 232: Train, Loglike in nats: 7.180940: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.903001: 100%|█| 256/256 [00:00<00:00, 3005.85it/s]\n",
"Epoch 233: Train, Loglike in nats: 7.182276: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.897264: 100%|█| 256/256 [00:00<00:00, 3164.32it/s]\n",
"Epoch 234: Train, Loglike in nats: 7.205443: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.981925: 100%|█| 256/256 [00:00<00:00, 3249.22it/s]\n",
"Epoch 235: Train, Loglike in nats: 7.215735: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.998847: 100%|█| 256/256 [00:00<00:00, 4806.43it/s]\n",
"Epoch 236: Train, Loglike in nats: 7.226037: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.854257: 100%|█| 256/256 [00:00<00:00, 3148.60it/s]\n",
"Epoch 237: Train, Loglike in nats: 7.204447: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.961008: 100%|█| 256/256 [00:00<00:00, 4398.85it/s]\n",
"Epoch 238: Train, Loglike in nats: 7.210616: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.983492: 100%|█| 256/256 [00:00<00:00, 3515.27it/s]\n",
"Epoch 239: Train, Loglike in nats: 7.245856: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.004025: 100%|█| 256/256 [00:00<00:00, 3712.49it/s]\n",
"Epoch 240: Train, Loglike in nats: 7.254111: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.970000: 100%|█| 256/256 [00:00<00:00, 2715.86it/s]\n",
"Epoch 241: Train, Loglike in nats: 7.248875: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.948079: 100%|█| 256/256 [00:00<00:00, 3404.53it/s]\n",
"Epoch 242: Train, Loglike in nats: 7.245370: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.983081: 100%|█| 256/256 [00:00<00:00, 4193.96it/s]\n",
"Epoch 243: Train, Loglike in nats: 7.256679: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.984150: 100%|█| 256/256 [00:00<00:00, 5436.86it/s]\n",
"Epoch 244: Train, Loglike in nats: 7.242065: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.965386: 100%|█| 256/256 [00:00<00:00, 4771.21it/s]\n",
"Epoch 245: Train, Loglike in nats: 7.256324: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.956080: 100%|█| 256/256 [00:00<00:00, 4771.76it/s]\n",
"Epoch 246: Train, Loglike in nats: 7.237340: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.065677: 100%|█| 256/256 [00:00<00:00, 4344.65it/s]\n",
"Epoch 247: Train, Loglike in nats: 7.264074: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.035882: 100%|█| 256/256 [00:00<00:00, 1993.49it/s]\n",
"Epoch 248: Train, Loglike in nats: 7.265508: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.046490: 100%|█| 256/256 [00:00<00:00, 4600.34it/s]\n",
"Epoch 249: Train, Loglike in nats: 7.254757: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.050386: 100%|█| 256/256 [00:00<00:00, 4308.74it/s]\n",
"Epoch 250: Train, Loglike in nats: 7.261817: 100%|█| 2304/2304 [00:02<00:00,\n",
"- Val, Loglike in nats: 6.974447: 100%|█| 256/256 [00:00<00:00, 2424.29it/s]\n",
"Epoch 251: Train, Loglike in nats: 7.272326: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.011924: 100%|█| 256/256 [00:00<00:00, 2394.45it/s]\n",
"Epoch 252: Train, Loglike in nats: 7.272482: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.034987: 100%|█| 256/256 [00:00<00:00, 2596.25it/s]\n",
"Epoch 253: Train, Loglike in nats: 7.273192: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.038661: 100%|█| 256/256 [00:00<00:00, 3571.27it/s]\n",
"Epoch 254: Train, Loglike in nats: 7.278097: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.051842: 100%|█| 256/256 [00:00<00:00, 1602.09it/s]\n",
"Epoch 255: Train, Loglike in nats: 7.274506: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.105392: 100%|█| 256/256 [00:00<00:00, 3935.99it/s]\n",
"Epoch 256: Train, Loglike in nats: 7.307196: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.070848: 100%|█| 256/256 [00:00<00:00, 3127.44it/s]\n",
"Epoch 257: Train, Loglike in nats: 7.287526: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.013665: 100%|█| 256/256 [00:00<00:00, 1881.80it/s]\n",
"Epoch 258: Train, Loglike in nats: 7.297325: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.028750: 100%|█| 256/256 [00:00<00:00, 1680.49it/s]\n",
"Epoch 259: Train, Loglike in nats: 7.298551: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.936582: 100%|█| 256/256 [00:00<00:00, 3390.14it/s]\n",
"Epoch 260: Train, Loglike in nats: 7.295615: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.098983: 100%|█| 256/256 [00:00<00:00, 3543.75it/s]\n",
"Epoch 261: Train, Loglike in nats: 7.278522: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.017798: 100%|█| 256/256 [00:00<00:00, 4716.12it/s]\n",
"Epoch 262: Train, Loglike in nats: 7.273079: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.047269: 100%|█| 256/256 [00:00<00:00, 4272.80it/s]\n",
"Epoch 263: Train, Loglike in nats: 7.304631: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.013604: 100%|█| 256/256 [00:00<00:00, 3246.06it/s]\n",
"Epoch 264: Train, Loglike in nats: 7.305711: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.999825: 100%|█| 256/256 [00:00<00:00, 3514.80it/s]\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 265: Train, Loglike in nats: 7.304676: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.008645: 100%|█| 256/256 [00:00<00:00, 3489.73it/s]\n",
"Epoch 266: Train, Loglike in nats: 7.279627: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.096175: 100%|█| 256/256 [00:00<00:00, 3291.83it/s]\n",
"Epoch 267: Train, Loglike in nats: 7.292471: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.023993: 100%|█| 256/256 [00:00<00:00, 4433.65it/s]\n",
"Epoch 268: Train, Loglike in nats: 7.318467: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.014885: 100%|█| 256/256 [00:00<00:00, 2584.37it/s]\n",
"Epoch 269: Train, Loglike in nats: 7.307246: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.107642: 100%|█| 256/256 [00:00<00:00, 3373.40it/s]\n",
"Epoch 270: Train, Loglike in nats: 7.321354: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.015020: 100%|█| 256/256 [00:00<00:00, 3856.87it/s]\n",
"Epoch 271: Train, Loglike in nats: 7.330850: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.065150: 100%|█| 256/256 [00:00<00:00, 3234.12it/s]\n",
"Epoch 272: Train, Loglike in nats: 7.298432: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.078562: 100%|█| 256/256 [00:00<00:00, 4202.66it/s]\n",
"Epoch 273: Train, Loglike in nats: 7.319243: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.084467: 100%|█| 256/256 [00:00<00:00, 1926.89it/s]\n",
"Epoch 274: Train, Loglike in nats: 7.308819: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.038202: 100%|█| 256/256 [00:00<00:00, 2776.60it/s]\n",
"Epoch 275: Train, Loglike in nats: 7.297086: 100%|█| 2304/2304 [00:02<00:00,\n",
"- Val, Loglike in nats: 7.083919: 100%|█| 256/256 [00:00<00:00, 1895.41it/s]\n",
"Epoch 276: Train, Loglike in nats: 7.316991: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.125168: 100%|█| 256/256 [00:00<00:00, 4807.23it/s]\n",
"Epoch 277: Train, Loglike in nats: 7.338279: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.100325: 100%|█| 256/256 [00:00<00:00, 3618.11it/s]\n",
"Epoch 278: Train, Loglike in nats: 7.316423: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.031301: 100%|█| 256/256 [00:00<00:00, 2889.69it/s]\n",
"Epoch 279: Train, Loglike in nats: 7.329750: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.976141: 100%|█| 256/256 [00:00<00:00, 4578.78it/s]\n",
"Epoch 280: Train, Loglike in nats: 7.328092: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.105322: 100%|█| 256/256 [00:00<00:00, 4358.87it/s]\n",
"Epoch 281: Train, Loglike in nats: 7.303649: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.059430: 100%|█| 256/256 [00:00<00:00, 4501.89it/s]\n",
"Epoch 282: Train, Loglike in nats: 7.344545: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.046972: 100%|█| 256/256 [00:00<00:00, 1959.23it/s]\n",
"Epoch 283: Train, Loglike in nats: 7.329625: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.020240: 100%|█| 256/256 [00:00<00:00, 2283.18it/s]\n",
"Epoch 284: Train, Loglike in nats: 7.335547: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.041233: 100%|█| 256/256 [00:00<00:00, 3231.71it/s]\n",
"Epoch 285: Train, Loglike in nats: 7.334932: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.936577: 100%|█| 256/256 [00:00<00:00, 3648.55it/s]\n",
"Epoch 286: Train, Loglike in nats: 7.341789: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.013749: 100%|█| 256/256 [00:00<00:00, 3991.15it/s]\n",
"Epoch 287: Train, Loglike in nats: 7.332159: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.064516: 100%|█| 256/256 [00:00<00:00, 4158.34it/s]\n",
"Epoch 288: Train, Loglike in nats: 7.351825: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.080848: 100%|█| 256/256 [00:00<00:00, 3911.26it/s]\n",
"Epoch 289: Train, Loglike in nats: 7.335212: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.069468: 100%|█| 256/256 [00:00<00:00, 4895.58it/s]\n",
"Epoch 290: Train, Loglike in nats: 7.331078: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.139069: 100%|█| 256/256 [00:00<00:00, 2995.73it/s]\n",
"Epoch 291: Train, Loglike in nats: 7.348450: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.058819: 100%|█| 256/256 [00:00<00:00, 3143.73it/s]\n",
"Epoch 292: Train, Loglike in nats: 7.327452: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.028440: 100%|█| 256/256 [00:00<00:00, 2791.38it/s]\n",
"Epoch 293: Train, Loglike in nats: 7.324356: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 6.991703: 100%|█| 256/256 [00:00<00:00, 4643.73it/s]\n",
"Epoch 294: Train, Loglike in nats: 7.326188: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.054939: 100%|█| 256/256 [00:00<00:00, 4476.46it/s]\n",
"Epoch 295: Train, Loglike in nats: 7.329691: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.035150: 100%|█| 256/256 [00:00<00:00, 2879.57it/s]\n",
"Epoch 296: Train, Loglike in nats: 7.334575: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.072271: 100%|█| 256/256 [00:00<00:00, 4211.15it/s]\n",
"Epoch 297: Train, Loglike in nats: 7.334792: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.126701: 100%|█| 256/256 [00:00<00:00, 2566.91it/s]\n",
"Epoch 298: Train, Loglike in nats: 7.358398: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.084436: 100%|█| 256/256 [00:00<00:00, 2220.77it/s]\n",
"Epoch 299: Train, Loglike in nats: 7.332334: 100%|█| 2304/2304 [00:01<00:00,\n",
"- Val, Loglike in nats: 7.055864: 100%|█| 256/256 [00:00<00:00, 4508.77it/s]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Generating simulations\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"256it [00:00, 3311.77it/s]\n",
"WARNING:root:Too few points to create valid contours\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Effective sample size for current/all rounds 41.5/41.5\n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"---------------------- Round: 1 ----------------------\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 0: Train, Loglike in nats: -111.074880: 100%|█| 2304/2304 [00:01<00:00\n",
"- Val, Loglike in nats: -2.992713: 100%|█| 256/256 [00:00<00:00, 1011.96it/s\n",
"Epoch 1: Train, Loglike in nats: -2.200279: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: -1.263758: 100%|█| 256/256 [00:00<00:00, 2931.44it/s\n",
"Epoch 2: Train, Loglike in nats: -0.458073: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 0.593628: 100%|█| 256/256 [00:00<00:00, 1573.11it/s]\n",
"Epoch 3: Train, Loglike in nats: 0.448375: 100%|█| 2304/2304 [00:01<00:00, 1\n",
"- Val, Loglike in nats: 1.122143: 100%|█| 256/256 [00:00<00:00, 2982.33it/s]\n",
"Epoch 4: Train, Loglike in nats: 0.888271: 100%|█| 2304/2304 [00:01<00:00, 1\n",
"- Val, Loglike in nats: 1.288034: 100%|█| 256/256 [00:00<00:00, 5056.57it/s]\n",
"Epoch 5: Train, Loglike in nats: 1.180175: 100%|█| 2304/2304 [00:01<00:00, 1\n",
"- Val, Loglike in nats: 1.176789: 100%|█| 256/256 [00:00<00:00, 1636.73it/s]\n",
"Epoch 6: Train, Loglike in nats: 0.437039: 100%|█| 2304/2304 [00:01<00:00, 1\n",
"- Val, Loglike in nats: 1.160046: 100%|█| 256/256 [00:00<00:00, 4627.80it/s]\n",
"Epoch 7: Train, Loglike in nats: 1.156423: 100%|█| 2304/2304 [00:01<00:00, 1\n",
"- Val, Loglike in nats: 1.305726: 100%|█| 256/256 [00:00<00:00, 4260.56it/s]\n",
"Epoch 8: Train, Loglike in nats: 1.439486: 100%|█| 2304/2304 [00:01<00:00, 1\n",
"- Val, Loglike in nats: 1.423906: 100%|█| 256/256 [00:00<00:00, 2631.07it/s]\n",
"Epoch 9: Train, Loglike in nats: 0.334956: 100%|█| 2304/2304 [00:01<00:00, 1\n",
"- Val, Loglike in nats: 0.881735: 100%|█| 256/256 [00:00<00:00, 4153.48it/s]\n",
"Epoch 10: Train, Loglike in nats: 1.285358: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 1.543077: 100%|█| 256/256 [00:00<00:00, 2772.33it/s]\n",
"Epoch 11: Train, Loglike in nats: 1.688851: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 2.028180: 100%|█| 256/256 [00:00<00:00, 2095.94it/s]\n",
"Epoch 12: Train, Loglike in nats: 1.396050: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 2.235658: 100%|█| 256/256 [00:00<00:00, 3411.78it/s]\n",
"Epoch 13: Train, Loglike in nats: 2.059073: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 1.859036: 100%|█| 256/256 [00:00<00:00, 4622.52it/s]\n",
"Epoch 14: Train, Loglike in nats: 1.812682: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.185026: 100%|█| 256/256 [00:00<00:00, 5180.90it/s]\n",
"Epoch 15: Train, Loglike in nats: 1.022742: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 1.421531: 100%|█| 256/256 [00:00<00:00, 5014.18it/s]\n",
"Epoch 16: Train, Loglike in nats: 1.589736: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.048828: 100%|█| 256/256 [00:00<00:00, 5227.18it/s]\n",
"Epoch 17: Train, Loglike in nats: 2.244287: 100%|█| 2304/2304 [00:01<00:00, \n",
"- Val, Loglike in nats: 2.458492: 100%|█| 256/256 [00:00<00:00, 5533.72it/s]\n",
"Epoch 18: Train, Loglike in nats: 2.408170: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.579831: 100%|█| 256/256 [00:00<00:00, 5378.61it/s]\n",
"Epoch 19: Train, Loglike in nats: 2.464422: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.574936: 100%|█| 256/256 [00:00<00:00, 6135.70it/s]\n",
"Epoch 20: Train, Loglike in nats: 1.841877: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 1.567079: 100%|█| 256/256 [00:00<00:00, 5631.44it/s]\n",
"Epoch 21: Train, Loglike in nats: 1.699927: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 1.867174: 100%|█| 256/256 [00:00<00:00, 5348.84it/s]\n",
"Epoch 22: Train, Loglike in nats: 2.252223: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.682520: 100%|█| 256/256 [00:00<00:00, 5644.35it/s]\n",
"Epoch 23: Train, Loglike in nats: 2.672771: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.756768: 100%|█| 256/256 [00:00<00:00, 6104.73it/s]\n",
"Epoch 24: Train, Loglike in nats: 2.675641: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.801300: 100%|█| 256/256 [00:00<00:00, 5990.53it/s]\n",
"Epoch 25: Train, Loglike in nats: 2.603892: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.554320: 100%|█| 256/256 [00:00<00:00, 5988.72it/s]\n",
"Epoch 26: Train, Loglike in nats: 2.436226: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.693611: 100%|█| 256/256 [00:00<00:00, 5991.53it/s]\n",
"Epoch 27: Train, Loglike in nats: 2.694747: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.908753: 100%|█| 256/256 [00:00<00:00, 6244.68it/s]\n",
"Epoch 28: Train, Loglike in nats: 2.786688: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.875337: 100%|█| 256/256 [00:00<00:00, 5821.00it/s]\n",
"Epoch 29: Train, Loglike in nats: 2.839208: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.993960: 100%|█| 256/256 [00:00<00:00, 5818.48it/s]\n",
"Epoch 30: Train, Loglike in nats: 2.429057: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.334218: 100%|█| 256/256 [00:00<00:00, 6680.41it/s]\n",
"Epoch 31: Train, Loglike in nats: 2.378411: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.966625: 100%|█| 256/256 [00:00<00:00, 7189.38it/s]\n",
"Epoch 32: Train, Loglike in nats: 2.730419: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.067818: 100%|█| 256/256 [00:00<00:00, 6680.07it/s]\n",
"Epoch 33: Train, Loglike in nats: 2.804069: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.090227: 100%|█| 256/256 [00:00<00:00, 6888.70it/s]\n",
"Epoch 34: Train, Loglike in nats: 2.904780: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.112225: 100%|█| 256/256 [00:00<00:00, 6688.31it/s]\n",
"Epoch 35: Train, Loglike in nats: 2.941289: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.174132: 100%|█| 256/256 [00:00<00:00, 6837.03it/s]\n",
"Epoch 36: Train, Loglike in nats: 3.059609: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.152189: 100%|█| 256/256 [00:00<00:00, 6466.77it/s]\n",
"Epoch 37: Train, Loglike in nats: 2.592032: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.111625: 100%|█| 256/256 [00:00<00:00, 6641.73it/s]\n",
"Epoch 38: Train, Loglike in nats: 3.075136: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.227358: 100%|█| 256/256 [00:00<00:00, 5478.39it/s]\n",
"Epoch 39: Train, Loglike in nats: 2.279360: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.156766: 100%|█| 256/256 [00:00<00:00, 6817.71it/s]\n",
"Epoch 40: Train, Loglike in nats: 1.492288: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.411548: 100%|█| 256/256 [00:00<00:00, 6492.02it/s]\n",
"Epoch 41: Train, Loglike in nats: 2.716307: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.030724: 100%|█| 256/256 [00:00<00:00, 6864.17it/s]\n",
"Epoch 42: Train, Loglike in nats: 2.965123: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.020025: 100%|█| 256/256 [00:00<00:00, 7440.32it/s]\n",
"Epoch 43: Train, Loglike in nats: 3.200281: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.818868: 100%|█| 256/256 [00:00<00:00, 6929.24it/s]\n",
"Epoch 44: Train, Loglike in nats: 3.175573: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.390544: 100%|█| 256/256 [00:00<00:00, 6885.48it/s]\n",
"Epoch 45: Train, Loglike in nats: 3.308076: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.383103: 100%|█| 256/256 [00:00<00:00, 6879.30it/s]\n",
"Epoch 46: Train, Loglike in nats: 3.367786: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.437356: 100%|█| 256/256 [00:00<00:00, 6793.51it/s]\n",
"Epoch 47: Train, Loglike in nats: 3.050139: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.672836: 100%|█| 256/256 [00:00<00:00, 7071.02it/s]\n",
"Epoch 48: Train, Loglike in nats: 3.210750: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.074204: 100%|█| 256/256 [00:00<00:00, 6853.70it/s]\n",
"Epoch 49: Train, Loglike in nats: 3.146906: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.012418: 100%|█| 256/256 [00:00<00:00, 7034.70it/s]\n",
"Epoch 50: Train, Loglike in nats: 3.432348: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.808176: 100%|█| 256/256 [00:00<00:00, 6720.08it/s]\n",
"Epoch 51: Train, Loglike in nats: 3.692205: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.634159: 100%|█| 256/256 [00:00<00:00, 6637.46it/s]\n",
"Epoch 52: Train, Loglike in nats: 3.002430: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.531565: 100%|█| 256/256 [00:00<00:00, 6952.58it/s]\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 53: Train, Loglike in nats: 3.671081: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.016607: 100%|█| 256/256 [00:00<00:00, 7086.66it/s]\n",
"Epoch 54: Train, Loglike in nats: 3.971538: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.120502: 100%|█| 256/256 [00:00<00:00, 6758.15it/s]\n",
"Epoch 55: Train, Loglike in nats: 3.898712: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.700991: 100%|█| 256/256 [00:00<00:00, 6854.84it/s]\n",
"Epoch 56: Train, Loglike in nats: 3.712658: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.855739: 100%|█| 256/256 [00:00<00:00, 6845.83it/s]\n",
"Epoch 57: Train, Loglike in nats: 3.984127: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.148764: 100%|█| 256/256 [00:00<00:00, 7051.10it/s]\n",
"Epoch 58: Train, Loglike in nats: 4.120578: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.072963: 100%|█| 256/256 [00:00<00:00, 7087.12it/s]\n",
"Epoch 59: Train, Loglike in nats: 4.048295: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.859898: 100%|█| 256/256 [00:00<00:00, 6881.86it/s]\n",
"Epoch 60: Train, Loglike in nats: 3.923438: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.131983: 100%|█| 256/256 [00:00<00:00, 6844.22it/s]\n",
"Epoch 61: Train, Loglike in nats: 3.902147: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.986555: 100%|█| 256/256 [00:00<00:00, 7084.27it/s]\n",
"Epoch 62: Train, Loglike in nats: 3.634935: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.726820: 100%|█| 256/256 [00:00<00:00, 6909.62it/s]\n",
"Epoch 63: Train, Loglike in nats: 3.760851: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.847198: 100%|█| 256/256 [00:00<00:00, 7023.52it/s]\n",
"Epoch 64: Train, Loglike in nats: 3.964635: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.103169: 100%|█| 256/256 [00:00<00:00, 7195.84it/s]\n",
"Epoch 65: Train, Loglike in nats: 4.114371: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.683813: 100%|█| 256/256 [00:00<00:00, 6828.81it/s]\n",
"Epoch 66: Train, Loglike in nats: 3.949405: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.926994: 100%|█| 256/256 [00:00<00:00, 6974.16it/s]\n",
"Epoch 67: Train, Loglike in nats: 4.134924: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.332856: 100%|█| 256/256 [00:00<00:00, 7034.15it/s]\n",
"Epoch 68: Train, Loglike in nats: 4.173901: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.312083: 100%|█| 256/256 [00:00<00:00, 6939.54it/s]\n",
"Epoch 69: Train, Loglike in nats: 4.215352: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.347766: 100%|█| 256/256 [00:00<00:00, 6808.16it/s]\n",
"Epoch 70: Train, Loglike in nats: 4.322240: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.269313: 100%|█| 256/256 [00:00<00:00, 6946.50it/s]\n",
"Epoch 71: Train, Loglike in nats: 4.059325: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.094560: 100%|█| 256/256 [00:00<00:00, 6932.96it/s]\n",
"Epoch 72: Train, Loglike in nats: 3.645744: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.082846: 100%|█| 256/256 [00:00<00:00, 6763.39it/s]\n",
"Epoch 73: Train, Loglike in nats: 4.212441: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.459779: 100%|█| 256/256 [00:00<00:00, 6779.27it/s]\n",
"Epoch 74: Train, Loglike in nats: 4.284605: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.385456: 100%|█| 256/256 [00:00<00:00, 7150.70it/s]\n",
"Epoch 75: Train, Loglike in nats: 4.297160: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.336185: 100%|█| 256/256 [00:00<00:00, 6877.23it/s]\n",
"Epoch 76: Train, Loglike in nats: 4.333346: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.342100: 100%|█| 256/256 [00:00<00:00, 7229.37it/s]\n",
"Epoch 77: Train, Loglike in nats: 4.265670: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.549646: 100%|█| 256/256 [00:00<00:00, 6966.38it/s]\n",
"Epoch 78: Train, Loglike in nats: 4.323771: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.423072: 100%|█| 256/256 [00:00<00:00, 6832.29it/s]\n",
"Epoch 79: Train, Loglike in nats: 4.381990: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.543144: 100%|█| 256/256 [00:00<00:00, 6808.33it/s]\n",
"Epoch 80: Train, Loglike in nats: 4.465274: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.600409: 100%|█| 256/256 [00:00<00:00, 6949.88it/s]\n",
"Epoch 81: Train, Loglike in nats: 3.981684: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.516016: 100%|█| 256/256 [00:00<00:00, 6877.27it/s]\n",
"Epoch 82: Train, Loglike in nats: 4.335044: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.345150: 100%|█| 256/256 [00:00<00:00, 6953.70it/s]\n",
"Epoch 83: Train, Loglike in nats: 4.396036: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.485052: 100%|█| 256/256 [00:00<00:00, 7030.10it/s]\n",
"Epoch 84: Train, Loglike in nats: 4.434524: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.509062: 100%|█| 256/256 [00:00<00:00, 6868.78it/s]\n",
"Epoch 85: Train, Loglike in nats: 4.457015: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.343770: 100%|█| 256/256 [00:00<00:00, 6982.10it/s]\n",
"Epoch 86: Train, Loglike in nats: 4.361277: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.330976: 100%|█| 256/256 [00:00<00:00, 7094.19it/s]\n",
"Epoch 87: Train, Loglike in nats: 4.476224: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.604631: 100%|█| 256/256 [00:00<00:00, 6675.88it/s]\n",
"Epoch 88: Train, Loglike in nats: 4.423406: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.412801: 100%|█| 256/256 [00:00<00:00, 6920.22it/s]\n",
"Epoch 89: Train, Loglike in nats: 3.997995: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.975684: 100%|█| 256/256 [00:00<00:00, 7119.27it/s]\n",
"Epoch 90: Train, Loglike in nats: 3.941249: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.520312: 100%|█| 256/256 [00:00<00:00, 7021.22it/s]\n",
"Epoch 91: Train, Loglike in nats: 4.215732: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.296104: 100%|█| 256/256 [00:00<00:00, 6826.34it/s]\n",
"Epoch 92: Train, Loglike in nats: 4.370192: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.090794: 100%|█| 256/256 [00:00<00:00, 7004.10it/s]\n",
"Epoch 93: Train, Loglike in nats: 3.842001: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.613758: 100%|█| 256/256 [00:00<00:00, 6998.80it/s]\n",
"Epoch 94: Train, Loglike in nats: 3.747036: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.108294: 100%|█| 256/256 [00:00<00:00, 6785.01it/s]\n",
"Epoch 95: Train, Loglike in nats: 4.273025: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.538351: 100%|█| 256/256 [00:00<00:00, 6952.35it/s]\n",
"Epoch 96: Train, Loglike in nats: 4.499550: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.592475: 100%|█| 256/256 [00:00<00:00, 6763.39it/s]\n",
"Epoch 97: Train, Loglike in nats: 4.576464: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.513281: 100%|█| 256/256 [00:00<00:00, 6665.56it/s]\n",
"Epoch 98: Train, Loglike in nats: 4.449008: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.438046: 100%|█| 256/256 [00:00<00:00, 6896.75it/s]\n",
"Epoch 99: Train, Loglike in nats: 4.426386: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.289808: 100%|█| 256/256 [00:00<00:00, 6713.53it/s]\n",
"Epoch 100: Train, Loglike in nats: 4.380172: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.494611: 100%|█| 256/256 [00:00<00:00, 6938.20it/s]\n",
"Epoch 101: Train, Loglike in nats: 4.503746: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.443243: 100%|█| 256/256 [00:00<00:00, 7057.82it/s]\n",
"Epoch 102: Train, Loglike in nats: 4.449413: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.437430: 100%|█| 256/256 [00:00<00:00, 7120.45it/s]\n",
"Epoch 103: Train, Loglike in nats: 4.535067: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.611163: 100%|█| 256/256 [00:00<00:00, 6977.20it/s]\n",
"Epoch 104: Train, Loglike in nats: 4.592947: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.731658: 100%|█| 256/256 [00:00<00:00, 6573.38it/s]\n",
"Epoch 105: Train, Loglike in nats: 4.625255: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.550815: 100%|█| 256/256 [00:00<00:00, 6880.05it/s]\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 106: Train, Loglike in nats: 4.543944: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.502524: 100%|█| 256/256 [00:00<00:00, 6871.33it/s]\n",
"Epoch 107: Train, Loglike in nats: 4.582432: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.757286: 100%|█| 256/256 [00:00<00:00, 7001.26it/s]\n",
"Epoch 108: Train, Loglike in nats: 4.609005: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.654587: 100%|█| 256/256 [00:00<00:00, 6992.42it/s]\n",
"Epoch 109: Train, Loglike in nats: 4.604053: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.633290: 100%|█| 256/256 [00:00<00:00, 7110.07it/s]\n",
"Epoch 110: Train, Loglike in nats: 4.634672: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.683434: 100%|█| 256/256 [00:00<00:00, 7032.40it/s]\n",
"Epoch 111: Train, Loglike in nats: 4.668680: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.722544: 100%|█| 256/256 [00:00<00:00, 7167.07it/s]\n",
"Epoch 112: Train, Loglike in nats: 4.610692: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.553889: 100%|█| 256/256 [00:00<00:00, 7078.11it/s]\n",
"Epoch 113: Train, Loglike in nats: 4.636194: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.729086: 100%|█| 256/256 [00:00<00:00, 6700.71it/s]\n",
"Epoch 114: Train, Loglike in nats: 4.638307: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.639952: 100%|█| 256/256 [00:00<00:00, 6839.47it/s]\n",
"Epoch 115: Train, Loglike in nats: 4.649648: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.688537: 100%|█| 256/256 [00:00<00:00, 7066.32it/s]\n",
"Epoch 116: Train, Loglike in nats: 4.646255: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.445121: 100%|█| 256/256 [00:00<00:00, 6767.48it/s]\n",
"Epoch 117: Train, Loglike in nats: 4.546423: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.684577: 100%|█| 256/256 [00:00<00:00, 7017.00it/s]\n",
"Epoch 118: Train, Loglike in nats: 4.664975: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.409183: 100%|█| 256/256 [00:00<00:00, 6955.05it/s]\n",
"Epoch 119: Train, Loglike in nats: 4.597261: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.618570: 100%|█| 256/256 [00:00<00:00, 6771.45it/s]\n",
"Epoch 120: Train, Loglike in nats: 4.581435: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.448646: 100%|█| 256/256 [00:00<00:00, 6969.95it/s]\n",
"Epoch 121: Train, Loglike in nats: 4.430024: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.712512: 100%|█| 256/256 [00:00<00:00, 7052.12it/s]\n",
"Epoch 122: Train, Loglike in nats: 4.619772: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.788047: 100%|█| 256/256 [00:00<00:00, 6864.96it/s]\n",
"Epoch 123: Train, Loglike in nats: 4.630033: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.620933: 100%|█| 256/256 [00:00<00:00, 7022.14it/s]\n",
"Epoch 124: Train, Loglike in nats: 4.619167: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.821687: 100%|█| 256/256 [00:00<00:00, 6705.77it/s]\n",
"Epoch 125: Train, Loglike in nats: 4.676330: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.700980: 100%|█| 256/256 [00:00<00:00, 6830.46it/s]\n",
"Epoch 126: Train, Loglike in nats: 4.709170: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.638464: 100%|█| 256/256 [00:00<00:00, 7211.50it/s]\n",
"Epoch 127: Train, Loglike in nats: 4.695473: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.770936: 100%|█| 256/256 [00:00<00:00, 6206.67it/s]\n",
"Epoch 128: Train, Loglike in nats: 4.760443: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.744103: 100%|█| 256/256 [00:00<00:00, 7209.08it/s]\n",
"Epoch 129: Train, Loglike in nats: 4.722085: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.267892: 100%|█| 256/256 [00:00<00:00, 7036.78it/s]\n",
"Epoch 130: Train, Loglike in nats: 4.662899: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.741573: 100%|█| 256/256 [00:00<00:00, 6607.48it/s]\n",
"Epoch 131: Train, Loglike in nats: 4.704491: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.733673: 100%|█| 256/256 [00:00<00:00, 6689.44it/s]\n",
"Epoch 132: Train, Loglike in nats: 4.668829: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.765700: 100%|█| 256/256 [00:00<00:00, 6842.52it/s]\n",
"Epoch 133: Train, Loglike in nats: 4.766939: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.748709: 100%|█| 256/256 [00:00<00:00, 6789.17it/s]\n",
"Epoch 134: Train, Loglike in nats: 4.688504: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.661104: 100%|█| 256/256 [00:00<00:00, 7044.95it/s]\n",
"Epoch 135: Train, Loglike in nats: 4.487069: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.504766: 100%|█| 256/256 [00:00<00:00, 6824.30it/s]\n",
"Epoch 136: Train, Loglike in nats: 4.740299: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.912984: 100%|█| 256/256 [00:00<00:00, 6779.66it/s]\n",
"Epoch 137: Train, Loglike in nats: 4.707845: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.808662: 100%|█| 256/256 [00:00<00:00, 6939.32it/s]\n",
"Epoch 138: Train, Loglike in nats: 4.725697: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.816910: 100%|█| 256/256 [00:00<00:00, 6482.85it/s]\n",
"Epoch 139: Train, Loglike in nats: 4.753620: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.815832: 100%|█| 256/256 [00:00<00:00, 6778.97it/s]\n",
"Epoch 140: Train, Loglike in nats: 4.828905: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.823953: 100%|█| 256/256 [00:00<00:00, 6570.20it/s]\n",
"Epoch 141: Train, Loglike in nats: 4.773109: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.783313: 100%|█| 256/256 [00:00<00:00, 6591.66it/s]\n",
"Epoch 142: Train, Loglike in nats: 4.819690: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.818953: 100%|█| 256/256 [00:00<00:00, 6597.61it/s]\n",
"Epoch 143: Train, Loglike in nats: 4.867340: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.853146: 100%|█| 256/256 [00:00<00:00, 6911.45it/s]\n",
"Epoch 144: Train, Loglike in nats: 4.760531: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.905544: 100%|█| 256/256 [00:00<00:00, 6779.32it/s]\n",
"Epoch 145: Train, Loglike in nats: 4.727601: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.756484: 100%|█| 256/256 [00:00<00:00, 6892.46it/s]\n",
"Epoch 146: Train, Loglike in nats: 4.596042: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.777543: 100%|█| 256/256 [00:00<00:00, 6797.43it/s]\n",
"Epoch 147: Train, Loglike in nats: 4.734501: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.892480: 100%|█| 256/256 [00:00<00:00, 6442.86it/s]\n",
"Epoch 148: Train, Loglike in nats: 4.830244: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.731425: 100%|█| 256/256 [00:00<00:00, 6402.37it/s]\n",
"Epoch 149: Train, Loglike in nats: 4.798859: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.859170: 100%|█| 256/256 [00:00<00:00, 6886.54it/s]\n",
"Epoch 150: Train, Loglike in nats: 4.822317: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.817088: 100%|█| 256/256 [00:00<00:00, 7012.74it/s]\n",
"Epoch 151: Train, Loglike in nats: 4.815515: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.852090: 100%|█| 256/256 [00:00<00:00, 6727.66it/s]\n",
"Epoch 152: Train, Loglike in nats: 4.815280: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.861661: 100%|█| 256/256 [00:00<00:00, 6851.16it/s]\n",
"Epoch 153: Train, Loglike in nats: 4.788223: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.944104: 100%|█| 256/256 [00:00<00:00, 6999.35it/s]\n",
"Epoch 154: Train, Loglike in nats: 4.833244: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.852159: 100%|█| 256/256 [00:00<00:00, 6862.11it/s]\n",
"Epoch 155: Train, Loglike in nats: 4.848967: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.835477: 100%|█| 256/256 [00:00<00:00, 6720.04it/s]\n",
"Epoch 156: Train, Loglike in nats: 4.882189: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.911913: 100%|█| 256/256 [00:00<00:00, 7465.46it/s]\n",
"Epoch 157: Train, Loglike in nats: 4.876556: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.876873: 100%|█| 256/256 [00:00<00:00, 7640.17it/s]\n",
"Epoch 158: Train, Loglike in nats: 4.815367: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.907768: 100%|█| 256/256 [00:00<00:00, 7642.51it/s]\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 159: Train, Loglike in nats: 4.878157: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.991665: 100%|█| 256/256 [00:00<00:00, 7407.36it/s]\n",
"Epoch 160: Train, Loglike in nats: 4.877523: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.859263: 100%|█| 256/256 [00:00<00:00, 7600.53it/s]\n",
"Epoch 161: Train, Loglike in nats: 4.894128: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.910856: 100%|█| 256/256 [00:00<00:00, 7237.99it/s]\n",
"Epoch 162: Train, Loglike in nats: 4.888952: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.894308: 100%|█| 256/256 [00:00<00:00, 7055.50it/s]\n",
"Epoch 163: Train, Loglike in nats: 4.890267: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.943397: 100%|█| 256/256 [00:00<00:00, 7306.95it/s]\n",
"Epoch 164: Train, Loglike in nats: 4.918437: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.972646: 100%|█| 256/256 [00:00<00:00, 7386.22it/s]\n",
"Epoch 165: Train, Loglike in nats: 4.951344: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.925364: 100%|█| 256/256 [00:00<00:00, 7045.00it/s]\n",
"Epoch 166: Train, Loglike in nats: 4.911215: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.964138: 100%|█| 256/256 [00:00<00:00, 7330.65it/s]\n",
"Epoch 167: Train, Loglike in nats: 4.883754: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.863797: 100%|█| 256/256 [00:00<00:00, 7727.37it/s]\n",
"Epoch 168: Train, Loglike in nats: 4.932081: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.081136: 100%|█| 256/256 [00:00<00:00, 7586.14it/s]\n",
"Epoch 169: Train, Loglike in nats: 4.828330: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.855214: 100%|█| 256/256 [00:00<00:00, 6998.94it/s]\n",
"Epoch 170: Train, Loglike in nats: 4.831960: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.973392: 100%|█| 256/256 [00:00<00:00, 7335.25it/s]\n",
"Epoch 171: Train, Loglike in nats: 4.858005: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.978397: 100%|█| 256/256 [00:00<00:00, 7498.25it/s]\n",
"Epoch 172: Train, Loglike in nats: 4.941252: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.814515: 100%|█| 256/256 [00:00<00:00, 7194.25it/s]\n",
"Epoch 173: Train, Loglike in nats: 4.938721: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.900439: 100%|█| 256/256 [00:00<00:00, 7271.27it/s]\n",
"Epoch 174: Train, Loglike in nats: 4.924385: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.058537: 100%|█| 256/256 [00:00<00:00, 7481.48it/s]\n",
"Epoch 175: Train, Loglike in nats: 4.909537: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.874205: 100%|█| 256/256 [00:00<00:00, 7550.29it/s]\n",
"Epoch 176: Train, Loglike in nats: 4.926949: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.992140: 100%|█| 256/256 [00:00<00:00, 7476.63it/s]\n",
"Epoch 177: Train, Loglike in nats: 4.960558: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.993335: 100%|█| 256/256 [00:00<00:00, 7665.59it/s]\n",
"Epoch 178: Train, Loglike in nats: 4.954484: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.720257: 100%|█| 256/256 [00:00<00:00, 7531.65it/s]\n",
"Epoch 179: Train, Loglike in nats: 4.889739: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.906847: 100%|█| 256/256 [00:00<00:00, 7340.27it/s]\n",
"Epoch 180: Train, Loglike in nats: 4.924509: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.981167: 100%|█| 256/256 [00:00<00:00, 7072.05it/s]\n",
"Epoch 181: Train, Loglike in nats: 4.943957: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.909654: 100%|█| 256/256 [00:00<00:00, 7086.66it/s]\n",
"Epoch 182: Train, Loglike in nats: 4.891324: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.814131: 100%|█| 256/256 [00:00<00:00, 7149.65it/s]\n",
"Epoch 183: Train, Loglike in nats: 4.849531: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.805203: 100%|█| 256/256 [00:00<00:00, 7639.46it/s]\n",
"Epoch 184: Train, Loglike in nats: 4.830100: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.874947: 100%|█| 256/256 [00:00<00:00, 7629.26it/s]\n",
"Epoch 185: Train, Loglike in nats: 4.969568: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.093703: 100%|█| 256/256 [00:00<00:00, 7564.49it/s]\n",
"Epoch 186: Train, Loglike in nats: 5.008568: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.956884: 100%|█| 256/256 [00:00<00:00, 7654.39it/s]\n",
"Epoch 187: Train, Loglike in nats: 4.979974: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.086188: 100%|█| 256/256 [00:00<00:00, 7655.75it/s]\n",
"Epoch 188: Train, Loglike in nats: 4.959536: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.992037: 100%|█| 256/256 [00:00<00:00, 7205.74it/s]\n",
"Epoch 189: Train, Loglike in nats: 4.972693: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.009516: 100%|█| 256/256 [00:00<00:00, 7379.62it/s]\n",
"Epoch 190: Train, Loglike in nats: 4.979783: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.060404: 100%|█| 256/256 [00:00<00:00, 7361.40it/s]\n",
"Epoch 191: Train, Loglike in nats: 4.945945: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.937216: 100%|█| 256/256 [00:00<00:00, 7354.85it/s]\n",
"Epoch 192: Train, Loglike in nats: 4.979626: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.894011: 100%|█| 256/256 [00:00<00:00, 7411.04it/s]\n",
"Epoch 193: Train, Loglike in nats: 4.960504: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.019113: 100%|█| 256/256 [00:00<00:00, 7607.91it/s]\n",
"Epoch 194: Train, Loglike in nats: 4.989235: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.070102: 100%|█| 256/256 [00:00<00:00, 7618.11it/s]\n",
"Epoch 195: Train, Loglike in nats: 4.929754: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.124246: 100%|█| 256/256 [00:00<00:00, 7594.72it/s]\n",
"Epoch 196: Train, Loglike in nats: 4.981910: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.961595: 100%|█| 256/256 [00:00<00:00, 7456.59it/s]\n",
"Epoch 197: Train, Loglike in nats: 4.936925: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.933918: 100%|█| 256/256 [00:00<00:00, 7230.83it/s]\n",
"Epoch 198: Train, Loglike in nats: 5.000826: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.017155: 100%|█| 256/256 [00:00<00:00, 7091.24it/s]\n",
"Epoch 199: Train, Loglike in nats: 5.037700: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.998076: 100%|█| 256/256 [00:00<00:00, 7337.56it/s]\n",
"Epoch 200: Train, Loglike in nats: 5.068087: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.087027: 100%|█| 256/256 [00:00<00:00, 7130.52it/s]\n",
"Epoch 201: Train, Loglike in nats: 5.031738: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.011127: 100%|█| 256/256 [00:00<00:00, 7283.90it/s]\n",
"Epoch 202: Train, Loglike in nats: 5.044392: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.142298: 100%|█| 256/256 [00:00<00:00, 7300.54it/s]\n",
"Epoch 203: Train, Loglike in nats: 5.013560: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.010705: 100%|█| 256/256 [00:00<00:00, 6451.61it/s]\n",
"Epoch 204: Train, Loglike in nats: 4.955142: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.136156: 100%|█| 256/256 [00:00<00:00, 7361.15it/s]\n",
"Epoch 205: Train, Loglike in nats: 5.023201: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.058125: 100%|█| 256/256 [00:00<00:00, 7313.22it/s]\n",
"Epoch 206: Train, Loglike in nats: 5.028421: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.043618: 100%|█| 256/256 [00:00<00:00, 7105.37it/s]\n",
"Epoch 207: Train, Loglike in nats: 5.024207: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.064460: 100%|█| 256/256 [00:00<00:00, 7073.40it/s]\n",
"Epoch 208: Train, Loglike in nats: 5.071890: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.099055: 100%|█| 256/256 [00:00<00:00, 7020.58it/s]\n",
"Epoch 209: Train, Loglike in nats: 5.043016: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.059133: 100%|█| 256/256 [00:00<00:00, 6684.40it/s]\n",
"Epoch 210: Train, Loglike in nats: 5.042257: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.132987: 100%|█| 256/256 [00:00<00:00, 7267.78it/s]\n",
"Epoch 211: Train, Loglike in nats: 5.046383: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.097592: 100%|█| 256/256 [00:00<00:00, 6788.32it/s]\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 212: Train, Loglike in nats: 5.006162: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.107549: 100%|█| 256/256 [00:00<00:00, 7571.96it/s]\n",
"Epoch 213: Train, Loglike in nats: 5.037832: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.116076: 100%|█| 256/256 [00:00<00:00, 7204.00it/s]\n",
"Epoch 214: Train, Loglike in nats: 5.043028: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.041624: 100%|█| 256/256 [00:00<00:00, 7200.09it/s]\n",
"Epoch 215: Train, Loglike in nats: 5.083967: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.090517: 100%|█| 256/256 [00:00<00:00, 7171.71it/s]\n",
"Epoch 216: Train, Loglike in nats: 5.062029: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.990519: 100%|█| 256/256 [00:00<00:00, 6112.58it/s]\n",
"Epoch 217: Train, Loglike in nats: 5.040047: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.076622: 100%|█| 256/256 [00:00<00:00, 7185.15it/s]\n",
"Epoch 218: Train, Loglike in nats: 5.018942: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.046337: 100%|█| 256/256 [00:00<00:00, 7473.36it/s]\n",
"Epoch 219: Train, Loglike in nats: 5.122868: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.024473: 100%|█| 256/256 [00:00<00:00, 7661.21it/s]\n",
"Epoch 220: Train, Loglike in nats: 5.030771: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.184031: 100%|█| 256/256 [00:00<00:00, 7381.19it/s]\n",
"Epoch 221: Train, Loglike in nats: 5.085471: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.068532: 100%|█| 256/256 [00:00<00:00, 7387.54it/s]\n",
"Epoch 222: Train, Loglike in nats: 5.048608: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.080677: 100%|█| 256/256 [00:00<00:00, 7656.68it/s]\n",
"Epoch 223: Train, Loglike in nats: 5.060910: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.176305: 100%|█| 256/256 [00:00<00:00, 7185.39it/s]\n",
"Epoch 224: Train, Loglike in nats: 5.106030: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.163081: 100%|█| 256/256 [00:00<00:00, 7374.30it/s]\n",
"Epoch 225: Train, Loglike in nats: 5.079444: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.074261: 100%|█| 256/256 [00:00<00:00, 7375.26it/s]\n",
"Epoch 226: Train, Loglike in nats: 5.070240: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.116743: 100%|█| 256/256 [00:00<00:00, 7364.53it/s]\n",
"Epoch 227: Train, Loglike in nats: 5.100734: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.150346: 100%|█| 256/256 [00:00<00:00, 7290.13it/s]\n",
"Epoch 228: Train, Loglike in nats: 5.087965: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.085164: 100%|█| 256/256 [00:00<00:00, 7203.37it/s]\n",
"Epoch 229: Train, Loglike in nats: 5.086281: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.165950: 100%|█| 256/256 [00:00<00:00, 6531.32it/s]\n",
"Epoch 230: Train, Loglike in nats: 5.069295: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.023100: 100%|█| 256/256 [00:00<00:00, 6515.62it/s]\n",
"Epoch 231: Train, Loglike in nats: 5.081991: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.089276: 100%|█| 256/256 [00:00<00:00, 7521.20it/s]\n",
"Epoch 232: Train, Loglike in nats: 5.097532: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.065683: 100%|█| 256/256 [00:00<00:00, 7097.85it/s]\n",
"Epoch 233: Train, Loglike in nats: 5.078255: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.193214: 100%|█| 256/256 [00:00<00:00, 7135.30it/s]\n",
"Epoch 234: Train, Loglike in nats: 5.106338: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.060592: 100%|█| 256/256 [00:00<00:00, 7350.22it/s]\n",
"Epoch 235: Train, Loglike in nats: 5.120595: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.114061: 100%|█| 256/256 [00:00<00:00, 7522.57it/s]\n",
"Epoch 236: Train, Loglike in nats: 5.113690: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.137984: 100%|█| 256/256 [00:00<00:00, 7432.95it/s]\n",
"Epoch 237: Train, Loglike in nats: 5.140368: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.163786: 100%|█| 256/256 [00:00<00:00, 7315.91it/s]\n",
"Epoch 238: Train, Loglike in nats: 5.123009: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.115195: 100%|█| 256/256 [00:00<00:00, 7411.10it/s]\n",
"Epoch 239: Train, Loglike in nats: 5.122023: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.115406: 100%|█| 256/256 [00:00<00:00, 7031.25it/s]\n",
"Epoch 240: Train, Loglike in nats: 5.123559: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.134575: 100%|█| 256/256 [00:00<00:00, 7162.24it/s]\n",
"Epoch 241: Train, Loglike in nats: 5.132092: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.148455: 100%|█| 256/256 [00:00<00:00, 7141.28it/s]\n",
"Epoch 242: Train, Loglike in nats: 5.121586: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.235629: 100%|█| 256/256 [00:00<00:00, 7366.71it/s]\n",
"Epoch 243: Train, Loglike in nats: 5.114081: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.156068: 100%|█| 256/256 [00:00<00:00, 7307.80it/s]\n",
"Epoch 244: Train, Loglike in nats: 5.122736: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.153957: 100%|█| 256/256 [00:00<00:00, 7497.25it/s]\n",
"Epoch 245: Train, Loglike in nats: 5.108798: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.181901: 100%|█| 256/256 [00:00<00:00, 7270.04it/s]\n",
"Epoch 246: Train, Loglike in nats: 5.126208: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.102291: 100%|█| 256/256 [00:00<00:00, 7216.78it/s]\n",
"Epoch 247: Train, Loglike in nats: 5.126986: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.186876: 100%|█| 256/256 [00:00<00:00, 7357.62it/s]\n",
"Epoch 248: Train, Loglike in nats: 5.134367: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.102363: 100%|█| 256/256 [00:00<00:00, 7323.25it/s]\n",
"Epoch 249: Train, Loglike in nats: 5.106711: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.206118: 100%|█| 256/256 [00:00<00:00, 7325.60it/s]\n",
"Epoch 250: Train, Loglike in nats: 5.075738: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.181466: 100%|█| 256/256 [00:00<00:00, 6652.14it/s]\n",
"Epoch 251: Train, Loglike in nats: 5.138174: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.117506: 100%|█| 256/256 [00:00<00:00, 7519.04it/s]\n",
"Epoch 252: Train, Loglike in nats: 5.104439: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.081573: 100%|█| 256/256 [00:00<00:00, 7447.49it/s]\n",
"Epoch 253: Train, Loglike in nats: 5.153634: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.205510: 100%|█| 256/256 [00:00<00:00, 7461.20it/s]\n",
"Epoch 254: Train, Loglike in nats: 5.143805: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.254789: 100%|█| 256/256 [00:00<00:00, 7462.66it/s]\n",
"Epoch 255: Train, Loglike in nats: 5.123546: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.200665: 100%|█| 256/256 [00:00<00:00, 7507.58it/s]\n",
"Epoch 256: Train, Loglike in nats: 5.177478: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.211371: 100%|█| 256/256 [00:00<00:00, 7220.23it/s]\n",
"Epoch 257: Train, Loglike in nats: 5.121754: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.106763: 100%|█| 256/256 [00:00<00:00, 6886.89it/s]\n",
"Epoch 258: Train, Loglike in nats: 5.167212: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.172981: 100%|█| 256/256 [00:00<00:00, 7445.48it/s]\n",
"Epoch 259: Train, Loglike in nats: 5.126055: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.169009: 100%|█| 256/256 [00:00<00:00, 7382.81it/s]\n",
"Epoch 260: Train, Loglike in nats: 5.150400: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.154891: 100%|█| 256/256 [00:00<00:00, 7690.68it/s]\n",
"Epoch 261: Train, Loglike in nats: 5.127478: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.113437: 100%|█| 256/256 [00:00<00:00, 7361.15it/s]\n",
"Epoch 262: Train, Loglike in nats: 5.146076: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.208222: 100%|█| 256/256 [00:00<00:00, 7639.08it/s]\n",
"Epoch 263: Train, Loglike in nats: 5.129072: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.155390: 100%|█| 256/256 [00:00<00:00, 6537.44it/s]\n",
"Epoch 264: Train, Loglike in nats: 5.156494: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.173073: 100%|█| 256/256 [00:00<00:00, 7292.21it/s]\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 265: Train, Loglike in nats: 5.163797: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.169765: 100%|█| 256/256 [00:00<00:00, 7081.80it/s]\n",
"Epoch 266: Train, Loglike in nats: 5.168819: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.263631: 100%|█| 256/256 [00:00<00:00, 7107.34it/s]\n",
"Epoch 267: Train, Loglike in nats: 5.158370: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.160183: 100%|█| 256/256 [00:00<00:00, 7384.54it/s]\n",
"Epoch 268: Train, Loglike in nats: 5.165156: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.183895: 100%|█| 256/256 [00:00<00:00, 7563.27it/s]\n",
"Epoch 269: Train, Loglike in nats: 5.146857: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.193723: 100%|█| 256/256 [00:00<00:00, 7254.03it/s]\n",
"Epoch 270: Train, Loglike in nats: 5.160407: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.196368: 100%|█| 256/256 [00:00<00:00, 7018.33it/s]\n",
"Epoch 271: Train, Loglike in nats: 5.167139: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.173479: 100%|█| 256/256 [00:00<00:00, 7325.80it/s]\n",
"Epoch 272: Train, Loglike in nats: 5.149775: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.257178: 100%|█| 256/256 [00:00<00:00, 7051.15it/s]\n",
"Epoch 273: Train, Loglike in nats: 5.142056: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.225978: 100%|█| 256/256 [00:00<00:00, 7036.22it/s]\n",
"Epoch 274: Train, Loglike in nats: 5.147622: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.198354: 100%|█| 256/256 [00:00<00:00, 7277.29it/s]\n",
"Epoch 275: Train, Loglike in nats: 5.151979: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.228982: 100%|█| 256/256 [00:00<00:00, 6993.79it/s]\n",
"Epoch 276: Train, Loglike in nats: 5.157323: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.244571: 100%|█| 256/256 [00:00<00:00, 7394.97it/s]\n",
"Epoch 277: Train, Loglike in nats: 5.173774: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.136812: 100%|█| 256/256 [00:00<00:00, 7471.38it/s]\n",
"Epoch 278: Train, Loglike in nats: 5.158896: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.059916: 100%|█| 256/256 [00:00<00:00, 7075.54it/s]\n",
"Epoch 279: Train, Loglike in nats: 5.152943: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.184673: 100%|█| 256/256 [00:00<00:00, 6883.71it/s]\n",
"Epoch 280: Train, Loglike in nats: 5.179606: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.204752: 100%|█| 256/256 [00:00<00:00, 7311.08it/s]\n",
"Epoch 281: Train, Loglike in nats: 5.170100: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.212417: 100%|█| 256/256 [00:00<00:00, 6885.08it/s]\n",
"Epoch 282: Train, Loglike in nats: 5.138464: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.184686: 100%|█| 256/256 [00:00<00:00, 7336.51it/s]\n",
"Epoch 283: Train, Loglike in nats: 5.171954: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.115675: 100%|█| 256/256 [00:00<00:00, 7474.92it/s]\n",
"Epoch 284: Train, Loglike in nats: 5.148740: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.175754: 100%|█| 256/256 [00:00<00:00, 7445.94it/s]\n",
"Epoch 285: Train, Loglike in nats: 5.161445: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.199004: 100%|█| 256/256 [00:00<00:00, 7134.31it/s]\n",
"Epoch 286: Train, Loglike in nats: 5.180827: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.148674: 100%|█| 256/256 [00:00<00:00, 7171.43it/s]\n",
"Epoch 287: Train, Loglike in nats: 5.151980: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.112621: 100%|█| 256/256 [00:00<00:00, 7166.88it/s]\n",
"Epoch 288: Train, Loglike in nats: 5.186673: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.162656: 100%|█| 256/256 [00:00<00:00, 6826.68it/s]\n",
"Epoch 289: Train, Loglike in nats: 5.174800: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.138060: 100%|█| 256/256 [00:00<00:00, 6726.95it/s]\n",
"Epoch 290: Train, Loglike in nats: 5.164673: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.187915: 100%|█| 256/256 [00:00<00:00, 7019.11it/s]\n",
"Epoch 291: Train, Loglike in nats: 5.160560: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.184336: 100%|█| 256/256 [00:00<00:00, 7123.99it/s]\n",
"Epoch 292: Train, Loglike in nats: 5.152422: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.210833: 100%|█| 256/256 [00:00<00:00, 6749.10it/s]\n",
"Epoch 293: Train, Loglike in nats: 5.127267: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.233580: 100%|█| 256/256 [00:00<00:00, 7171.00it/s]\n",
"Epoch 294: Train, Loglike in nats: 5.148996: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.155513: 100%|█| 256/256 [00:00<00:00, 6926.92it/s]\n",
"Epoch 295: Train, Loglike in nats: 5.161234: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.148299: 100%|█| 256/256 [00:00<00:00, 7294.09it/s]\n",
"Epoch 296: Train, Loglike in nats: 5.160137: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.261293: 100%|█| 256/256 [00:00<00:00, 7371.16it/s]\n",
"Epoch 297: Train, Loglike in nats: 5.185086: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.247223: 100%|█| 256/256 [00:00<00:00, 7659.41it/s]\n",
"Epoch 298: Train, Loglike in nats: 5.146867: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.186357: 100%|█| 256/256 [00:00<00:00, 7685.34it/s]\n",
"Epoch 299: Train, Loglike in nats: 5.156577: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.203336: 100%|█| 256/256 [00:00<00:00, 7704.14it/s]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Generating simulations\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"256it [00:00, 2968.50it/s]\n",
"WARNING:root:Too few points to create valid contours\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Effective sample size for current/all rounds 832.6/874.1\n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"---------------------- Round: 2 ----------------------\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 0: Train, Loglike in nats: -25.459032: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 0.716796: 100%|█| 256/256 [00:00<00:00, 1351.22it/s]\n",
"Epoch 1: Train, Loglike in nats: 1.816165: 100%|█| 2304/2304 [00:00<00:00, 3\n",
"- Val, Loglike in nats: 2.938455: 100%|█| 256/256 [00:00<00:00, 6964.07it/s]\n",
"Epoch 2: Train, Loglike in nats: 3.767795: 100%|█| 2304/2304 [00:00<00:00, 3\n",
"- Val, Loglike in nats: 4.019073: 100%|█| 256/256 [00:00<00:00, 7134.31it/s]\n",
"Epoch 3: Train, Loglike in nats: 3.963702: 100%|█| 2304/2304 [00:00<00:00, 3\n",
"- Val, Loglike in nats: 3.872475: 100%|█| 256/256 [00:00<00:00, 7504.43it/s]\n",
"Epoch 4: Train, Loglike in nats: 4.032058: 100%|█| 2304/2304 [00:00<00:00, 3\n",
"- Val, Loglike in nats: 4.027811: 100%|█| 256/256 [00:00<00:00, 7622.49it/s]\n",
"Epoch 5: Train, Loglike in nats: 4.110664: 100%|█| 2304/2304 [00:00<00:00, 3\n",
"- Val, Loglike in nats: 4.155464: 100%|█| 256/256 [00:00<00:00, 7589.03it/s]\n",
"Epoch 6: Train, Loglike in nats: 3.645802: 100%|█| 2304/2304 [00:00<00:00, 3\n",
"- Val, Loglike in nats: 3.652211: 100%|█| 256/256 [00:00<00:00, 7381.60it/s]\n",
"Epoch 7: Train, Loglike in nats: 3.993701: 100%|█| 2304/2304 [00:00<00:00, 3\n",
"- Val, Loglike in nats: 4.068994: 100%|█| 256/256 [00:00<00:00, 7692.06it/s]\n",
"Epoch 8: Train, Loglike in nats: 3.871829: 100%|█| 2304/2304 [00:00<00:00, 3\n",
"- Val, Loglike in nats: 2.962914: 100%|█| 256/256 [00:00<00:00, 7362.67it/s]\n",
"Epoch 9: Train, Loglike in nats: 3.788942: 100%|█| 2304/2304 [00:00<00:00, 3\n",
"- Val, Loglike in nats: 3.853029: 100%|█| 256/256 [00:00<00:00, 7348.31it/s]\n",
"Epoch 10: Train, Loglike in nats: 4.076739: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.330261: 100%|█| 256/256 [00:00<00:00, 7055.69it/s]\n",
"Epoch 11: Train, Loglike in nats: 4.015337: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.818393: 100%|█| 256/256 [00:00<00:00, 7074.56it/s]\n",
"Epoch 12: Train, Loglike in nats: 3.927076: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.837148: 100%|█| 256/256 [00:00<00:00, 7640.66it/s]\n",
"Epoch 13: Train, Loglike in nats: 4.027839: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.168886: 100%|█| 256/256 [00:00<00:00, 7754.72it/s]\n",
"Epoch 14: Train, Loglike in nats: 4.161034: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.196741: 100%|█| 256/256 [00:00<00:00, 6817.19it/s]\n",
"Epoch 15: Train, Loglike in nats: 4.160670: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.282686: 100%|█| 256/256 [00:00<00:00, 7199.12it/s]\n",
"Epoch 16: Train, Loglike in nats: 4.191583: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.225052: 100%|█| 256/256 [00:00<00:00, 7080.63it/s]\n",
"Epoch 17: Train, Loglike in nats: 4.277606: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.220875: 100%|█| 256/256 [00:00<00:00, 7280.99it/s]\n",
"Epoch 18: Train, Loglike in nats: 4.190850: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.346900: 100%|█| 256/256 [00:00<00:00, 7307.00it/s]\n",
"Epoch 19: Train, Loglike in nats: 4.216282: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.283581: 100%|█| 256/256 [00:00<00:00, 7467.02it/s]\n",
"Epoch 20: Train, Loglike in nats: 4.134266: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.185882: 100%|█| 256/256 [00:00<00:00, 7212.95it/s]\n",
"Epoch 21: Train, Loglike in nats: 3.997634: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.268462: 100%|█| 256/256 [00:00<00:00, 7552.79it/s]\n",
"Epoch 22: Train, Loglike in nats: 4.168719: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.337598: 100%|█| 256/256 [00:00<00:00, 7675.01it/s]\n",
"Epoch 23: Train, Loglike in nats: 4.205741: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.159449: 100%|█| 256/256 [00:00<00:00, 7698.07it/s]\n",
"Epoch 24: Train, Loglike in nats: 4.218801: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.149680: 100%|█| 256/256 [00:00<00:00, 7372.42it/s]\n",
"Epoch 25: Train, Loglike in nats: 4.212697: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.200192: 100%|█| 256/256 [00:00<00:00, 6942.95it/s]\n",
"Epoch 26: Train, Loglike in nats: 4.326603: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.346785: 100%|█| 256/256 [00:00<00:00, 7016.41it/s]\n",
"Epoch 27: Train, Loglike in nats: 4.271183: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.317245: 100%|█| 256/256 [00:00<00:00, 7195.65it/s]\n",
"Epoch 28: Train, Loglike in nats: 4.270976: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.308356: 100%|█| 256/256 [00:00<00:00, 7128.34it/s]\n",
"Epoch 29: Train, Loglike in nats: 4.235686: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.278309: 100%|█| 256/256 [00:00<00:00, 7133.31it/s]\n",
"Epoch 30: Train, Loglike in nats: 4.255909: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.134195: 100%|█| 256/256 [00:00<00:00, 7478.82it/s]\n",
"Epoch 31: Train, Loglike in nats: 4.155123: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.390512: 100%|█| 256/256 [00:00<00:00, 7189.82it/s]\n",
"Epoch 32: Train, Loglike in nats: 4.295762: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.168796: 100%|█| 256/256 [00:00<00:00, 6975.11it/s]\n",
"Epoch 33: Train, Loglike in nats: 4.149734: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.189861: 100%|█| 256/256 [00:00<00:00, 6938.60it/s]\n",
"Epoch 34: Train, Loglike in nats: 4.081147: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.286496: 100%|█| 256/256 [00:00<00:00, 7264.04it/s]\n",
"Epoch 35: Train, Loglike in nats: 4.284154: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.275317: 100%|█| 256/256 [00:00<00:00, 6957.62it/s]\n",
"Epoch 36: Train, Loglike in nats: 4.278087: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.305576: 100%|█| 256/256 [00:00<00:00, 7134.73it/s]\n",
"Epoch 37: Train, Loglike in nats: 4.237536: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.134045: 100%|█| 256/256 [00:00<00:00, 7243.71it/s]\n",
"Epoch 38: Train, Loglike in nats: 4.065331: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.244261: 100%|█| 256/256 [00:00<00:00, 6994.15it/s]\n",
"Epoch 39: Train, Loglike in nats: 4.031012: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.330646: 100%|█| 256/256 [00:00<00:00, 7187.99it/s]\n",
"Epoch 40: Train, Loglike in nats: 3.858065: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.224121: 100%|█| 256/256 [00:00<00:00, 6951.00it/s]\n",
"Epoch 41: Train, Loglike in nats: 4.246456: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.343836: 100%|█| 256/256 [00:00<00:00, 6944.08it/s]\n",
"Epoch 42: Train, Loglike in nats: 4.154380: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.020471: 100%|█| 256/256 [00:00<00:00, 6986.09it/s]\n",
"Epoch 43: Train, Loglike in nats: 4.299876: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.407318: 100%|█| 256/256 [00:00<00:00, 6982.37it/s]\n",
"Epoch 44: Train, Loglike in nats: 4.406497: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.413136: 100%|█| 256/256 [00:00<00:00, 6903.01it/s]\n",
"Epoch 45: Train, Loglike in nats: 4.393619: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.465865: 100%|█| 256/256 [00:00<00:00, 7166.21it/s]\n",
"Epoch 46: Train, Loglike in nats: 4.213091: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.308416: 100%|█| 256/256 [00:00<00:00, 7097.52it/s]\n",
"Epoch 47: Train, Loglike in nats: 4.110359: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.968832: 100%|█| 256/256 [00:00<00:00, 7168.84it/s]\n",
"Epoch 48: Train, Loglike in nats: 4.037476: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.417141: 100%|█| 256/256 [00:00<00:00, 6913.32it/s]\n",
"Epoch 49: Train, Loglike in nats: 4.274728: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.454271: 100%|█| 256/256 [00:00<00:00, 6829.94it/s]\n",
"Epoch 50: Train, Loglike in nats: 4.260391: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.253131: 100%|█| 256/256 [00:00<00:00, 6899.37it/s]\n",
"Epoch 51: Train, Loglike in nats: 3.777712: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 2.940146: 100%|█| 256/256 [00:00<00:00, 7249.28it/s]\n",
"Epoch 52: Train, Loglike in nats: 3.553122: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.019976: 100%|█| 256/256 [00:00<00:00, 7343.68it/s]\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 53: Train, Loglike in nats: 4.342750: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.482338: 100%|█| 256/256 [00:00<00:00, 7395.17it/s]\n",
"Epoch 54: Train, Loglike in nats: 4.363355: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.321406: 100%|█| 256/256 [00:00<00:00, 7216.39it/s]\n",
"Epoch 55: Train, Loglike in nats: 4.389521: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.506277: 100%|█| 256/256 [00:00<00:00, 7238.53it/s]\n",
"Epoch 56: Train, Loglike in nats: 4.442504: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.441913: 100%|█| 256/256 [00:00<00:00, 6860.18it/s]\n",
"Epoch 57: Train, Loglike in nats: 4.446343: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.528080: 100%|█| 256/256 [00:00<00:00, 6780.04it/s]\n",
"Epoch 58: Train, Loglike in nats: 4.398529: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.333248: 100%|█| 256/256 [00:00<00:00, 7196.03it/s]\n",
"Epoch 59: Train, Loglike in nats: 4.094214: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.507918: 100%|█| 256/256 [00:00<00:00, 7111.11it/s]\n",
"Epoch 60: Train, Loglike in nats: 4.185894: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.279068: 100%|█| 256/256 [00:00<00:00, 6841.47it/s]\n",
"Epoch 61: Train, Loglike in nats: 4.398495: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.427952: 100%|█| 256/256 [00:00<00:00, 7374.30it/s]\n",
"Epoch 62: Train, Loglike in nats: 4.371146: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.536351: 100%|█| 256/256 [00:00<00:00, 7373.69it/s]\n",
"Epoch 63: Train, Loglike in nats: 4.218025: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.235493: 100%|█| 256/256 [00:00<00:00, 6744.99it/s]\n",
"Epoch 64: Train, Loglike in nats: 3.175480: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.932615: 100%|█| 256/256 [00:00<00:00, 6793.34it/s]\n",
"Epoch 65: Train, Loglike in nats: 4.174471: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.202407: 100%|█| 256/256 [00:00<00:00, 7067.11it/s]\n",
"Epoch 66: Train, Loglike in nats: 4.304278: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.359468: 100%|█| 256/256 [00:00<00:00, 7021.59it/s]\n",
"Epoch 67: Train, Loglike in nats: 4.264257: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.129058: 100%|█| 256/256 [00:00<00:00, 7119.50it/s]\n",
"Epoch 68: Train, Loglike in nats: 4.313629: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.475007: 100%|█| 256/256 [00:00<00:00, 7116.91it/s]\n",
"Epoch 69: Train, Loglike in nats: 4.310327: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.398302: 100%|█| 256/256 [00:00<00:00, 7525.26it/s]\n",
"Epoch 70: Train, Loglike in nats: 4.390965: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.438315: 100%|█| 256/256 [00:00<00:00, 7156.99it/s]\n",
"Epoch 71: Train, Loglike in nats: 4.354778: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.349362: 100%|█| 256/256 [00:00<00:00, 6882.30it/s]\n",
"Epoch 72: Train, Loglike in nats: 4.372137: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.310641: 100%|█| 256/256 [00:00<00:00, 6938.78it/s]\n",
"Epoch 73: Train, Loglike in nats: 4.348265: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.427760: 100%|█| 256/256 [00:00<00:00, 6915.77it/s]\n",
"Epoch 74: Train, Loglike in nats: 4.298110: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.478752: 100%|█| 256/256 [00:00<00:00, 6765.52it/s]\n",
"Epoch 75: Train, Loglike in nats: 4.462824: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.478666: 100%|█| 256/256 [00:00<00:00, 7319.95it/s]\n",
"Epoch 76: Train, Loglike in nats: 4.403349: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.274703: 100%|█| 256/256 [00:00<00:00, 7136.92it/s]\n",
"Epoch 77: Train, Loglike in nats: 4.356835: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.414157: 100%|█| 256/256 [00:00<00:00, 7097.52it/s]\n",
"Epoch 78: Train, Loglike in nats: 4.387547: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.316439: 100%|█| 256/256 [00:00<00:00, 7243.90it/s]\n",
"Epoch 79: Train, Loglike in nats: 4.408562: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.153494: 100%|█| 256/256 [00:00<00:00, 7394.31it/s]\n",
"Epoch 80: Train, Loglike in nats: 4.334682: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.572682: 100%|█| 256/256 [00:00<00:00, 6920.45it/s]\n",
"Epoch 81: Train, Loglike in nats: 4.440049: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.425889: 100%|█| 256/256 [00:00<00:00, 6747.11it/s]\n",
"Epoch 82: Train, Loglike in nats: 4.376562: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.360289: 100%|█| 256/256 [00:00<00:00, 6831.20it/s]\n",
"Epoch 83: Train, Loglike in nats: 4.436251: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.424048: 100%|█| 256/256 [00:00<00:00, 7158.23it/s]\n",
"Epoch 84: Train, Loglike in nats: 4.411824: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.485960: 100%|█| 256/256 [00:00<00:00, 6871.20it/s]\n",
"Epoch 85: Train, Loglike in nats: 4.428925: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.540188: 100%|█| 256/256 [00:00<00:00, 7150.22it/s]\n",
"Epoch 86: Train, Loglike in nats: 4.453360: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.578872: 100%|█| 256/256 [00:00<00:00, 7080.21it/s]\n",
"Epoch 87: Train, Loglike in nats: 4.497378: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.507009: 100%|█| 256/256 [00:00<00:00, 6860.31it/s]\n",
"Epoch 88: Train, Loglike in nats: 4.459049: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.503995: 100%|█| 256/256 [00:00<00:00, 7383.07it/s]\n",
"Epoch 89: Train, Loglike in nats: 4.460312: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.429508: 100%|█| 256/256 [00:00<00:00, 6975.84it/s]\n",
"Epoch 90: Train, Loglike in nats: 4.229008: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.471846: 100%|█| 256/256 [00:00<00:00, 6882.08it/s]\n",
"Epoch 91: Train, Loglike in nats: 4.401881: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.413384: 100%|█| 256/256 [00:00<00:00, 7392.01it/s]\n",
"Epoch 92: Train, Loglike in nats: 4.441780: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.487418: 100%|█| 256/256 [00:00<00:00, 7172.48it/s]\n",
"Epoch 93: Train, Loglike in nats: 4.333367: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.381265: 100%|█| 256/256 [00:00<00:00, 7363.27it/s]\n",
"Epoch 94: Train, Loglike in nats: 4.354813: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.307774: 100%|█| 256/256 [00:00<00:00, 7229.85it/s]\n",
"Epoch 95: Train, Loglike in nats: 4.456592: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.496818: 100%|█| 256/256 [00:00<00:00, 7201.10it/s]\n",
"Epoch 96: Train, Loglike in nats: 4.472215: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.544455: 100%|█| 256/256 [00:00<00:00, 7270.83it/s]\n",
"Epoch 97: Train, Loglike in nats: 4.485637: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.525896: 100%|█| 256/256 [00:00<00:00, 6695.78it/s]\n",
"Epoch 98: Train, Loglike in nats: 4.486530: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.595785: 100%|█| 256/256 [00:00<00:00, 6998.94it/s]\n",
"Epoch 99: Train, Loglike in nats: 4.505122: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.633898: 100%|█| 256/256 [00:00<00:00, 6894.85it/s]\n",
"Epoch 100: Train, Loglike in nats: 4.478048: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.492528: 100%|█| 256/256 [00:00<00:00, 7030.46it/s]\n",
"Epoch 101: Train, Loglike in nats: 4.346496: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.432290: 100%|█| 256/256 [00:00<00:00, 7077.27it/s]\n",
"Epoch 102: Train, Loglike in nats: 4.246643: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.041471: 100%|█| 256/256 [00:00<00:00, 6970.18it/s]\n",
"Epoch 103: Train, Loglike in nats: 4.245991: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.494583: 100%|█| 256/256 [00:00<00:00, 7093.16it/s]\n",
"Epoch 104: Train, Loglike in nats: 4.366352: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.602152: 100%|█| 256/256 [00:00<00:00, 7102.22it/s]\n",
"Epoch 105: Train, Loglike in nats: 4.543465: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.552619: 100%|█| 256/256 [00:00<00:00, 6626.11it/s]\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 106: Train, Loglike in nats: 4.551825: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.550610: 100%|█| 256/256 [00:00<00:00, 7086.70it/s]\n",
"Epoch 107: Train, Loglike in nats: 4.387917: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.460163: 100%|█| 256/256 [00:00<00:00, 7424.62it/s]\n",
"Epoch 108: Train, Loglike in nats: 4.390234: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.568196: 100%|█| 256/256 [00:00<00:00, 7318.70it/s]\n",
"Epoch 109: Train, Loglike in nats: 4.547569: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.484544: 100%|█| 256/256 [00:00<00:00, 7530.96it/s]\n",
"Epoch 110: Train, Loglike in nats: 4.574683: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.680196: 100%|█| 256/256 [00:00<00:00, 7685.39it/s]\n",
"Epoch 111: Train, Loglike in nats: 4.556639: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.572621: 100%|█| 256/256 [00:00<00:00, 7163.05it/s]\n",
"Epoch 112: Train, Loglike in nats: 4.444817: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.281064: 100%|█| 256/256 [00:00<00:00, 7220.81it/s]\n",
"Epoch 113: Train, Loglike in nats: 4.513227: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.576050: 100%|█| 256/256 [00:00<00:00, 6993.19it/s]\n",
"Epoch 114: Train, Loglike in nats: 4.556698: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.611020: 100%|█| 256/256 [00:00<00:00, 6960.69it/s]\n",
"Epoch 115: Train, Loglike in nats: 4.482500: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.572237: 100%|█| 256/256 [00:00<00:00, 7431.77it/s]\n",
"Epoch 116: Train, Loglike in nats: 4.456186: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.655474: 100%|█| 256/256 [00:00<00:00, 7649.59it/s]\n",
"Epoch 117: Train, Loglike in nats: 4.539685: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.389518: 100%|█| 256/256 [00:00<00:00, 7395.02it/s]\n",
"Epoch 118: Train, Loglike in nats: 4.500516: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.401004: 100%|█| 256/256 [00:00<00:00, 7670.68it/s]\n",
"Epoch 119: Train, Loglike in nats: 4.529864: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.557676: 100%|█| 256/256 [00:00<00:00, 7647.73it/s]\n",
"Epoch 120: Train, Loglike in nats: 4.468634: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.548501: 100%|█| 256/256 [00:00<00:00, 7038.95it/s]\n",
"Epoch 121: Train, Loglike in nats: 4.563500: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.578938: 100%|█| 256/256 [00:00<00:00, 6985.73it/s]\n",
"Epoch 122: Train, Loglike in nats: 4.608435: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.721342: 100%|█| 256/256 [00:00<00:00, 7192.42it/s]\n",
"Epoch 123: Train, Loglike in nats: 4.580320: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.665948: 100%|█| 256/256 [00:00<00:00, 7351.02it/s]\n",
"Epoch 124: Train, Loglike in nats: 4.532303: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.632061: 100%|█| 256/256 [00:00<00:00, 7418.57it/s]\n",
"Epoch 125: Train, Loglike in nats: 4.606361: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.534810: 100%|█| 256/256 [00:00<00:00, 7414.73it/s]\n",
"Epoch 126: Train, Loglike in nats: 4.556992: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.507894: 100%|█| 256/256 [00:00<00:00, 7521.47it/s]\n",
"Epoch 127: Train, Loglike in nats: 4.586454: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.712366: 100%|█| 256/256 [00:00<00:00, 7189.58it/s]\n",
"Epoch 128: Train, Loglike in nats: 4.546159: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.611938: 100%|█| 256/256 [00:00<00:00, 7467.95it/s]\n",
"Epoch 129: Train, Loglike in nats: 4.439271: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.402538: 100%|█| 256/256 [00:00<00:00, 6923.79it/s]\n",
"Epoch 130: Train, Loglike in nats: 4.560505: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.643305: 100%|█| 256/256 [00:00<00:00, 6947.04it/s]\n",
"Epoch 131: Train, Loglike in nats: 4.563086: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.585982: 100%|█| 256/256 [00:00<00:00, 7034.89it/s]\n",
"Epoch 132: Train, Loglike in nats: 4.550482: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.686270: 100%|█| 256/256 [00:00<00:00, 6998.62it/s]\n",
"Epoch 133: Train, Loglike in nats: 4.442138: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.416639: 100%|█| 256/256 [00:00<00:00, 7178.91it/s]\n",
"Epoch 134: Train, Loglike in nats: 4.222669: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.360689: 100%|█| 256/256 [00:00<00:00, 7434.55it/s]\n",
"Epoch 135: Train, Loglike in nats: 4.293054: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.129561: 100%|█| 256/256 [00:00<00:00, 7182.89it/s]\n",
"Epoch 136: Train, Loglike in nats: 4.519616: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.504741: 100%|█| 256/256 [00:00<00:00, 7097.52it/s]\n",
"Epoch 137: Train, Loglike in nats: 4.533878: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.677875: 100%|█| 256/256 [00:00<00:00, 7451.37it/s]\n",
"Epoch 138: Train, Loglike in nats: 4.656576: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.739008: 100%|█| 256/256 [00:00<00:00, 7105.93it/s]\n",
"Epoch 139: Train, Loglike in nats: 4.671229: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.676918: 100%|█| 256/256 [00:00<00:00, 7023.47it/s]\n",
"Epoch 140: Train, Loglike in nats: 4.676200: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.684293: 100%|█| 256/256 [00:00<00:00, 7212.52it/s]\n",
"Epoch 141: Train, Loglike in nats: 4.588891: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.553392: 100%|█| 256/256 [00:00<00:00, 6866.59it/s]\n",
"Epoch 142: Train, Loglike in nats: 4.572881: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.708003: 100%|█| 256/256 [00:00<00:00, 7046.29it/s]\n",
"Epoch 143: Train, Loglike in nats: 4.680336: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.651568: 100%|█| 256/256 [00:00<00:00, 7356.51it/s]\n",
"Epoch 144: Train, Loglike in nats: 4.633306: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.630590: 100%|█| 256/256 [00:00<00:00, 7171.28it/s]\n",
"Epoch 145: Train, Loglike in nats: 4.517875: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.712124: 100%|█| 256/256 [00:00<00:00, 7206.80it/s]\n",
"Epoch 146: Train, Loglike in nats: 4.617796: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.585333: 100%|█| 256/256 [00:00<00:00, 7511.73it/s]\n",
"Epoch 147: Train, Loglike in nats: 4.565910: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.725930: 100%|█| 256/256 [00:00<00:00, 7219.45it/s]\n",
"Epoch 148: Train, Loglike in nats: 4.627827: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.651860: 100%|█| 256/256 [00:00<00:00, 7304.71it/s]\n",
"Epoch 149: Train, Loglike in nats: 4.667055: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.736513: 100%|█| 256/256 [00:00<00:00, 7077.50it/s]\n",
"Epoch 150: Train, Loglike in nats: 4.628793: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.659184: 100%|█| 256/256 [00:00<00:00, 7264.88it/s]\n",
"Epoch 151: Train, Loglike in nats: 4.675322: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.702706: 100%|█| 256/256 [00:00<00:00, 7119.17it/s]\n",
"Epoch 152: Train, Loglike in nats: 4.658490: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.599602: 100%|█| 256/256 [00:00<00:00, 7242.88it/s]\n",
"Epoch 153: Train, Loglike in nats: 4.668788: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.724724: 100%|█| 256/256 [00:00<00:00, 7219.60it/s]\n",
"Epoch 154: Train, Loglike in nats: 4.626144: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.700100: 100%|█| 256/256 [00:00<00:00, 7323.50it/s]\n",
"Epoch 155: Train, Loglike in nats: 4.593838: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.745340: 100%|█| 256/256 [00:00<00:00, 7048.28it/s]\n",
"Epoch 156: Train, Loglike in nats: 4.697848: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.755816: 100%|█| 256/256 [00:00<00:00, 6970.72it/s]\n",
"Epoch 157: Train, Loglike in nats: 4.710353: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.692124: 100%|█| 256/256 [00:00<00:00, 6988.41it/s]\n",
"Epoch 158: Train, Loglike in nats: 4.563360: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.593971: 100%|█| 256/256 [00:00<00:00, 7050.22it/s]\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 159: Train, Loglike in nats: 4.617407: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.748337: 100%|█| 256/256 [00:00<00:00, 7138.05it/s]\n",
"Epoch 160: Train, Loglike in nats: 4.683209: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.714194: 100%|█| 256/256 [00:00<00:00, 7395.27it/s]\n",
"Epoch 161: Train, Loglike in nats: 4.758719: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.775281: 100%|█| 256/256 [00:00<00:00, 7218.77it/s]\n",
"Epoch 162: Train, Loglike in nats: 4.698996: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.720401: 100%|█| 256/256 [00:00<00:00, 7337.36it/s]\n",
"Epoch 163: Train, Loglike in nats: 4.685947: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.656868: 100%|█| 256/256 [00:00<00:00, 7658.48it/s]\n",
"Epoch 164: Train, Loglike in nats: 4.625983: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.719793: 100%|█| 256/256 [00:00<00:00, 7166.21it/s]\n",
"Epoch 165: Train, Loglike in nats: 4.623974: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.793547: 100%|█| 256/256 [00:00<00:00, 7296.77it/s]\n",
"Epoch 166: Train, Loglike in nats: 4.707324: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.729740: 100%|█| 256/256 [00:00<00:00, 7019.30it/s]\n",
"Epoch 167: Train, Loglike in nats: 4.699100: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.714632: 100%|█| 256/256 [00:00<00:00, 7040.33it/s]\n",
"Epoch 168: Train, Loglike in nats: 4.756079: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.839397: 100%|█| 256/256 [00:00<00:00, 6954.47it/s]\n",
"Epoch 169: Train, Loglike in nats: 4.684510: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.763081: 100%|█| 256/256 [00:00<00:00, 7609.52it/s]\n",
"Epoch 170: Train, Loglike in nats: 4.718971: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.857027: 100%|█| 256/256 [00:00<00:00, 7461.77it/s]\n",
"Epoch 171: Train, Loglike in nats: 4.633773: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.685323: 100%|█| 256/256 [00:00<00:00, 7470.03it/s]\n",
"Epoch 172: Train, Loglike in nats: 4.613327: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.654237: 100%|█| 256/256 [00:00<00:00, 7586.08it/s]\n",
"Epoch 173: Train, Loglike in nats: 4.739897: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.742795: 100%|█| 256/256 [00:00<00:00, 7202.12it/s]\n",
"Epoch 174: Train, Loglike in nats: 4.713966: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.818064: 100%|█| 256/256 [00:00<00:00, 7062.23it/s]\n",
"Epoch 175: Train, Loglike in nats: 4.730965: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.659272: 100%|█| 256/256 [00:00<00:00, 7077.92it/s]\n",
"Epoch 176: Train, Loglike in nats: 4.753417: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.769886: 100%|█| 256/256 [00:00<00:00, 7000.26it/s]\n",
"Epoch 177: Train, Loglike in nats: 4.663913: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.757805: 100%|█| 256/256 [00:00<00:00, 6957.26it/s]\n",
"Epoch 178: Train, Loglike in nats: 4.670643: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.615166: 100%|█| 256/256 [00:00<00:00, 7381.14it/s]\n",
"Epoch 179: Train, Loglike in nats: 4.734698: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.681443: 100%|█| 256/256 [00:00<00:00, 7462.24it/s]\n",
"Epoch 180: Train, Loglike in nats: 4.728320: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.766556: 100%|█| 256/256 [00:00<00:00, 7562.57it/s]\n",
"Epoch 181: Train, Loglike in nats: 4.744843: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.778620: 100%|█| 256/256 [00:00<00:00, 7395.02it/s]\n",
"Epoch 182: Train, Loglike in nats: 4.715136: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.793665: 100%|█| 256/256 [00:00<00:00, 7264.88it/s]\n",
"Epoch 183: Train, Loglike in nats: 4.690339: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.787398: 100%|█| 256/256 [00:00<00:00, 6929.38it/s]\n",
"Epoch 184: Train, Loglike in nats: 4.762077: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.802665: 100%|█| 256/256 [00:00<00:00, 7123.33it/s]\n",
"Epoch 185: Train, Loglike in nats: 4.740115: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.858251: 100%|█| 256/256 [00:00<00:00, 7128.77it/s]\n",
"Epoch 186: Train, Loglike in nats: 4.780403: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.793300: 100%|█| 256/256 [00:00<00:00, 7011.96it/s]\n",
"Epoch 187: Train, Loglike in nats: 4.747309: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.833971: 100%|█| 256/256 [00:00<00:00, 7154.18it/s]\n",
"Epoch 188: Train, Loglike in nats: 4.767345: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.774062: 100%|█| 256/256 [00:00<00:00, 7219.79it/s]\n",
"Epoch 189: Train, Loglike in nats: 4.728542: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.710522: 100%|█| 256/256 [00:00<00:00, 7635.44it/s]\n",
"Epoch 190: Train, Loglike in nats: 4.754664: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.782536: 100%|█| 256/256 [00:00<00:00, 7116.86it/s]\n",
"Epoch 191: Train, Loglike in nats: 4.741723: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.585524: 100%|█| 256/256 [00:00<00:00, 7150.46it/s]\n",
"Epoch 192: Train, Loglike in nats: 4.726345: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.567327: 100%|█| 256/256 [00:00<00:00, 6772.69it/s]\n",
"Epoch 193: Train, Loglike in nats: 4.703058: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.692782: 100%|█| 256/256 [00:00<00:00, 6847.28it/s]\n",
"Epoch 194: Train, Loglike in nats: 4.706592: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.642871: 100%|█| 256/256 [00:00<00:00, 7210.87it/s]\n",
"Epoch 195: Train, Loglike in nats: 4.706992: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.749754: 100%|█| 256/256 [00:00<00:00, 7498.72it/s]\n",
"Epoch 196: Train, Loglike in nats: 4.716857: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.799282: 100%|█| 256/256 [00:00<00:00, 7269.21it/s]\n",
"Epoch 197: Train, Loglike in nats: 4.744892: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.752505: 100%|█| 256/256 [00:00<00:00, 6991.87it/s]\n",
"Epoch 198: Train, Loglike in nats: 4.802556: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.759777: 100%|█| 256/256 [00:00<00:00, 6946.68it/s]\n",
"Epoch 199: Train, Loglike in nats: 4.775708: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.745548: 100%|█| 256/256 [00:00<00:00, 6697.41it/s]\n",
"Epoch 200: Train, Loglike in nats: 4.737733: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.688066: 100%|█| 256/256 [00:00<00:00, 7143.99it/s]\n",
"Epoch 201: Train, Loglike in nats: 4.764923: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.741889: 100%|█| 256/256 [00:00<00:00, 6443.17it/s]\n",
"Epoch 202: Train, Loglike in nats: 4.791975: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.947197: 100%|█| 256/256 [00:00<00:00, 6858.30it/s]\n",
"Epoch 203: Train, Loglike in nats: 4.774905: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.867058: 100%|█| 256/256 [00:00<00:00, 7260.95it/s]\n",
"Epoch 204: Train, Loglike in nats: 4.721918: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.797095: 100%|█| 256/256 [00:00<00:00, 6910.34it/s]\n",
"Epoch 205: Train, Loglike in nats: 4.757875: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.796571: 100%|█| 256/256 [00:00<00:00, 7313.42it/s]\n",
"Epoch 206: Train, Loglike in nats: 4.787943: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.828769: 100%|█| 256/256 [00:00<00:00, 7104.80it/s]\n",
"Epoch 207: Train, Loglike in nats: 4.780513: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.793428: 100%|█| 256/256 [00:00<00:00, 6721.14it/s]\n",
"Epoch 208: Train, Loglike in nats: 4.802980: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.704525: 100%|█| 256/256 [00:00<00:00, 6579.62it/s]\n",
"Epoch 209: Train, Loglike in nats: 4.761561: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.738401: 100%|█| 256/256 [00:00<00:00, 6956.54it/s]\n",
"Epoch 210: Train, Loglike in nats: 4.719154: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.782625: 100%|█| 256/256 [00:00<00:00, 6079.29it/s]\n",
"Epoch 211: Train, Loglike in nats: 4.791929: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.890286: 100%|█| 256/256 [00:00<00:00, 5768.46it/s]\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 212: Train, Loglike in nats: 4.791015: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.843431: 100%|█| 256/256 [00:00<00:00, 6860.31it/s]\n",
"Epoch 213: Train, Loglike in nats: 4.828598: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.838082: 100%|█| 256/256 [00:00<00:00, 7258.30it/s]\n",
"Epoch 214: Train, Loglike in nats: 4.775691: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.769201: 100%|█| 256/256 [00:00<00:00, 7301.78it/s]\n",
"Epoch 215: Train, Loglike in nats: 4.811089: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.845944: 100%|█| 256/256 [00:00<00:00, 7051.38it/s]\n",
"Epoch 216: Train, Loglike in nats: 4.803718: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.749101: 100%|█| 256/256 [00:00<00:00, 7175.88it/s]\n",
"Epoch 217: Train, Loglike in nats: 4.814103: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.904918: 100%|█| 256/256 [00:00<00:00, 7270.44it/s]\n",
"Epoch 218: Train, Loglike in nats: 4.810892: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.854766: 100%|█| 256/256 [00:00<00:00, 7411.45it/s]\n",
"Epoch 219: Train, Loglike in nats: 4.849734: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.791654: 100%|█| 256/256 [00:00<00:00, 7396.29it/s]\n",
"Epoch 220: Train, Loglike in nats: 4.773797: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.812427: 100%|█| 256/256 [00:00<00:00, 7094.01it/s]\n",
"Epoch 221: Train, Loglike in nats: 4.800320: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.796589: 100%|█| 256/256 [00:00<00:00, 6970.54it/s]\n",
"Epoch 222: Train, Loglike in nats: 4.805697: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.763232: 100%|█| 256/256 [00:00<00:00, 6855.76it/s]\n",
"Epoch 223: Train, Loglike in nats: 4.784628: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.870389: 100%|█| 256/256 [00:00<00:00, 7452.71it/s]\n",
"Epoch 224: Train, Loglike in nats: 4.836376: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.840377: 100%|█| 256/256 [00:00<00:00, 7538.05it/s]\n",
"Epoch 225: Train, Loglike in nats: 4.786961: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.804414: 100%|█| 256/256 [00:00<00:00, 7522.78it/s]\n",
"Epoch 226: Train, Loglike in nats: 4.807696: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.806514: 100%|█| 256/256 [00:00<00:00, 7561.19it/s]\n",
"Epoch 227: Train, Loglike in nats: 4.820759: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.864990: 100%|█| 256/256 [00:00<00:00, 7714.33it/s]\n",
"Epoch 228: Train, Loglike in nats: 4.809407: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.887897: 100%|█| 256/256 [00:00<00:00, 7280.44it/s]\n",
"Epoch 229: Train, Loglike in nats: 4.837718: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.869583: 100%|█| 256/256 [00:00<00:00, 7524.47it/s]\n",
"Epoch 230: Train, Loglike in nats: 4.765549: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.856167: 100%|█| 256/256 [00:00<00:00, 7221.20it/s]\n",
"Epoch 231: Train, Loglike in nats: 4.817754: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.792692: 100%|█| 256/256 [00:00<00:00, 7524.15it/s]\n",
"Epoch 232: Train, Loglike in nats: 4.817926: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.865110: 100%|█| 256/256 [00:00<00:00, 7695.75it/s]\n",
"Epoch 233: Train, Loglike in nats: 4.803393: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.823866: 100%|█| 256/256 [00:00<00:00, 7665.37it/s]\n",
"Epoch 234: Train, Loglike in nats: 4.809716: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.768380: 100%|█| 256/256 [00:00<00:00, 7732.05it/s]\n",
"Epoch 235: Train, Loglike in nats: 4.805282: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.789787: 100%|█| 256/256 [00:00<00:00, 7712.67it/s]\n",
"Epoch 236: Train, Loglike in nats: 4.819271: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.893731: 100%|█| 256/256 [00:00<00:00, 7578.00it/s]\n",
"Epoch 237: Train, Loglike in nats: 4.860694: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.866067: 100%|█| 256/256 [00:00<00:00, 7447.90it/s]\n",
"Epoch 238: Train, Loglike in nats: 4.851028: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.828249: 100%|█| 256/256 [00:00<00:00, 7291.96it/s]\n",
"Epoch 239: Train, Loglike in nats: 4.821402: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.855254: 100%|█| 256/256 [00:00<00:00, 7527.42it/s]\n",
"Epoch 240: Train, Loglike in nats: 4.830735: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.796936: 100%|█| 256/256 [00:00<00:00, 7384.80it/s]\n",
"Epoch 241: Train, Loglike in nats: 4.848983: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.883338: 100%|█| 256/256 [00:00<00:00, 7595.10it/s]\n",
"Epoch 242: Train, Loglike in nats: 4.853800: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.961061: 100%|█| 256/256 [00:00<00:00, 7761.56it/s]\n",
"Epoch 243: Train, Loglike in nats: 4.838485: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.889397: 100%|█| 256/256 [00:00<00:00, 7675.01it/s]\n",
"Epoch 244: Train, Loglike in nats: 4.840094: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.884178: 100%|█| 256/256 [00:00<00:00, 7436.30it/s]\n",
"Epoch 245: Train, Loglike in nats: 4.843608: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.896363: 100%|█| 256/256 [00:00<00:00, 7100.62it/s]\n",
"Epoch 246: Train, Loglike in nats: 4.827901: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.865753: 100%|█| 256/256 [00:00<00:00, 6716.34it/s]\n",
"Epoch 247: Train, Loglike in nats: 4.872555: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.925087: 100%|█| 256/256 [00:00<00:00, 7156.08it/s]\n",
"Epoch 248: Train, Loglike in nats: 4.872678: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.903154: 100%|█| 256/256 [00:00<00:00, 6961.46it/s]\n",
"Epoch 249: Train, Loglike in nats: 4.852192: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.876648: 100%|█| 256/256 [00:00<00:00, 7194.83it/s]\n",
"Epoch 250: Train, Loglike in nats: 4.808098: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.883813: 100%|█| 256/256 [00:00<00:00, 6944.30it/s]\n",
"Epoch 251: Train, Loglike in nats: 4.846334: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.809102: 100%|█| 256/256 [00:00<00:00, 7051.19it/s]\n",
"Epoch 252: Train, Loglike in nats: 4.823839: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.847595: 100%|█| 256/256 [00:00<00:00, 6974.16it/s]\n",
"Epoch 253: Train, Loglike in nats: 4.840242: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.874866: 100%|█| 256/256 [00:00<00:00, 6789.35it/s]\n",
"Epoch 254: Train, Loglike in nats: 4.838672: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.911481: 100%|█| 256/256 [00:00<00:00, 6948.08it/s]\n",
"Epoch 255: Train, Loglike in nats: 4.843149: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.881008: 100%|█| 256/256 [00:00<00:00, 7160.28it/s]\n",
"Epoch 256: Train, Loglike in nats: 4.875273: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.907650: 100%|█| 256/256 [00:00<00:00, 6743.42it/s]\n",
"Epoch 257: Train, Loglike in nats: 4.843125: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.834339: 100%|█| 256/256 [00:00<00:00, 6912.38it/s]\n",
"Epoch 258: Train, Loglike in nats: 4.877197: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.887118: 100%|█| 256/256 [00:00<00:00, 6960.69it/s]\n",
"Epoch 259: Train, Loglike in nats: 4.845052: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.876871: 100%|█| 256/256 [00:00<00:00, 6924.95it/s]\n",
"Epoch 260: Train, Loglike in nats: 4.878731: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.888623: 100%|█| 256/256 [00:00<00:00, 6750.29it/s]\n",
"Epoch 261: Train, Loglike in nats: 4.847645: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.865180: 100%|█| 256/256 [00:00<00:00, 6772.13it/s]\n",
"Epoch 262: Train, Loglike in nats: 4.869785: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.931800: 100%|█| 256/256 [00:00<00:00, 6877.63it/s]\n",
"Epoch 263: Train, Loglike in nats: 4.863113: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.900752: 100%|█| 256/256 [00:00<00:00, 6767.48it/s]\n",
"Epoch 264: Train, Loglike in nats: 4.870215: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.864139: 100%|█| 256/256 [00:00<00:00, 6190.14it/s]\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 265: Train, Loglike in nats: 4.872549: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.920706: 100%|█| 256/256 [00:00<00:00, 6733.99it/s]\n",
"Epoch 266: Train, Loglike in nats: 4.873785: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.993923: 100%|█| 256/256 [00:00<00:00, 7380.94it/s]\n",
"Epoch 267: Train, Loglike in nats: 4.878537: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.871083: 100%|█| 256/256 [00:00<00:00, 7297.61it/s]\n",
"Epoch 268: Train, Loglike in nats: 4.872050: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.880050: 100%|█| 256/256 [00:00<00:00, 6932.20it/s]\n",
"Epoch 269: Train, Loglike in nats: 4.860022: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.914714: 100%|█| 256/256 [00:00<00:00, 6968.05it/s]\n",
"Epoch 270: Train, Loglike in nats: 4.867594: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.940704: 100%|█| 256/256 [00:00<00:00, 6755.86it/s]\n",
"Epoch 271: Train, Loglike in nats: 4.877998: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.887568: 100%|█| 256/256 [00:00<00:00, 7434.96it/s]\n",
"Epoch 272: Train, Loglike in nats: 4.859782: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.921072: 100%|█| 256/256 [00:00<00:00, 7000.26it/s]\n",
"Epoch 273: Train, Loglike in nats: 4.866823: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.944920: 100%|█| 256/256 [00:00<00:00, 6926.38it/s]\n",
"Epoch 274: Train, Loglike in nats: 4.876299: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.920372: 100%|█| 256/256 [00:00<00:00, 7201.34it/s]\n",
"Epoch 275: Train, Loglike in nats: 4.858252: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.941661: 100%|█| 256/256 [00:00<00:00, 7293.05it/s]\n",
"Epoch 276: Train, Loglike in nats: 4.851760: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.926747: 100%|█| 256/256 [00:00<00:00, 6778.25it/s]\n",
"Epoch 277: Train, Loglike in nats: 4.897157: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.863073: 100%|█| 256/256 [00:00<00:00, 6747.15it/s]\n",
"Epoch 278: Train, Loglike in nats: 4.871885: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.803133: 100%|█| 256/256 [00:00<00:00, 6960.69it/s]\n",
"Epoch 279: Train, Loglike in nats: 4.851917: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.947403: 100%|█| 256/256 [00:00<00:00, 6725.85it/s]\n",
"Epoch 280: Train, Loglike in nats: 4.887821: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.908969: 100%|█| 256/256 [00:00<00:00, 7295.93it/s]\n",
"Epoch 281: Train, Loglike in nats: 4.879801: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.930233: 100%|█| 256/256 [00:00<00:00, 7011.78it/s]\n",
"Epoch 282: Train, Loglike in nats: 4.855300: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.889888: 100%|█| 256/256 [00:00<00:00, 7344.94it/s]\n",
"Epoch 283: Train, Loglike in nats: 4.884274: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.826675: 100%|█| 256/256 [00:00<00:00, 6856.85it/s]\n",
"Epoch 284: Train, Loglike in nats: 4.876749: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.901900: 100%|█| 256/256 [00:00<00:00, 7211.45it/s]\n",
"Epoch 285: Train, Loglike in nats: 4.867153: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.936841: 100%|█| 256/256 [00:00<00:00, 6757.13it/s]\n",
"Epoch 286: Train, Loglike in nats: 4.879673: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.882199: 100%|█| 256/256 [00:00<00:00, 6971.49it/s]\n",
"Epoch 287: Train, Loglike in nats: 4.860911: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.806086: 100%|█| 256/256 [00:00<00:00, 7022.79it/s]\n",
"Epoch 288: Train, Loglike in nats: 4.887481: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.871099: 100%|█| 256/256 [00:00<00:00, 7152.03it/s]\n",
"Epoch 289: Train, Loglike in nats: 4.886433: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.864386: 100%|█| 256/256 [00:00<00:00, 7095.55it/s]\n",
"Epoch 290: Train, Loglike in nats: 4.872854: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.877355: 100%|█| 256/256 [00:00<00:00, 7314.67it/s]\n",
"Epoch 291: Train, Loglike in nats: 4.867135: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.947392: 100%|█| 256/256 [00:00<00:00, 7001.81it/s]\n",
"Epoch 292: Train, Loglike in nats: 4.858369: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.927133: 100%|█| 256/256 [00:00<00:00, 7379.87it/s]\n",
"Epoch 293: Train, Loglike in nats: 4.850555: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.929248: 100%|█| 256/256 [00:00<00:00, 6640.25it/s]\n",
"Epoch 294: Train, Loglike in nats: 4.865536: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.871136: 100%|█| 256/256 [00:00<00:00, 7134.26it/s]\n",
"Epoch 295: Train, Loglike in nats: 4.879278: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.866585: 100%|█| 256/256 [00:00<00:00, 6871.90it/s]\n",
"Epoch 296: Train, Loglike in nats: 4.875204: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.968857: 100%|█| 256/256 [00:00<00:00, 7205.84it/s]\n",
"Epoch 297: Train, Loglike in nats: 4.891515: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.951466: 100%|█| 256/256 [00:00<00:00, 7109.74it/s]\n",
"Epoch 298: Train, Loglike in nats: 4.865585: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.940246: 100%|█| 256/256 [00:00<00:00, 6996.43it/s]\n",
"Epoch 299: Train, Loglike in nats: 4.873398: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.907986: 100%|█| 256/256 [00:00<00:00, 6986.91it/s]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Generating simulations\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"256it [00:00, 2977.96it/s]\n",
"WARNING:root:Too few points to create valid contours\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Effective sample size for current/all rounds 981.6/1855.7\n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"---------------------- Round: 3 ----------------------\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 0: Train, Loglike in nats: -2.254916: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 3.607208: 100%|█| 256/256 [00:00<00:00, 1334.22it/s]\n",
"Epoch 1: Train, Loglike in nats: 3.961391: 100%|█| 2304/2304 [00:00<00:00, 3\n",
"- Val, Loglike in nats: 4.249711: 100%|█| 256/256 [00:00<00:00, 6990.96it/s]\n",
"Epoch 2: Train, Loglike in nats: 4.261970: 100%|█| 2304/2304 [00:00<00:00, 3\n",
"- Val, Loglike in nats: 4.421974: 100%|█| 256/256 [00:00<00:00, 7016.22it/s]\n",
"Epoch 3: Train, Loglike in nats: 4.304015: 100%|█| 2304/2304 [00:00<00:00, 3\n",
"- Val, Loglike in nats: 4.360326: 100%|█| 256/256 [00:00<00:00, 7141.28it/s]\n",
"Epoch 4: Train, Loglike in nats: 4.237622: 100%|█| 2304/2304 [00:00<00:00, 3\n",
"- Val, Loglike in nats: 4.493383: 100%|█| 256/256 [00:00<00:00, 7184.57it/s]\n",
"Epoch 5: Train, Loglike in nats: 4.404403: 100%|█| 2304/2304 [00:00<00:00, 3\n",
"- Val, Loglike in nats: 4.459258: 100%|█| 256/256 [00:00<00:00, 6830.51it/s]\n",
"Epoch 6: Train, Loglike in nats: 4.350708: 100%|█| 2304/2304 [00:00<00:00, 3\n",
"- Val, Loglike in nats: 4.407907: 100%|█| 256/256 [00:00<00:00, 7303.42it/s]\n",
"Epoch 7: Train, Loglike in nats: 4.355108: 100%|█| 2304/2304 [00:00<00:00, 3\n",
"- Val, Loglike in nats: 4.503678: 100%|█| 256/256 [00:00<00:00, 7277.48it/s]\n",
"Epoch 8: Train, Loglike in nats: 4.348162: 100%|█| 2304/2304 [00:00<00:00, 3\n",
"- Val, Loglike in nats: 4.218352: 100%|█| 256/256 [00:00<00:00, 7318.01it/s]\n",
"Epoch 9: Train, Loglike in nats: 4.191212: 100%|█| 2304/2304 [00:00<00:00, 3\n",
"- Val, Loglike in nats: 3.973791: 100%|█| 256/256 [00:00<00:00, 7148.41it/s]\n",
"Epoch 10: Train, Loglike in nats: 4.302415: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.426634: 100%|█| 256/256 [00:00<00:00, 7131.32it/s]\n",
"Epoch 11: Train, Loglike in nats: 4.252123: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.293779: 100%|█| 256/256 [00:00<00:00, 7329.65it/s]\n",
"Epoch 12: Train, Loglike in nats: 4.432020: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.513666: 100%|█| 256/256 [00:00<00:00, 7017.55it/s]\n",
"Epoch 13: Train, Loglike in nats: 4.391188: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.558033: 100%|█| 256/256 [00:00<00:00, 6985.37it/s]\n",
"Epoch 14: Train, Loglike in nats: 4.186291: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.399681: 100%|█| 256/256 [00:00<00:00, 7000.67it/s]\n",
"Epoch 15: Train, Loglike in nats: 4.360615: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.640914: 100%|█| 256/256 [00:00<00:00, 7112.24it/s]\n",
"Epoch 16: Train, Loglike in nats: 4.545840: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.590468: 100%|█| 256/256 [00:00<00:00, 7168.27it/s]\n",
"Epoch 17: Train, Loglike in nats: 4.498627: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.552762: 100%|█| 256/256 [00:00<00:00, 7284.10it/s]\n",
"Epoch 18: Train, Loglike in nats: 4.507410: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.645383: 100%|█| 256/256 [00:00<00:00, 7103.63it/s]\n",
"Epoch 19: Train, Loglike in nats: 4.481674: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.461302: 100%|█| 256/256 [00:00<00:00, 7131.75it/s]\n",
"Epoch 20: Train, Loglike in nats: 4.441385: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.054482: 100%|█| 256/256 [00:00<00:00, 7162.43it/s]\n",
"Epoch 21: Train, Loglike in nats: 4.240643: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.526018: 100%|█| 256/256 [00:00<00:00, 7248.01it/s]\n",
"Epoch 22: Train, Loglike in nats: 4.449009: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.404407: 100%|█| 256/256 [00:00<00:00, 7274.03it/s]\n",
"Epoch 23: Train, Loglike in nats: 4.430830: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.533774: 100%|█| 256/256 [00:00<00:00, 7162.19it/s]\n",
"Epoch 24: Train, Loglike in nats: 4.519465: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.677872: 100%|█| 256/256 [00:00<00:00, 7493.49it/s]\n",
"Epoch 25: Train, Loglike in nats: 4.518265: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.522072: 100%|█| 256/256 [00:00<00:00, 7223.92it/s]\n",
"Epoch 26: Train, Loglike in nats: 4.562395: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.534011: 100%|█| 256/256 [00:00<00:00, 7121.25it/s]\n",
"Epoch 27: Train, Loglike in nats: 4.584528: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.586026: 100%|█| 256/256 [00:00<00:00, 7351.68it/s]\n",
"Epoch 28: Train, Loglike in nats: 4.462471: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.472850: 100%|█| 256/256 [00:00<00:00, 7036.82it/s]\n",
"Epoch 29: Train, Loglike in nats: 4.472151: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.533405: 100%|█| 256/256 [00:00<00:00, 7249.43it/s]\n",
"Epoch 30: Train, Loglike in nats: 4.451222: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.603378: 100%|█| 256/256 [00:00<00:00, 7152.84it/s]\n",
"Epoch 31: Train, Loglike in nats: 4.409621: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.555896: 100%|█| 256/256 [00:00<00:00, 7419.44it/s]\n",
"Epoch 32: Train, Loglike in nats: 4.554532: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.501821: 100%|█| 256/256 [00:00<00:00, 7298.61it/s]\n",
"Epoch 33: Train, Loglike in nats: 4.552832: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.754763: 100%|█| 256/256 [00:00<00:00, 7293.25it/s]\n",
"Epoch 34: Train, Loglike in nats: 4.606837: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.579247: 100%|█| 256/256 [00:00<00:00, 7115.02it/s]\n",
"Epoch 35: Train, Loglike in nats: 4.419414: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.602200: 100%|█| 256/256 [00:00<00:00, 7263.50it/s]\n",
"Epoch 36: Train, Loglike in nats: 4.608782: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.745889: 100%|█| 256/256 [00:00<00:00, 6894.19it/s]\n",
"Epoch 37: Train, Loglike in nats: 4.432496: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.760521: 100%|█| 256/256 [00:00<00:00, 7105.41it/s]\n",
"Epoch 38: Train, Loglike in nats: 4.587770: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.662574: 100%|█| 256/256 [00:00<00:00, 7150.84it/s]\n",
"Epoch 39: Train, Loglike in nats: 4.458916: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.421184: 100%|█| 256/256 [00:00<00:00, 7190.44it/s]\n",
"Epoch 40: Train, Loglike in nats: 4.348884: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.565992: 100%|█| 256/256 [00:00<00:00, 7223.29it/s]\n",
"Epoch 41: Train, Loglike in nats: 4.546117: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.673806: 100%|█| 256/256 [00:00<00:00, 6989.78it/s]\n",
"Epoch 42: Train, Loglike in nats: 4.599654: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.709753: 100%|█| 256/256 [00:00<00:00, 7053.51it/s]\n",
"Epoch 43: Train, Loglike in nats: 4.625075: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.603225: 100%|█| 256/256 [00:00<00:00, 7310.28it/s]\n",
"Epoch 44: Train, Loglike in nats: 4.659478: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.630083: 100%|█| 256/256 [00:00<00:00, 7016.96it/s]\n",
"Epoch 45: Train, Loglike in nats: 4.682000: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.732234: 100%|█| 256/256 [00:00<00:00, 6962.40it/s]\n",
"Epoch 46: Train, Loglike in nats: 4.554792: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.663981: 100%|█| 256/256 [00:00<00:00, 7155.23it/s]\n",
"Epoch 47: Train, Loglike in nats: 4.573297: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.564854: 100%|█| 256/256 [00:00<00:00, 7175.12it/s]\n",
"Epoch 48: Train, Loglike in nats: 4.541779: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.738482: 100%|█| 256/256 [00:00<00:00, 7169.51it/s]\n",
"Epoch 49: Train, Loglike in nats: 4.509175: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.634123: 100%|█| 256/256 [00:00<00:00, 7196.81it/s]\n",
"Epoch 50: Train, Loglike in nats: 4.552632: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.652342: 100%|█| 256/256 [00:00<00:00, 7038.71it/s]\n",
"Epoch 51: Train, Loglike in nats: 4.531078: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.396230: 100%|█| 256/256 [00:00<00:00, 7016.13it/s]\n",
"Epoch 52: Train, Loglike in nats: 4.492283: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.382210: 100%|█| 256/256 [00:00<00:00, 7154.80it/s]\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 53: Train, Loglike in nats: 4.460400: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.672035: 100%|█| 256/256 [00:00<00:00, 7438.20it/s]\n",
"Epoch 54: Train, Loglike in nats: 4.339007: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.002027: 100%|█| 256/256 [00:00<00:00, 6945.16it/s]\n",
"Epoch 55: Train, Loglike in nats: 4.277437: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.693196: 100%|█| 256/256 [00:00<00:00, 7248.06it/s]\n",
"Epoch 56: Train, Loglike in nats: 4.443362: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.625434: 100%|█| 256/256 [00:00<00:00, 7293.85it/s]\n",
"Epoch 57: Train, Loglike in nats: 4.621117: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.820796: 100%|█| 256/256 [00:00<00:00, 7194.25it/s]\n",
"Epoch 58: Train, Loglike in nats: 4.700033: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.580562: 100%|█| 256/256 [00:00<00:00, 7323.30it/s]\n",
"Epoch 59: Train, Loglike in nats: 4.378896: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.219565: 100%|█| 256/256 [00:00<00:00, 7196.08it/s]\n",
"Epoch 60: Train, Loglike in nats: 4.540074: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.595256: 100%|█| 256/256 [00:00<00:00, 7024.49it/s]\n",
"Epoch 61: Train, Loglike in nats: 4.683227: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.320597: 100%|█| 256/256 [00:00<00:00, 7355.05it/s]\n",
"Epoch 62: Train, Loglike in nats: 4.550050: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.652726: 100%|█| 256/256 [00:00<00:00, 7323.95it/s]\n",
"Epoch 63: Train, Loglike in nats: 4.577566: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.612271: 100%|█| 256/256 [00:00<00:00, 6993.56it/s]\n",
"Epoch 64: Train, Loglike in nats: 4.538843: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.601538: 100%|█| 256/256 [00:00<00:00, 7233.85it/s]\n",
"Epoch 65: Train, Loglike in nats: 4.654364: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.568342: 100%|█| 256/256 [00:00<00:00, 7060.33it/s]\n",
"Epoch 66: Train, Loglike in nats: 4.562939: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.683165: 100%|█| 256/256 [00:00<00:00, 7133.88it/s]\n",
"Epoch 67: Train, Loglike in nats: 4.586874: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.757002: 100%|█| 256/256 [00:00<00:00, 7076.89it/s]\n",
"Epoch 68: Train, Loglike in nats: 4.575104: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.729914: 100%|█| 256/256 [00:00<00:00, 7020.77it/s]\n",
"Epoch 69: Train, Loglike in nats: 4.206414: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.259872: 100%|█| 256/256 [00:00<00:00, 7159.42it/s]\n",
"Epoch 70: Train, Loglike in nats: 4.548915: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.706856: 100%|█| 256/256 [00:00<00:00, 7115.26it/s]\n",
"Epoch 71: Train, Loglike in nats: 4.698675: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.801560: 100%|█| 256/256 [00:00<00:00, 7087.08it/s]\n",
"Epoch 72: Train, Loglike in nats: 4.531759: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.770474: 100%|█| 256/256 [00:00<00:00, 7160.43it/s]\n",
"Epoch 73: Train, Loglike in nats: 4.546275: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.720428: 100%|█| 256/256 [00:00<00:00, 7092.60it/s]\n",
"Epoch 74: Train, Loglike in nats: 4.661378: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.873534: 100%|█| 256/256 [00:00<00:00, 7276.25it/s]\n",
"Epoch 75: Train, Loglike in nats: 4.740895: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.830359: 100%|█| 256/256 [00:00<00:00, 7191.02it/s]\n",
"Epoch 76: Train, Loglike in nats: 4.737348: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.647756: 100%|█| 256/256 [00:00<00:00, 7183.13it/s]\n",
"Epoch 77: Train, Loglike in nats: 4.607565: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.744749: 100%|█| 256/256 [00:00<00:00, 7257.27it/s]\n",
"Epoch 78: Train, Loglike in nats: 4.523176: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.638309: 100%|█| 256/256 [00:00<00:00, 7021.78it/s]\n",
"Epoch 79: Train, Loglike in nats: 4.543652: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.789568: 100%|█| 256/256 [00:00<00:00, 7170.23it/s]\n",
"Epoch 80: Train, Loglike in nats: 4.694047: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.837515: 100%|█| 256/256 [00:00<00:00, 7103.63it/s]\n",
"Epoch 81: Train, Loglike in nats: 4.742090: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.797641: 100%|█| 256/256 [00:00<00:00, 7084.79it/s]\n",
"Epoch 82: Train, Loglike in nats: 4.536091: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.481468: 100%|█| 256/256 [00:00<00:00, 7161.43it/s]\n",
"Epoch 83: Train, Loglike in nats: 4.666183: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.797166: 100%|█| 256/256 [00:00<00:00, 7439.96it/s]\n",
"Epoch 84: Train, Loglike in nats: 4.700233: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.829692: 100%|█| 256/256 [00:00<00:00, 7095.93it/s]\n",
"Epoch 85: Train, Loglike in nats: 4.740786: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.909887: 100%|█| 256/256 [00:00<00:00, 6972.26it/s]\n",
"Epoch 86: Train, Loglike in nats: 4.708581: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.658979: 100%|█| 256/256 [00:00<00:00, 6973.21it/s]\n",
"Epoch 87: Train, Loglike in nats: 4.672133: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.794995: 100%|█| 256/256 [00:00<00:00, 7088.06it/s]\n",
"Epoch 88: Train, Loglike in nats: 4.684402: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.797291: 100%|█| 256/256 [00:00<00:00, 7179.72it/s]\n",
"Epoch 89: Train, Loglike in nats: 4.559476: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.670976: 100%|█| 256/256 [00:00<00:00, 6810.32it/s]\n",
"Epoch 90: Train, Loglike in nats: 4.641824: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.758659: 100%|█| 256/256 [00:00<00:00, 6940.26it/s]\n",
"Epoch 91: Train, Loglike in nats: 4.744356: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.585352: 100%|█| 256/256 [00:00<00:00, 7100.25it/s]\n",
"Epoch 92: Train, Loglike in nats: 4.690108: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.651466: 100%|█| 256/256 [00:00<00:00, 7128.39it/s]\n",
"Epoch 93: Train, Loglike in nats: 4.684790: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.754060: 100%|█| 256/256 [00:00<00:00, 7134.50it/s]\n",
"Epoch 94: Train, Loglike in nats: 4.719005: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.782381: 100%|█| 256/256 [00:00<00:00, 7038.81it/s]\n",
"Epoch 95: Train, Loglike in nats: 4.661931: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.718342: 100%|█| 256/256 [00:00<00:00, 6701.04it/s]\n",
"Epoch 96: Train, Loglike in nats: 4.701336: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.633593: 100%|█| 256/256 [00:00<00:00, 7056.80it/s]\n",
"Epoch 97: Train, Loglike in nats: 4.697102: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.844307: 100%|█| 256/256 [00:00<00:00, 6608.98it/s]\n",
"Epoch 98: Train, Loglike in nats: 4.685904: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.681110: 100%|█| 256/256 [00:00<00:00, 7009.90it/s]\n",
"Epoch 99: Train, Loglike in nats: 4.685625: 100%|█| 2304/2304 [00:00<00:00, \n",
"- Val, Loglike in nats: 4.812140: 100%|█| 256/256 [00:00<00:00, 6770.81it/s]\n",
"Epoch 100: Train, Loglike in nats: 4.724505: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.898593: 100%|█| 256/256 [00:00<00:00, 7114.88it/s]\n",
"Epoch 101: Train, Loglike in nats: 4.776207: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.743017: 100%|█| 256/256 [00:00<00:00, 6914.43it/s]\n",
"Epoch 102: Train, Loglike in nats: 4.724447: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.904695: 100%|█| 256/256 [00:00<00:00, 6824.90it/s]\n",
"Epoch 103: Train, Loglike in nats: 4.760356: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.831684: 100%|█| 256/256 [00:00<00:00, 7001.99it/s]\n",
"Epoch 104: Train, Loglike in nats: 4.720137: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.777143: 100%|█| 256/256 [00:00<00:00, 7022.14it/s]\n",
"Epoch 105: Train, Loglike in nats: 4.749445: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.824291: 100%|█| 256/256 [00:00<00:00, 7024.30it/s]\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 106: Train, Loglike in nats: 4.739383: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.662352: 100%|█| 256/256 [00:00<00:00, 7214.89it/s]\n",
"Epoch 107: Train, Loglike in nats: 4.663712: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.730576: 100%|█| 256/256 [00:00<00:00, 7186.11it/s]\n",
"Epoch 108: Train, Loglike in nats: 4.685017: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.662242: 100%|█| 256/256 [00:00<00:00, 7161.81it/s]\n",
"Epoch 109: Train, Loglike in nats: 4.420206: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.799919: 100%|█| 256/256 [00:00<00:00, 7222.85it/s]\n",
"Epoch 110: Train, Loglike in nats: 4.586950: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.839162: 100%|█| 256/256 [00:00<00:00, 6641.77it/s]\n",
"Epoch 111: Train, Loglike in nats: 4.687782: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.866652: 100%|█| 256/256 [00:00<00:00, 7072.79it/s]\n",
"Epoch 112: Train, Loglike in nats: 4.816356: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.844614: 100%|█| 256/256 [00:00<00:00, 7073.03it/s]\n",
"Epoch 113: Train, Loglike in nats: 4.815367: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.844851: 100%|█| 256/256 [00:00<00:00, 7283.70it/s]\n",
"Epoch 114: Train, Loglike in nats: 4.801442: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.782068: 100%|█| 256/256 [00:00<00:00, 7006.38it/s]\n",
"Epoch 115: Train, Loglike in nats: 4.712310: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.613576: 100%|█| 256/256 [00:00<00:00, 7382.81it/s]\n",
"Epoch 116: Train, Loglike in nats: 4.729708: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.861105: 100%|█| 256/256 [00:00<00:00, 7097.29it/s]\n",
"Epoch 117: Train, Loglike in nats: 4.768230: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.556072: 100%|█| 256/256 [00:00<00:00, 7212.13it/s]\n",
"Epoch 118: Train, Loglike in nats: 4.769103: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.824241: 100%|█| 256/256 [00:00<00:00, 7655.69it/s]\n",
"Epoch 119: Train, Loglike in nats: 4.774367: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.812512: 100%|█| 256/256 [00:00<00:00, 7284.05it/s]\n",
"Epoch 120: Train, Loglike in nats: 4.763802: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.676857: 100%|█| 256/256 [00:00<00:00, 7510.00it/s]\n",
"Epoch 121: Train, Loglike in nats: 4.766626: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.737663: 100%|█| 256/256 [00:00<00:00, 6953.12it/s]\n",
"Epoch 122: Train, Loglike in nats: 4.828807: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.924353: 100%|█| 256/256 [00:00<00:00, 7384.74it/s]\n",
"Epoch 123: Train, Loglike in nats: 4.796009: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.932484: 100%|█| 256/256 [00:00<00:00, 7399.45it/s]\n",
"Epoch 124: Train, Loglike in nats: 4.831570: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.895680: 100%|█| 256/256 [00:00<00:00, 7358.02it/s]\n",
"Epoch 125: Train, Loglike in nats: 4.801763: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.716289: 100%|█| 256/256 [00:00<00:00, 7310.73it/s]\n",
"Epoch 126: Train, Loglike in nats: 4.754300: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.878354: 100%|█| 256/256 [00:00<00:00, 7261.59it/s]\n",
"Epoch 127: Train, Loglike in nats: 4.798734: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.861428: 100%|█| 256/256 [00:00<00:00, 7362.67it/s]\n",
"Epoch 128: Train, Loglike in nats: 4.764201: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.812524: 100%|█| 256/256 [00:00<00:00, 7426.78it/s]\n",
"Epoch 129: Train, Loglike in nats: 4.738397: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.804435: 100%|█| 256/256 [00:00<00:00, 6923.03it/s]\n",
"Epoch 130: Train, Loglike in nats: 4.825358: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.922467: 100%|█| 256/256 [00:00<00:00, 7172.91it/s]\n",
"Epoch 131: Train, Loglike in nats: 4.783786: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.763267: 100%|█| 256/256 [00:00<00:00, 7425.91it/s]\n",
"Epoch 132: Train, Loglike in nats: 4.808064: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.885164: 100%|█| 256/256 [00:00<00:00, 6899.37it/s]\n",
"Epoch 133: Train, Loglike in nats: 4.829371: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.793253: 100%|█| 256/256 [00:00<00:00, 6901.23it/s]\n",
"Epoch 134: Train, Loglike in nats: 4.769795: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.841720: 100%|█| 256/256 [00:00<00:00, 6932.60it/s]\n",
"Epoch 135: Train, Loglike in nats: 4.839640: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.795256: 100%|█| 256/256 [00:00<00:00, 7117.99it/s]\n",
"Epoch 136: Train, Loglike in nats: 4.870009: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.964134: 100%|█| 256/256 [00:00<00:00, 6903.45it/s]\n",
"Epoch 137: Train, Loglike in nats: 4.784118: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.947754: 100%|█| 256/256 [00:00<00:00, 7205.40it/s]\n",
"Epoch 138: Train, Loglike in nats: 4.821724: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.933723: 100%|█| 256/256 [00:00<00:00, 6872.65it/s]\n",
"Epoch 139: Train, Loglike in nats: 4.878417: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.897148: 100%|█| 256/256 [00:00<00:00, 6874.90it/s]\n",
"Epoch 140: Train, Loglike in nats: 4.879096: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.883657: 100%|█| 256/256 [00:00<00:00, 6566.95it/s]\n",
"Epoch 141: Train, Loglike in nats: 4.792603: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.710927: 100%|█| 256/256 [00:00<00:00, 7164.82it/s]\n",
"Epoch 142: Train, Loglike in nats: 4.771148: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.925281: 100%|█| 256/256 [00:00<00:00, 7520.57it/s]\n",
"Epoch 143: Train, Loglike in nats: 4.852556: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.768024: 100%|█| 256/256 [00:00<00:00, 7472.47it/s]\n",
"Epoch 144: Train, Loglike in nats: 4.802715: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.756022: 100%|█| 256/256 [00:00<00:00, 7406.34it/s]\n",
"Epoch 145: Train, Loglike in nats: 4.688688: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.933115: 100%|█| 256/256 [00:00<00:00, 7404.81it/s]\n",
"Epoch 146: Train, Loglike in nats: 4.875092: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.831896: 100%|█| 256/256 [00:00<00:00, 7476.43it/s]\n",
"Epoch 147: Train, Loglike in nats: 4.818489: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.981203: 100%|█| 256/256 [00:00<00:00, 7582.07it/s]\n",
"Epoch 148: Train, Loglike in nats: 4.865102: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.907088: 100%|█| 256/256 [00:00<00:00, 7416.88it/s]\n",
"Epoch 149: Train, Loglike in nats: 4.859244: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.958753: 100%|█| 256/256 [00:00<00:00, 7447.44it/s]\n",
"Epoch 150: Train, Loglike in nats: 4.854468: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.884654: 100%|█| 256/256 [00:00<00:00, 5605.46it/s]\n",
"Epoch 151: Train, Loglike in nats: 4.817355: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.817810: 100%|█| 256/256 [00:00<00:00, 7374.60it/s]\n",
"Epoch 152: Train, Loglike in nats: 4.823654: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.976997: 100%|█| 256/256 [00:00<00:00, 6717.06it/s]\n",
"Epoch 153: Train, Loglike in nats: 4.867379: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.957842: 100%|█| 256/256 [00:00<00:00, 6813.08it/s]\n",
"Epoch 154: Train, Loglike in nats: 4.846090: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.912516: 100%|█| 256/256 [00:00<00:00, 7653.68it/s]\n",
"Epoch 155: Train, Loglike in nats: 4.832615: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.962831: 100%|█| 256/256 [00:00<00:00, 7428.89it/s]\n",
"Epoch 156: Train, Loglike in nats: 4.877176: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.918849: 100%|█| 256/256 [00:00<00:00, 7795.88it/s]\n",
"Epoch 157: Train, Loglike in nats: 4.891723: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.851194: 100%|█| 256/256 [00:00<00:00, 7380.53it/s]\n",
"Epoch 158: Train, Loglike in nats: 4.667483: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.476809: 100%|█| 256/256 [00:00<00:00, 6922.68it/s]\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 159: Train, Loglike in nats: 4.769931: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.942445: 100%|█| 256/256 [00:00<00:00, 6408.49it/s]\n",
"Epoch 160: Train, Loglike in nats: 4.880798: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.913004: 100%|█| 256/256 [00:00<00:00, 7519.89it/s]\n",
"Epoch 161: Train, Loglike in nats: 4.930751: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.947358: 100%|█| 256/256 [00:00<00:00, 7085.35it/s]\n",
"Epoch 162: Train, Loglike in nats: 4.842791: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.882313: 100%|█| 256/256 [00:00<00:00, 7413.40it/s]\n",
"Epoch 163: Train, Loglike in nats: 4.823536: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.852108: 100%|█| 256/256 [00:00<00:00, 7592.84it/s]\n",
"Epoch 164: Train, Loglike in nats: 4.855254: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.936172: 100%|█| 256/256 [00:00<00:00, 7433.62it/s]\n",
"Epoch 165: Train, Loglike in nats: 4.891049: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.928318: 100%|█| 256/256 [00:00<00:00, 7505.95it/s]\n",
"Epoch 166: Train, Loglike in nats: 4.855692: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.800160: 100%|█| 256/256 [00:00<00:00, 7678.96it/s]\n",
"Epoch 167: Train, Loglike in nats: 4.809008: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.626600: 100%|█| 256/256 [00:00<00:00, 7681.88it/s]\n",
"Epoch 168: Train, Loglike in nats: 4.860285: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.986201: 100%|█| 256/256 [00:00<00:00, 7557.20it/s]\n",
"Epoch 169: Train, Loglike in nats: 4.852093: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.891025: 100%|█| 256/256 [00:00<00:00, 7441.50it/s]\n",
"Epoch 170: Train, Loglike in nats: 4.848801: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.009564: 100%|█| 256/256 [00:00<00:00, 7502.07it/s]\n",
"Epoch 171: Train, Loglike in nats: 4.877434: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.937558: 100%|█| 256/256 [00:00<00:00, 7402.92it/s]\n",
"Epoch 172: Train, Loglike in nats: 4.828379: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.926861: 100%|█| 256/256 [00:00<00:00, 7486.07it/s]\n",
"Epoch 173: Train, Loglike in nats: 4.918975: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.916042: 100%|█| 256/256 [00:00<00:00, 7650.46it/s]\n",
"Epoch 174: Train, Loglike in nats: 4.867292: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.995896: 100%|█| 256/256 [00:00<00:00, 7810.79it/s]\n",
"Epoch 175: Train, Loglike in nats: 4.893898: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.887014: 100%|█| 256/256 [00:00<00:00, 7407.82it/s]\n",
"Epoch 176: Train, Loglike in nats: 4.930756: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.928427: 100%|█| 256/256 [00:00<00:00, 7560.98it/s]\n",
"Epoch 177: Train, Loglike in nats: 4.860454: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.842382: 100%|█| 256/256 [00:00<00:00, 7499.40it/s]\n",
"Epoch 178: Train, Loglike in nats: 4.826075: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.800313: 100%|█| 256/256 [00:00<00:00, 7620.22it/s]\n",
"Epoch 179: Train, Loglike in nats: 4.847323: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.832878: 100%|█| 256/256 [00:00<00:00, 7808.41it/s]\n",
"Epoch 180: Train, Loglike in nats: 4.785418: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.888988: 100%|█| 256/256 [00:00<00:00, 7248.65it/s]\n",
"Epoch 181: Train, Loglike in nats: 4.887916: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.003945: 100%|█| 256/256 [00:00<00:00, 7689.74it/s]\n",
"Epoch 182: Train, Loglike in nats: 4.878324: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.002037: 100%|█| 256/256 [00:00<00:00, 7656.84it/s]\n",
"Epoch 183: Train, Loglike in nats: 4.905340: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.948775: 100%|█| 256/256 [00:00<00:00, 7621.51it/s]\n",
"Epoch 184: Train, Loglike in nats: 4.917200: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.958248: 100%|█| 256/256 [00:00<00:00, 7435.99it/s]\n",
"Epoch 185: Train, Loglike in nats: 4.902764: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.019319: 100%|█| 256/256 [00:00<00:00, 7558.79it/s]\n",
"Epoch 186: Train, Loglike in nats: 4.950947: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.973337: 100%|█| 256/256 [00:00<00:00, 7723.82it/s]\n",
"Epoch 187: Train, Loglike in nats: 4.858890: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.047011: 100%|█| 256/256 [00:00<00:00, 7556.35it/s]\n",
"Epoch 188: Train, Loglike in nats: 4.907679: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.997971: 100%|█| 256/256 [00:00<00:00, 7443.20it/s]\n",
"Epoch 189: Train, Loglike in nats: 4.877380: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.917530: 100%|█| 256/256 [00:00<00:00, 7613.79it/s]\n",
"Epoch 190: Train, Loglike in nats: 4.917055: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.976613: 100%|█| 256/256 [00:00<00:00, 7703.15it/s]\n",
"Epoch 191: Train, Loglike in nats: 4.903774: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.851151: 100%|█| 256/256 [00:00<00:00, 7672.49it/s]\n",
"Epoch 192: Train, Loglike in nats: 4.885668: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.779591: 100%|█| 256/256 [00:00<00:00, 7464.83it/s]\n",
"Epoch 193: Train, Loglike in nats: 4.913061: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.933357: 100%|█| 256/256 [00:00<00:00, 7167.88it/s]\n",
"Epoch 194: Train, Loglike in nats: 4.854201: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.808918: 100%|█| 256/256 [00:00<00:00, 6969.23it/s]\n",
"Epoch 195: Train, Loglike in nats: 4.858738: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.976485: 100%|█| 256/256 [00:00<00:00, 6687.23it/s]\n",
"Epoch 196: Train, Loglike in nats: 4.883326: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.989956: 100%|█| 256/256 [00:00<00:00, 7390.89it/s]\n",
"Epoch 197: Train, Loglike in nats: 4.916571: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.880527: 100%|█| 256/256 [00:00<00:00, 6527.46it/s]\n",
"Epoch 198: Train, Loglike in nats: 4.944581: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.918352: 100%|█| 256/256 [00:00<00:00, 5142.64it/s]\n",
"Epoch 199: Train, Loglike in nats: 4.919453: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.931729: 100%|█| 256/256 [00:00<00:00, 7274.77it/s]\n",
"Epoch 200: Train, Loglike in nats: 4.863971: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.930439: 100%|█| 256/256 [00:00<00:00, 7322.25it/s]\n",
"Epoch 201: Train, Loglike in nats: 4.931427: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.953879: 100%|█| 256/256 [00:00<00:00, 6751.95it/s]\n",
"Epoch 202: Train, Loglike in nats: 4.940583: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.093653: 100%|█| 256/256 [00:00<00:00, 7374.14it/s]\n",
"Epoch 203: Train, Loglike in nats: 4.933789: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.026846: 100%|█| 256/256 [00:00<00:00, 7542.97it/s]\n",
"Epoch 204: Train, Loglike in nats: 4.901556: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.076484: 100%|█| 256/256 [00:00<00:00, 7426.52it/s]\n",
"Epoch 205: Train, Loglike in nats: 4.891478: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.938434: 100%|█| 256/256 [00:00<00:00, 7284.35it/s]\n",
"Epoch 206: Train, Loglike in nats: 4.905023: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.002440: 100%|█| 256/256 [00:00<00:00, 7523.89it/s]\n",
"Epoch 207: Train, Loglike in nats: 4.922543: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.946003: 100%|█| 256/256 [00:00<00:00, 7008.90it/s]\n",
"Epoch 208: Train, Loglike in nats: 4.933520: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.929842: 100%|█| 256/256 [00:00<00:00, 7421.34it/s]\n",
"Epoch 209: Train, Loglike in nats: 4.905619: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.001271: 100%|█| 256/256 [00:00<00:00, 7330.65it/s]\n",
"Epoch 210: Train, Loglike in nats: 4.923283: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.988173: 100%|█| 256/256 [00:00<00:00, 7337.76it/s]\n",
"Epoch 211: Train, Loglike in nats: 4.934201: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.032679: 100%|█| 256/256 [00:00<00:00, 7585.39it/s]\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Epoch 212: Train, Loglike in nats: 4.943468: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.998631: 100%|█| 256/256 [00:00<00:00, 7383.53it/s]\n",
"Epoch 213: Train, Loglike in nats: 4.959769: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.023635: 100%|█| 256/256 [00:00<00:00, 7492.13it/s]\n",
"Epoch 214: Train, Loglike in nats: 4.937681: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.956433: 100%|█| 256/256 [00:00<00:00, 7380.73it/s]\n",
"Epoch 215: Train, Loglike in nats: 4.959364: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.026244: 100%|█| 256/256 [00:00<00:00, 6990.14it/s]\n",
"Epoch 216: Train, Loglike in nats: 4.938967: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.952768: 100%|█| 256/256 [00:00<00:00, 7575.97it/s]\n",
"Epoch 217: Train, Loglike in nats: 4.959960: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.049692: 100%|█| 256/256 [00:00<00:00, 7559.86it/s]\n",
"Epoch 218: Train, Loglike in nats: 4.943532: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.989525: 100%|█| 256/256 [00:00<00:00, 7709.45it/s]\n",
"Epoch 219: Train, Loglike in nats: 4.998487: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.930694: 100%|█| 256/256 [00:00<00:00, 7233.46it/s]\n",
"Epoch 220: Train, Loglike in nats: 4.922991: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.968363: 100%|█| 256/256 [00:00<00:00, 7298.41it/s]\n",
"Epoch 221: Train, Loglike in nats: 4.944018: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.964511: 100%|█| 256/256 [00:00<00:00, 7098.88it/s]\n",
"Epoch 222: Train, Loglike in nats: 4.959001: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.976989: 100%|█| 256/256 [00:00<00:00, 7077.13it/s]\n",
"Epoch 223: Train, Loglike in nats: 4.942852: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.032462: 100%|█| 256/256 [00:00<00:00, 7606.34it/s]\n",
"Epoch 224: Train, Loglike in nats: 4.977668: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.983855: 100%|█| 256/256 [00:00<00:00, 7353.84it/s]\n",
"Epoch 225: Train, Loglike in nats: 4.939078: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.935115: 100%|█| 256/256 [00:00<00:00, 7761.56it/s]\n",
"Epoch 226: Train, Loglike in nats: 4.953270: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.013212: 100%|█| 256/256 [00:00<00:00, 7523.68it/s]\n",
"Epoch 227: Train, Loglike in nats: 4.954148: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.978697: 100%|█| 256/256 [00:00<00:00, 7312.62it/s]\n",
"Epoch 228: Train, Loglike in nats: 4.941857: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.009906: 100%|█| 256/256 [00:00<00:00, 7447.95it/s]\n",
"Epoch 229: Train, Loglike in nats: 4.969720: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.016267: 100%|█| 256/256 [00:00<00:00, 7122.43it/s]\n",
"Epoch 230: Train, Loglike in nats: 4.936580: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.001699: 100%|█| 256/256 [00:00<00:00, 7477.26it/s]\n",
"Epoch 231: Train, Loglike in nats: 4.953486: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.947896: 100%|█| 256/256 [00:00<00:00, 7471.64it/s]\n",
"Epoch 232: Train, Loglike in nats: 4.949690: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.970058: 100%|█| 256/256 [00:00<00:00, 7692.99it/s]\n",
"Epoch 233: Train, Loglike in nats: 4.943972: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.994866: 100%|█| 256/256 [00:00<00:00, 7354.45it/s]\n",
"Epoch 234: Train, Loglike in nats: 4.946190: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.979137: 100%|█| 256/256 [00:00<00:00, 7566.36it/s]\n",
"Epoch 235: Train, Loglike in nats: 4.940177: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.925819: 100%|█| 256/256 [00:00<00:00, 7471.38it/s]\n",
"Epoch 236: Train, Loglike in nats: 4.947596: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.070110: 100%|█| 256/256 [00:00<00:00, 6911.85it/s]\n",
"Epoch 237: Train, Loglike in nats: 4.999313: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.032080: 100%|█| 256/256 [00:00<00:00, 7347.65it/s]\n",
"Epoch 238: Train, Loglike in nats: 4.981619: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.981657: 100%|█| 256/256 [00:00<00:00, 7010.22it/s]\n",
"Epoch 239: Train, Loglike in nats: 4.946543: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.985532: 100%|█| 256/256 [00:00<00:00, 7266.55it/s]\n",
"Epoch 240: Train, Loglike in nats: 4.963146: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.995767: 100%|█| 256/256 [00:00<00:00, 7344.94it/s]\n",
"Epoch 241: Train, Loglike in nats: 4.983130: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.043330: 100%|█| 256/256 [00:00<00:00, 7408.03it/s]\n",
"Epoch 242: Train, Loglike in nats: 4.977412: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.067501: 100%|█| 256/256 [00:00<00:00, 7592.63it/s]\n",
"Epoch 243: Train, Loglike in nats: 4.972563: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.040345: 100%|█| 256/256 [00:00<00:00, 7690.40it/s]\n",
"Epoch 244: Train, Loglike in nats: 4.975473: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.030294: 100%|█| 256/256 [00:00<00:00, 7333.95it/s]\n",
"Epoch 245: Train, Loglike in nats: 4.971058: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.016021: 100%|█| 256/256 [00:00<00:00, 6894.14it/s]\n",
"Epoch 246: Train, Loglike in nats: 4.964948: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.992126: 100%|█| 256/256 [00:00<00:00, 7434.08it/s]\n",
"Epoch 247: Train, Loglike in nats: 4.999891: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.046591: 100%|█| 256/256 [00:00<00:00, 7580.89it/s]\n",
"Epoch 248: Train, Loglike in nats: 5.001117: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.025182: 100%|█| 256/256 [00:00<00:00, 7697.79it/s]\n",
"Epoch 249: Train, Loglike in nats: 4.974158: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.001335: 100%|█| 256/256 [00:00<00:00, 7637.94it/s]\n",
"Epoch 250: Train, Loglike in nats: 4.944972: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 5.019472: 100%|█| 256/256 [00:00<00:00, 7649.10it/s]\n",
"Epoch 251: Train, Loglike in nats: 4.971719: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.952545: 100%|█| 256/256 [00:00<00:00, 7522.78it/s]\n",
"Epoch 252: Train, Loglike in nats: 4.947555: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.969583: 100%|█| 256/256 [00:00<00:00, 6915.19it/s]\n",
"Epoch 253: Train, Loglike in nats: 4.972092: 100%|█| 2304/2304 [00:00<00:00,\n",
"- Val, Loglike in nats: 4.998250: 100%|█| 256/256 [00:00<00:00, 7167.64it/s]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"early stopping, loading state dict from epoch 202\n",
"Generating simulations\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"256it [00:00, 3215.27it/s]\n",
"WARNING:root:Too few points to create valid contours\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Effective sample size for current/all rounds 321.0/2176.7\n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Early stop: Surrogate posterior did not improve for this round\n"
]
}
],
"source": [
"engine.fit(\n",
" x_obs=x_obs,\n",
" y_true=y_true,\n",
" n_sims=2560, # N = 5120 per round\n",
" n_rounds=5, # Max 5 rounds. (also see early_stop_traing)\n",
" n_epochs=300,\n",
" batch_size=128,\n",
" lr=0.0005,\n",
" early_stop_train=True, # If sampling efficiency is reduced for this round, stop and revert to previous round\n",
" early_stop_patience=50, # Within a round, wait this many epochs before stopping training\n",
" noise=np.array([0.1]*50) # assume homogeneous noise; used for importance sampling\n",
")"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"plt.rcParams['figure.dpi']=100\n",
"plt.figure(figsize=(5,5))\n",
"plt.subplot(211)\n",
"plt.plot(np.array(engine.neff)/2560*100)\n",
"plt.plot([0,4],[20,20],label='ANPE: 20%')\n",
"plt.legend()\n",
"plt.ylabel('Sampling Efficiency [%]')\n",
"plt.subplot(212)\n",
"plt.plot([np.sum(engine.neff[:i+1]) for i in range(4)])\n",
"plt.ylabel('Effective Sample Size')\n",
"plt.xlabel('# Round')\n",
"plt.tight_layout()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"As we see, after the 3rd round of training the sampling efficiency is as high as 40%, meaning the NDE posterior is nearly the true posterior. In comparison, for ANPE we used more training samples and ended up with a lower sampling efficiency of 20%, showcasing the efficacy of SNPE (active learning).\n",
"\n",
"As expected, the 4th round of training degraded the result due to the use of NLL loss (see paper for details). Therefore, the network state is automatically reverted to the 3rd round. Note that every training round contributed to the final effective sample.\n",
"\n",
"To generate more effective posterior samples, let us draw 10000 more samples from the 3nd round NDE. The first corner plot is the (unweighted) NDE surrogate posterior. The second one shows the importance reweighted posterior, which is asymptotically exact"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"'test/2/266.pth'"
]
},
"execution_count": 14,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"engine.best_params"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {
"scrolled": false
},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"WARNING:root:Too few points to create valid contours\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"surrogate posterior\n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Generating simulations\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"1000it [00:00, 2913.62it/s]\n",
"WARNING:root:Too few points to create valid contours\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Effective Sample Size = 3757.0\n",
"Sampling efficiency = 37.6%\n"
]
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAtUAAALpCAYAAAB8C9q9AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAADl6UlEQVR4nOzdd3hTZfsH8G/SpGmbTjpogZaWIZSyl8geIlCggBYUUaaCP0BEhoCKiIIsfRVwAIoC4utARFnKnkWFUhCEtswOaKF7JelKzu+PkvM2tIWWJE2Tfj/XlctzTs64U88hd57znOeWCIIggIiIiIiIHpnU0gEQEREREVk7JtVEREREREZiUk1EREREZCQm1URERERERmJSTURERERkJCbVRERERERGYlJNRERERGQkJtVEREREREZiUk1EREREZCQm1URERERERmJSTWTFfvnlF3Tv3h1eXl5QKBRo1KgRZs2ahczMzIdue+3aNbzyyito37495HI5AgMDK1wvNDQUzs7O8PLywtSpU6FSqUz8SYiIiKybzNIBENGjy8jIQO/evTF37ly4ubnh4sWLWLx4Mf755x8cOnTogdteunQJu3fvRufOnSEIQrmJeHZ2Nvr27Yt69eph27ZtyMjIwKxZs3D37l1s377dXB+LiIjI6kgEQRAsHQQRlRUXF4egoCDcvHmzwlbk8nz55ZeYPHky4uPjERAQUOF6Op0OUmnJzapXXnkFf/zxB+Li4gzWWblyJRYtWoT4+Hj4+PgAKGkdf+aZZxAZGYkOHTpU+XMRERHZInb/ILIxderUAQAUFRU9cD19Qv0ge/fuRd++fcWEGgDCwsLg7OyM3bt3GxcoERGRDWFSbcUkEgkkEkmVtwsMDIREIinTKqm3bds29O7dGx4eHlAqlWjTpg1Wrlz50CStJjB37G+88Yb4d1+yZIlFYylNq9UiPz8fkZGRWLx4MUJDQ9G4cWOj93v58mUEBwcbLJPJZHjssccQHR1t9P6JiIhsBZNqMjBz5kyMGjUKERER6Ny5MwYOHIiEhATMmzcPffv2hUajsXSIFTJ37KdOncJHH31UqR8yjxKLIAgoLi4WX1qtFkBJwlx6eXk9tjw9PeHo6IhOnTrBz88PP/30k1GfVS8zMxPu7u5llnt4eCAjI8MkxyAiIrIFTKproUOHDiE6Ohr169c3WP7rr79i9erVcHZ2xt9//419+/Zh+/btuHr1Klq1aoWTJ09i4cKFFor6wcwdu1qtxvjx4+Hn54dhw4aZJZbNmzdDLpeLryZNmgAAmjRpYrB88+bNZbY9evQoIiIisG7dOly+fBlDhw4Vk3IiIiIyPybVtVDjxo3RvHlzyOVyg+UffPABAGD+/Plo3769uNzLywuff/45AODTTz9FdnZ29QVbSeaOfcGCBbh69So2bNgANzc3s8QydOhQnDlzRnzt3LkTALBz506D5UOHDi1zzLZt26Jr166YMmUKfv75Zxw5cgQ7dux4pM9amoeHB7Kyssosz8zMFPtuExEREZPqGqF03+gvv/wSHTp0gFKphLu7O0JDQ/HXX389dB/bt29H9+7d4erqCqVSiW7dumHv3r3lrlten+rbt2/jzJkzAIDnn3++zDbdu3eHv78/CgoKKtyvpZg79qNHj2Lt2rUYO3YsQkNDzRaLp6cnOnbsKL5atWoFAGjVqpXBck9PzwfG0L59e0gkEly7dq0qH7NcwcHBZfpOa7VaXLlypUxfayIiotqMSXUNMmvWLEyZMgVOTk4YNmwY/P398fvvv6NHjx4PbHVctGgRRo4cCQAIDQ1F06ZNcerUKQwZMqTSrZXnzp0DUDJyRFBQULnrdOzY0WDdmsKcsefl5WHixImoW7cuPvnkE4vGUlkREREQBAGNGjUyel+hoaE4cuQIUlNTxWW7du1CXl4eBg8ebPT+iYiIbAWT6hpk3bp1OHjwIE6cOIH//ve/uHjxIlauXIni4mJMmDABKSkp5W63Zs0a/Pnnnzh58iR++OEHnD9/HosWLYIgCJg/f36ljn3z5k0AeOC4xv7+/gbrVsb48ePFlviqvI4ePVrpY5grdgCYM2cObt68iS+++AIeHh4WjaU8AwYMwMqVK7Fnzx4cPHgQq1atwqhRo9C6dWsMHz5cXG/Lli2QyWQ4duyYuEytVuPnn3/Gzz//jBs3bhjMx8fHAwCmTJmCOnXqYNiwYfj999/x3//+F6+88gqGDRsm/jggIiIiVlSsUaZMmYK+ffsaLJs7dy5++uknREZG4quvvsKbb75ZZrv33nsPjz/+uMGyBQsW4JNPPsGVK1eQmJgoJnIVyc3NBQAolcoK13F2dgYA5OTkVOrzACXdHR6Fr69vpdc1V+z79+/H+vXr8dxzzxkkqJaIpSKdO3fG1q1bxQQ9MDAQU6dOxaxZs2Bvby+up9PpoNVqDUYOSUlJEe9w6Onnv/nmG4wfPx7u7u44fPgwZsyYgfDwcDg4OGDkyJH48MMPjY6diIjIljCprkHGjRtX7vKxY8ciMjISR48eLTepLu/BNYVCgUaNGuHcuXO4ffv2Q5Nqc3nppZfw0ksvWeTYxsjOzsakSZPg7e2NtWvXWiSGwMDAcofPK+3999/H+++//9B9jR8/HuPHj6/y/gHgsccewx9//PHQ9YiIiGozJtU1SEV9cPXLb926Ve77FXU1cHV1BQDk5+c/9NguLi4AAJVKVeE6eXl5BvutKcwR+8yZM3Hr1i38+OOP8PLysmgsREREVPMxqbYiFbUqVqbc9MMEBgYCABITEytcR/+eft3K+Oqrr3Dy5MkqxzN//nw0b968UuuaI/YdO3ZAJpPh888/F4fB04uJiQEAbNy4EQcPHoSvry9++OEHs8XyMNeuXcOMGTNw/PhxODg4YNSoUVi1atUDu6BUddsff/wRH330Ef799184ODigXbt22Lp1K/z8/EzyGYiIiKwdk+oa5ObNm2jbtm2Z5fqh7xo0aGC2Y7dr1w4AkJ6ejps3b5bbah4ZGQkABmMvP8zJkyfLLVbyMOPHj690Um2u2IuLiw0e7LtfXFwc4uLi0LBhQ7PHUpHs7Gz07dsX9erVw7Zt25CRkYFZs2bh7t272L59u0m2/eijj7BgwQLMmTMHK1asgEqlwvHjxyt1B4SIiKjWEMjiAAgAhFmzZpX7fufOnQUAwpIlS8rdriK9evUSAAhHjhwxWN6wYUMBgHDz5k2D5Z06dSr3OIIgCCdOnBAACAqFQsjKyqrcB6tG1Rn7uHHjBADC+++/b/FYVqxYITg4OAh3794Vl23fvl0AIERGRhq97dWrVwW5XC58/vnnRsdKRERkyzikXg3yxRdflBlK7uOPP8bp06fh4uKCSZMmmfX4+ocgly9fjqioKHF5eno6pk6dCgCYPn36QysKWsKjxr5gwQI0b94cCxYssHgsj2Lv3r3o27cvfHx8xGVhYWFwdnbG7t27jd7266+/hr29vdnPPSIiImvHpLoG0Q+p16tXLzz//PNo3bo1Zs2aBTs7O3z99ddVGmbuUQwfPhwzZsxAXl4eunTpgkGDBiE8PBxNmjTBxYsX0a1bt0qNNGEJjxp7cnIyYmNjkZycbPFYHsXly5fLVDaUyWR47LHHylRCfJRtT506hebNm2Pz5s1o2LAhZDIZ2rRpg99//90k8RMREdkKJtU1yMcff4zPP/8cOTk5+PXXXxEfH4+BAwfi+PHjCA8Pr5YYVq9ejR9//BFPPPEETp06hb1796JBgwZYvnw5Dh8+DEdHx2qJ41HUpNirK5bMzEy4u7uXWe7h4YGMjAyjt71z5w5iY2OxaNEiLFmyBHv37kVAQADCwsJw6dIlU3wEIiIimyARhEoMVEtmJZFIAFQ8ugfVDps2bcKECRMeut62bdvEH1lyuRyLFi3C22+/bbDOk08+CalUiv3791e4n8ps+9hjj+Hq1avYs2cPQkNDAQCFhYVo3Lgx+vTpgy1btlT1YxIREdkkjv5BVEOMGDECXbp0eeh69evXF6c9PDyQlZVVZp3MzEw0bdr0gfupzLb60ux9+vQR37e3t0e3bt3YUk1ERFQKk2qiGsLNza3KDy8GBweX6Tut1Wpx5coVhIWFGb1tSEgITp8+XWZbQRA4pB4REVEp7FNNZMVCQ0Nx5MgRpKamist27dqFvLw8DB482Oht9cn1oUOHxHUKCgpw8uRJdOzY0ZQfhYiIyKqxTzWRFcvKykLLli0REBCAhQsXIjMzE7NmzUKXLl3w66+/iutt2bIFEydOxKFDh9CrV69Kb6vT6dC9e3fcuHEDy5Ytg6+vL9auXYujR4/izJkzCAkJscCnJiIiqnnYUk1kxdzd3XH48GG4uroiPDwcr776KoYPH46tW7carKfT6aDVag0ehq3MtlKpFLt370ZoaChmz56Np59+Gnl5eTh8+DATaiIiolLYUk1EREREZCS2VBMRERERGYlJNRERERGRkTikngnpdDokJSXBxcVFLOhCRCUEQUBubi7q1asHqZS/54mIyLYwqTahpKQk+Pv7WzoMohotMTERDRo0sHQYREREJsWk2oRcXFwAlCQNrq6u4nKtVotTp04BAB5//HH8/fffAICuXbvCzs6u+gMlMiWVCqhXDwCgTUzEqX/+AVD2/M7JyYG/v794nRAREdkSjv5hQjk5OXBzc0N2drZBUk1k01QqwNm5ZDovD1Aqy12N1wcREdkydmwkIiIiIjISu39Ug6KiImzYsAEAMGHCBHzzzTcAgMmTJ0Mul1syNCKTKioqwobPPgPA85uIiGoXdv8woYpub6tUKjjfuz1+9+5d1K1bFwCQl5cHZQW3yomsRqnuH6q7d+FcwfnN7h9ERGTL2P2DiIiIiMhITKqJiIiIiIzEpJqIiIiIyEhMqomIiIiIjMSkmoiIiIjISEyqiYiIiIiMxHGqq4FCocDu3bsBAK6uruK0QqGwZFhEJlf6XOf5TUREtQnHqTYhjsNLtRLLlBMREbH7BxERERGRsdj9oxoUFRXhu+++AwCMGjUKP/30EwBgzJgxLONMNqWoqAjfbdoEgOc3ERHVLuz+YUIsU061EsuUExERsaXaHM6fPy8m0QCg0WjE6QsXLlgiJCIiIiIyIybVZtCrV68K3+vfv784nZiYiObNm1dHSERERERkRkyqzWDDhg3o0KGDOK/RaNC9e3cAwIEDB8TEOj093SLxEREREZFpMak2g2bNmqF9+/bivEqlEqdbt25tiZCIiIiIyIw4pB4RERERkZHYUk1EJlP6Qdzz58/D0dFRnM/Ly7NESERERNWCSXU1UCgU4tjUrq6uWL58OebPn88xfMkmJCYmwv/e9JOlHsTVP0dARERUGzCprgYymQwjR44U5/v374/58+dDJuOfn6xfenq6mFRHnDwJXanW6dLOnj2LyZMnV19gRERE1YhZHRGZTNu2bYEKChqx+wcREdkyJtXVoLi4GDt27AAADB06FAcOHBCXE9mS4uJi7Ni2DQAwYsQI3o0hIqJag9941aCgoACjRo0CUFKmfP78+QCAoqIiS4ZFZHKlz/W8vDwm1UREVGtwSD0iIiIiIiMxqSYiIiIiMhKTaiIiIiIiIzGpJiIiIiIyEpNqIiIiIiIjMakmIiIiIjISx7uqBvb29vjmm28AAM7Ozli0aBEWL17M4cbI5pQ+1+3t7S0cDRERUfVhVlcN5HI5xo8fL86HhYVh8eLFkMvllguKyAzuP9eJiIhqC3b/ICIiIiIyEluqq0FxcTH27dsHAOjXrx9OnDghLieyJcXFxdi3Zw8AYMCAAeziREREtQa/8apBQUEBhgwZAqCkTPnMmTMBsEw52Z7S5zrLlBMRUW3C7h9EREREREZiUk1EREREZCQm1URERERERmJSTURERERkJCbVRERERERGYlJNRERERGQkjndVDezt7fHpp58CKClTPm/ePKxYsYLDjZHNKX2us0w5ERHVJszqqoFcLse0adPE+VGjRmHFihUsU0425/5znYiIqLZg9w8iIiIiIiOxpboaaLVasTR5165dERkZKS4nsiVarRYnjh4FAPTo0QN2dnaWDYiIiKiaMKmuBvn5+ejTpw+AkjLlU6ZMAQAUFhZaMiwikyt9rufl5UGpVFo4IiIiourB7h9EREREREZiUk1EREREZCQm1URERERERmJSTURERERkJCbVRERERERGYlJNRERERGQkDqlXDeRyOVauXAkAcHJywmuvvYbVq1ezTDnZnNLnOiuGEhFRbcKsrhrY29tj7ty54vzYsWOxevVqJh1kc+4/14mIiGoLdv8gIiIiIjISW6qrgVarRVRUFACgTZs2uHTpkricyJZotVpEnTkDAGjfvj3LlBMRUa3BpLoa5Ofno3PnzgBKypSPHTsWAMuUk+0pfa6zTDkREdUm7P5BRERERGQkJtVEREREREZiUk1EREREZCQm1URERERERmJSTURERERkJCbVRERERERG4pB61UAul2PRokUASsqUT548GRs2bGCZcrI5pc91VgwlIqLahFldNbC3t8e7774rzk+ZMgUbNmxg0kE25/5znYiIqLZg9w8iIiIiIiOxpboa6HQ6REdHAwCaNWuG69evi8uJbIlOp0P0pUsAgODgYEil/N1ORES1A5PqaqDRaNCyZUsAJWXKR40aBQAoKCiwZFhEJlf6XGeZciIiqk3YjEREREREZCQm1URERERERmJSTURERERkJCbVRERERERGYlJNRERERGQkJtVEREREREbikHrVQC6XY86cOQBKypS/+OKL+Pbbb1mmnGxO6XOdFUOJiKg2YVZXDezt7bFq1SpxfubMmfj222+ZdJDNuf9cJyIiqi3Y/YOIiIiIyEhsqa4GOp0OCQkJAIAGDRogKSlJXE5kS3Q6HRLi4gAAAQEBLFNORES1BpPqaqDRaBAUFASgpEz50KFDAbBMOdme0uc6y5QTEVFtwmYkIiIiIiIjMakmIiIiIjISk2oiIiIiIiMxqSYiIiIiMhKTaiIiIiIiIzGpJiIiIiIyEofUqwYymQxTp04FADg4OGDkyJHYtm0b7OzsLBwZkWmVPtdlMv7zQkREtQe/9aqBQqHAZ599Js7Pnz8f27Ztg729vQWjIjK9+891IiKi2oLdP4iIiIiIjMSW6mogCALS0tIAAJ6ensjMzBSXE9kSQRCQlpoKAPDy8oJEIrFwRERERNWDSXU1UKvV8PHxAVBSpvzJJ58EAOTn51syLCKTU6vV8KlbFwDLlBMRUe3C7h9EREREREZiUk1EREREZCQm1URERERERmJSTURERERkJCbVRERERERGYlJNRERERGQkDqlXDWQyGcaNGwegpEz5kCFDsHv3bpYpJ5tT+lxnmXIiIqpN+K1XDRQKBTZt2iTOL168GLt372aZcrI595/rREREtQW7fxARERERGYkt1dVAEASo1WoAgKOjIzQajbicyJYIggC1SgUAcHJyYplyIiKqNZhUVwO1Wg1nZ2cAJWXKu3fvDoBlysn2qNVqOLNMORER1ULs/kFEREREZCQm1URERERERmJSTURERERkJCbVRERERERGYlJNRERERGQkJtVEREREREbikHrVwM7ODuHh4QAAe3t79OvXD4cOHYJUyt80ZFtKn+t2dnYWjoaIiKj6MKmuBg4ODti2bZs4v3LlSnTo0AEKhcKCURGZ3v3nOhERUW3BplIiIiIiIiMxqSYiIiIiMhKT6mqgUqkgkUggkUiQkpKCDh06AAA0Go2FIyMyrdLnukqlsnQ4RERE1YZJNRERERGRkZhUExEREREZiUk1EREREZGRmFQTERERERmJSTURERERkZGYVBMRERERGYkVFauBnZ0dQkNDAZSUKe/WrRsiIiJYppxsTulznWXKiYioNmFSXQ0cHBywZ88ecX7NmjUsU0426f5znYiIqLZgUykRERERkZGYVBMRERERGYlJdTVQqVRQKpVQKpVISUlBt27dALBMOdme0uc6y5QTEVFtwj7V1UStVovT+fn5FoyEyLxKn+tERES1BVuqiYiIiIiMxKSaiIiIiMhITKqJiIiIiIzEpJqIiIiIyEhMqomIiIiIjMTRP6qBVCpFr169AAAymQwdOnTA2bNnIZFILBwZkWmVPtelUv5mJyKi2oNJdTVwdHTE0aNHxfkNGzagQ4cOcHBwsFxQRGZw/7lORERUW7ApiYiIiIjISEyqiYiIiIiMxKS6GqhUKnh7e8Pb2xspKSno168fAJYpJ9tT+lxnmXIiIqpN2Ke6mqSlpYnTWVlZlguEyMxKn+tERES1BVuqiYiIiIiMxKSaiIiIiMhITKqJiIiIiIzEpJqIiIiIyEhMqomIiIiIjMTRP6qBVCpFx44dAZSUKW/RogUuX77MMuVkc0qf6yxTTkREtQmT6mrg6OiIM2fOiPPffvsty5STTbr/XCciIqot2JRERERERGQkJtVEREREREZiUl0N1Go1AgMDERgYiLS0NAwZMgQAy5ST7Sl9rqvVakuHQ0REVG3Yp7oaCIKA+Ph4AIBOp0NycrKFIyIyj9LnuiAIFo6GiIio+rClmoiIiIjISEyqiYiIiIiMxKSaiIiIiMhITKqJiIiIiIzEpJqIiIiIyEgc/aMaSCQStGjRAkBJ6eZGjRrhxo0bFo6KyPRKn+sSicTC0RAREVUfJtXVwMnJCZcuXRLnt23bhg4dOsDR0dGCURGZ3v3nOhERUW3B7h9EREREREZiUk1EREREZCQm1dVArVYjJCQEISEhSEtLw8iRIwGwTDnZntLnOsuUExFRbcI+1dVAEARcvnwZQEmZcj6kSLaq9LnOMuVERFSbsKWaiIiIiMhITKqJiIiIiIzEpJqIiIiIyEhMqomIiIiIjMSkmoiIiIjISBz9oxpIJBI0bNgQQEmZcj8/PyQnJ1s4KiLTK32us0w5ERHVJkyqq4GTkxPi4uLE+d27d7NMOdmk+891IiKi2oLdP4iIiIiIjMSkmoiIiIjISEyqq4FGo0GnTp3QqVMnZGRk4MUXXwQA5OfnWzgyItMqfa5rNBpLh0NERFRt2Ke6Guh0OkRGRgIAiouLWcaZbFbpc12n01k4GiIiourDlmoiIiIiIiMxqSYiIiIiMhKTaiIiIiIiIzGpJiIiIiIyEpNqIiIiIiIjcfSPauLl5SVOu7u7Iysry3LBEJlR6XOdiIiotmBSXQ2USiVSU1PF+UOHDrFMOdmk+891IiKi2oLdP4iIiIiIjMSkmoiIbN4vv/yC7t27w8vLCwqFAo0aNcKsWbOQmZlp0m1/++03dOnSBa6urvDx8cHgwYNx/vx5M3wiIqppmFRXA41Gg969e6N3797IyMjA5MmTAbBMOdme0uc6y5RTTZKRkYHevXtj48aN2LdvH15//XVs2bIF4eHhJtv2wIEDGDFiBJo2bYqff/4Z69evR0pKCvr164fk5GRzfTQiqiHYp7oa6HQ6HDt2DEBJmfKzZ88CYJlysj2lz3WWKafqEhcXh6CgINy8eROBgYHlrvPSSy8ZzPfu3RsODg6YPHkyEhISEBAQUOH+K7vtf//7XzRs2BBbtmyBRCIBALRp0waNGzfGvn37MH78+Ef/kERU47GlmoiIaqU6deoAAIqKikyybVFREZydncWEGgDc3NwA8EcmUW3ApJqIaqU2bdpAIpFAoVAgPT3d0uFUm9jYWKxduxbjx49Hq1atIJPJIJFIsGTJEqP3XVhYiDVr1qB79+6oU6cOHBwc0KBBAwwaNAg//vij2batCq1Wi/z8fERGRmLx4sUIDQ1F48aNTbLtxIkTERMTg48//hiZmZlITEzEq6++Cn9/fzz99NMm+wxEVDOx+wcR1TpnzpzBhQsXAJQkc1u3bsVrr71m4aiqxxdffIHVq1ebfL+3bt3CgAEDcPnyZXh5eaFbt25QKpVITEzE8ePHoVQq8eyzz5pkW0EQoNVqxXn9tFarRXFxsbjczs7OoNUYADw9PZGdnQ0AeOqpp/DTTz9V+jM+bNu+ffvil19+wZgxYzBr1iwAQFBQEA4ePAh3d/dKH4eIrBNbqomo1tm4cSMAoH79+gbztUHLli0xZ84cfPfdd4iOjsaLL75o9D41Gg369++Py5cv491330VSUhJ27dqFH374AREREUhNTcXbb79tsm03b94MuVwuvpo0aQIAaNKkicHyzZs3lzne0aNHERERgXXr1uHy5csYOnSoQYL+IA/b9q+//sLYsWPx/PPP4+DBg/j111/h5+eHQYMG4e7du5X9cxKRlWJLNRHVKmq1Gt9//z0A4Ntvv0VYWBguXryIM2fOoFOnThaOzvzuf+hOKjW+bWXZsmWIiYnB5MmTsWjRojLvOzk5oW3btibbdujQoThz5ow4n5ycjLCwMOzcuRN+fn7i8qCgoDL70++ra9euaNu2Lbp06YIdO3ZUahSQh2376quvomvXrli3bp24TZ8+fdCwYUN8/PHHWL58+UOPQUTWi0l1NXFychKnHRwcOJwe2azS53pNtG3bNuTk5KBly5bo06cPnn32WWzcuBEbN26sFUm1qRUVFeGLL74AAMydO7datvX09ISnp6c4HxcXBwBo1apVhaN/lKd9+/aQSCS4du1apbd50LaXLl3CoEGDDNZzdXVFkyZNcPXq1Sofg4isC5PqaqBUKqFSqcT5iIgIliknm3T/uV4T6bt6TJw4Ufzvxo0b8cMPP+Djjz/mdVlFUVFRSEtLQ7169dCkSRNcvHgRv/zyC5KSkuDh4YEePXpg0KBB5baIG7OtKUREREAQBDRq1Mgk2wYGBiIyMtJgvZycHFy7dg29evUyOl4iqtmYVBNRrXHlyhWcOHECcrkcL7zwAoCSW/nNmzdHTEwMfv755yr3MR4/fny5fXcf5siRI+jdu3eVt6tp9A98NmjQAPPnz8fKlSsNxuBfsWIF2rVrh19//bXMWNDGbFtVAwYMQL9+/RASEgKFQoFz585h1apVaN26NYYPHy6ut2XLFkycOBGHDh0SE+HKbjt9+nRMmzYNU6ZMwTPPPIO8vDx89NFHKCgowMsvv2xU/ERU8zGpJqJa4+uvvwYAhIWFwdvbW1w+ceJEvPHGG9i4cWOVk+ru3bs/Uiy+vr6PtF1Nox+O8Ny5czh9+jSmTZuGGTNmwNfXV5w/d+4cBg8ejKioKMjlcpNsW1WdO3fG1q1bcfPmTQAlrcpTp07FrFmzYG9vL66n0+mg1WoNkvvKbvt///d/cHBwwKeffor//ve/cHR0RPv27XHkyBE0a9bskWMnIishkMlkZ2cLAIRjx44ZLNdoNEJoaKgQGhoqZGZmCt26dRMACKdOnbJQpESmc+7kSUEABAEQNGlp4rmu0WgM1jt27JgAQMjOzrZInEVFRYKvr68AQNizZ4/Be3fu3BFkMpkgkUiEa9euWSS+h5k7d67QrFmzKr9OnDjxwP2OGzdOACC8//77jxTXBx98IAAQAAijR48u8358fLzg4OAgABC2bNlism2JiGoatlRXA61Wi7179wIoGRM3IiICACtske0pfa5Xdpiy6rJnzx7cuXMH9evXx4ABAwzeq1u3LkJDQ7Fz5058/fXXWLp0qYWirFhSUhJiY2OrvF1eXp4ZovkfFxcXcXrKlCll3g8ICMDgwYOxfft2HDx40OBOgDHbEhHVNEyqiahW0D+gmJ+fX+5DY7dv3wYAbNq0Ce+99x7s7Owqtd+vvvoKJ0+erHI88+fPR/PmzSu9/tatW7F169YqH8fcSj+oV9EDf/rlycnJJtv2UVy7dg0zZszA8ePH4eDggFGjRmHVqlVQKpUm2fbo0aPo06dPmW1DQkLw77//Gh0/EdVsTKqJyOYlJyeLLejp6eni3aLyJCUl4Y8//sDgwYMrte+TJ08+0oOK48ePr1JSXVPph5YTBAFpaWnw9/cvs05aWhoAwNnZ2WTbVlV2djb69u2LevXqYdu2bcjIyMCsWbNw9+5dbN++3aTbfvXVVwgJCRHna/owk0RkGkyqLSgmJuahw3d5eXkZ/dQ7UW23adMmaLVaPP744/jrr78qXG/evHlYuXIlNm7cWOmketOmTdi0aZOJIrU+vr6+6N69O06cOIGDBw+iXbt2Bu8XFRXh2LFjAEoe+DPVtlW1fv16pKamIjIyEj4+PgAAR0dHPPPMMzh79iw6dOhgsm1DQkLQpUsXo+IlIuvDpNqC7q9sVh4nJydER0czsSaLSEhIEFsKK3IlJgZtqyecR6Yf9WPcuHEPXG/s2LFYuXIldu/ejdTUVIMRQmq7BQsWYMeOHRgxYgSWLVtm8N6iRYvw5JNPYtmyZejRo4eYUBYXF2P27Nm4ceMGXFxcMGHChDL7NWbbqti7dy/69u0rJsVAySgwzs7O2L179wOTamO2JaLag0m1BZ08efKBLdXR0dF44YUXkJaWxqSaql1CQgKCg4OhVqsfuJ4TgOeqJ6RHcuzYMVy7dg0KhQLPPffgSENCQtC+fXtERUVhy5YtmD17djVFWX2ioqIwdepUcf769esASlpjd+/eLS7fsWOHQcnv5ORkxMbGltu3uV+/fnj//fexcOFC9OjRA507d4avry+ioqIQFxcHR0dHfP/996hbt65Jt62Ky5cvY+zYsQbLZDIZHnvsMURHR5t022HDhiEtLQ1eXl4YNmwYli9fjjp16hgVPxHVfEyqTUi4N66pSqVCTk6OuLx0hbnc3FxxulGjRg98QEb/1P7Zs2fN/gQ/0f1iY2OhVquxYcOGB46xK83PR8690TRUpc7vnJwcgxFA9NeBUGr83+qgf0Bx6NCh8PDweOj6Y8eORVRUFDZu3GiTSXVOTg7+/vvvMstv3bqFW7duifMFBQVV2u/bb7+Nzp0745NPPsHff/+NM2fOwNfXF+PHj8e8efMe2H/cmG0rKzMzE+7u7mWWe3h4ICMjwyTburm5Yfbs2ejduzecnZ3x999/Y9myZfjzzz8RGRkJhUJh7McgohpMIlT3N5wNu3XrVrkP2hDR/yQmJqJBgwaWDoOs2KZNmyrVHWTbtm0IDw8HAMjlcixatAhvv/22wTpPPvkkpFIp9u/fX+F+jNn2wIEDeOqpp/DNN99g/PjxD42ZiKwXW6pNqF69ekhMTISLiwskEonF4sjJyYG/vz8SExPh6upqsTgehTXHDlh3/OaOXRAE5Obmol69eibfN9UuI0aMqNSDgPXr1xenPTw8kJWVVWadzMxMNG3a9IH7MWbb/v37o06dOjhz5gyTaiIbx6TahKRSaY1qgXN1dbW6xE7PmmMHrDt+c8bu5uZmlv1S7eLm5lblcyk4OLhM/2etVosrV64gLCzMbNvqWbKhhYiqh9TSARAREZlbaGgojhw5gtTUVHHZrl27kJeX99DhE43Zdt++fcjIyDB6SEAiqvnYp9oG5eTkwM3NDdnZ2VbXWmrNsQPWHb81x070MFlZWWjZsiUCAgKwcOFCZGZmYtasWejSpQt+/fVXcb0tW7Zg4sSJOHTokFh5s7LbvvDCCwgKCkKHDh3g4uKCv//+GytWrECjRo3w999/w97evpo/NRFVJ7ZU2yCFQoFFixZZ5ZPm1hw7YN3xW3PsRA/j7u6Ow4cPw9XVFeHh4Xj11VcxfPjwMqXfdTodtFqtwSg1ld02JCQEv/76K8aOHYuBAwfiyy+/xKRJk3D06FEm1ES1AFuqiYiIiIiMxJZqIiIiIiIjMakmIiIiIjISh9QzIZ1Oh6SkJIuPU01UE5Uep1oqrfj3PK8joopV9joiourHpNqEkpKSWFGR6CEeVlGR1xHRw7EyKVHNw6TahFxcXADAKqvpEVVJYiLQoQNQUFDhKhoAzzRsiDqtWyMkJAShoaHo2bOneJ1U5P7rSKvV4tSpUwCAxx9/HH///TcAoGvXrrCzszPN5yGqQbRaLf46fBjd7pVYR1ISoFQC+F/l04ddR0RU/ZhUm5D+VrU1V9MjqpSCggcm1ADgCkAdH4+/bt1CfHw8GjZsCODhleXKu45KF9h4WLENIlswaODA/824uopJtR67RhHVPEyqzUGlAtiCRrZMo6nUao4AFFotrl24gB82bjRvTERERBbEpNoc6tWzdARENUJEqemcyEi4PcI+ioqKsGHDBgDAhAkT8M033wAAJk+eDLlcbnyQRDVMUVERvl6/HlMsHQgRVQmLv5iQWOYZJbe+ieh/cgC4AQ8tg35/uXSVSgVnZ2cAwN27d1G3bl0AQF5eHpT33RInsgUqlQo+zs5Q6Rfk5Rn0qS59fRBRzcGWanNISirpA0dkg1QqFeJ37kSLl19+6LrdAJy/N+3m5gZkZ5szNCIiIothUm0OSmWZh0qIbEl+JdfTAFDfm+7Xsyd27dplpoiIiIgsiyPHE1GVqFQqyKrYl7l58+b4v//7PzNFREREZHm1uqVaq9VynFuiKlIqlXCoXx/FMhlkxcUVrqcBkAagXbt2WL9+PZo1a1ZtMRIREVW3WttSfe3aNWzYsAG3b9+2dChEVkWpVKL+E0/g7H//iyVPP40xgYHie90AtL/3aqNQoMPw4fj222/RqVMnC0VLRERUPWplS/WFCxfQu3dvTJgwAWp1SY9PnU4HqbRqvzEKCgpQUKoARk5OjknjJKqplEolWoaG4o69PaTNmwMffAAA0Dz2GPIEAa1atULv3r3RpUsXBJZKuomIiGxVrUuqk5OTER4ejkmTJmHVqlXicrVaLQ7bVVnLli3D4sWLTR0ikVVQKpV44okn0MDDQ0yq582bh0atWpm0ZVqhUGD37t0ASqos6qcVCoXJjkFUkygUCmz/+WdAX6aciKxCrRun+tChQ3j77bdx8uRJAMCsWbMQExMDjUaDkSNH4tVXX630vsprqfb39+f4oVS7qFTAvR+kKTduQOnjU+740ZUdX5fj8BLB4LriONVE1qHWtVQnJibCzs4OdnZ26NOnDxwdHdG5c2doNBrMnDkT8fHx+PDDDyu1L4VCwdYyolJ8fHw4nCQREdVKtS6pbtasGS5duoSPPvoITk5O2LBhA+rXrw8A6NatG8LDw9GnTx8MHjzYwpESEVBSsvm7774DAIwaNQo//fQTAGDMmDEsU042qaioCD9u3YoXLB0IEVVJrUqqBUFAcHAwwsLC8N1330Gr1YoJtSAI6N+/P9q0aYObN29aOFIi0issLMSECRMAAKGhoeL0yJEjmVSTTSosLMSUV15hUk1kZWrVkHoSiQTu7u4YOnQo8vLycPHiRezfv198z9nZGR4eHnBwcLBwpER0/vx5REVF4fz58+KyCxcuGLyfkJBggciIiIjKqnUt1RKJBOHh4dDpdHjvvfcwadIkvP/++2jUqBF+//13REdHo1+/fpYOlajW69WrV5ll/fv3F6e7d+8OJycnREdHIyAgoDpDIyIiKsMmk+qKKiVKJBJxPOpRo0bBy8sLP/30E/7v//4PTZs2hUQiwZ49exAUFGSBqImotA0bNqBDhw7QaDTo3r07AODAgQNiYv3VV1/hpZdeQlpaGpNqIiKyOJtLqq9cuYJdu3bh+eefh5+fX5n3pVIpiouLIZPJ0LdvX/Tt2xfvvPMO5HI5ZDIZPDw8LBA1Ed2vWbNmaN++PVQqlbisdevW4nTz5s0tERYREVG5bCqpvnbtGp544glkZmYiPT0ds2bNgpeXl8E6giBAJjP82L6+vlWupkhEREREpGczSbVKpcKyZcsQFhaGTp06Yfr06SguLsYbb7xhkFhLJBIAwKpVq5Cfn4+FCxcyoSYiIiIio9hMUi2VStGhQwd4enri2WefhZeXF5577jkAKJNYZ2Rk4OzZs4iLi8O0adNQp04dS4VNRA+hUCjEsaldXV3FaQ6nR7YmISEBaWlpKC4uxvvvvQe88w6AkpFudI6OAIC8vDxLhkhED2AzSbWjoyPGjRsnlkceNWoUBEHA6NGjIQgC5s+fD09PT2i1WkilUnz++ecoKChgQk1Uw8lkMowcOVKc109HRUVZKiQik0tISEBwcDDUajUAwAnArHvvdeveHWqLRUZElWUzSTUAMaHWJ87PPvssBEHA888/D4lEgpkzZ2LVqlWIi4vDDz/8wISaiIhqhLS0NKjVamzduhXBwcGQajTAvVFvIk6eFFuqz549i8mTJ1syVCKqgE0l1Xp2dnYQBAE6nQ7PPfccJBIJXnzxRezcuRPXr1/H6dOnoVAoLB0mEVVCcXExduzYAQAYOnQodu3aBQBo2LChJcMiMovg4GC0bt0aO7//Hm3vLWvbti1wr9GI3T+Iai6bTKqB/z2QKAgCnn32WWzYsEGs0NaqVSsLR0dElVVQUIBRo0YBAO7evStOnzx50pJhEZlNQUEBXhw7Fk9bOhAiqhKbTaqBksRaq9Vi7ty5OHLkCM6fP8+EmoiIiIhMrlaMJRcSEoKoqCiDwhFERERERKZi0y3VQEn/6okTJ4rdQYiIiIiITK1WtFQzoSYiIiIic6oVSTURERERkTkxqSYiIiIiMpLN96kmIutmb2+Pb775BgDg7OwsTstk/OeLbJO9vT3Wr1sHvPKKpUMhoirgtxIR1WhyuRzjx48X5/XTLFNOtkoul+OFF15gUk1kZdj9g4iIiIjISGypJqIarbi4GPv27QMA9OvXD4cOHQIAeHt7WzIsIrMpLi7GwT/+wEBLB0JEVcKkmohqtIKCAgwZMgRASZly/TTLlJOtKigowDPh4VBZOhAiqhJ2/yAiIiIiMhKTaiIiIiIiIzGpJiIiIiIyEpNqIiIiIiIjMakmIiIiIjISk2oiIiIiIiNxSD0iqtHs7e3x6aefAigpU66fZplyslX29vb4z0cfAbNnWzoUIqoCfisRUY0ml8sxbdo0cV4/zTLlZKvkcjmmTJnCpJrIyrD7BxERERGRkdhSTUQ1mlarxYkTJwAAXbt2xalTpwAASqXSkmERmY1Wq0XE8ePoaelAiKhKmFQTUY2Wn5+PPn36ACgpU66fZplyslX5+fkYFBrKMuVEVobdP4iIiIiIjMSkmoiIiIjISEyqiYiIiIiMVGuTakEQDOZ1Op2FIiEiIiIia1crH1S8evUqNmzYAJVKhcDAQLzxxhuQSmvt7wsiIiIiMlKtS6ovXryIvn37ok+fPkhPT8dff/0FV1dXvPLKKwBKWrAlEkml9lVQUICCggJxPicnxywxExEREVHNVquS6rS0NLzwwguYOHEiVqxYgezsbIwePRoajUZcRyKRQKfTVarletmyZVi8eLE5Qyaq9eRyOVauXAkAcHJyEqdZppxslVwux9IlS4C337Z0KERUBbXqWykhIQGFhYWYPHkyAMDNzQ2+vr44efIkIiMj4ebmhs8//xxSqbRSifWCBQswa9YscT4nJwf+/v5m/QxEtY29vT3mzp0rzuunWaacbJW9vT1mzpzJpJrIytSqjsRKpRJqtRpbt25FcXEx3n//fXz77bdo2rQpfHx8cPjwYfTo0QMAKtVSrVAo4OrqavAiIiIiotqnVrVU+/n54bnnnsOXX36JiIgIHDlyBD/++COefvppAEBYWBief/55HDt2DL169bJwtEQElJRs1rdKt2nTBv/884+FIyIyL61Wi/Nnz6KDpQMhoiqpVUm1q6sr3n77bbzyyitITExEcnIyevbsafC+s7MzXFxcLBglEZWWn5+Pzp07AygpU66fZplyslX5+fno2asXy5QTWZla1f0DAFxcXBAUFIT69etDoVAgOjpafO+3336Ds7Mz6tevb8EIiYiIiMja2GRLtU6ngyAIsLOzM1hWup+0i4sLJBIJ3n//fXh4eMDZ2Rk7duzA4cOHUbduXUuETURERERWyuZaqi9fvoyxY8diwIAB+L//+z/s2bMHQMmDh1qtFkDJWNQ+Pj7YsmULGjdujJycHMjlckRERKBt27YWjJ6IiIiIrJFNtVTHxsaia9euGDRoEDp16oTff/8dkZGROHjwID7++GPY2dmhsLAQ9vb20Ol0CA4Oxn/+8x84OjqiqKgIcrnc0h+BiIiIiKyQzbRUC4KALVu2YMCAAfj++++xbNkynDhxAsOHD8fRo0fFsant7e0BALt27UJKSgocHR0BsJAEERERET06m0mqJRIJkpKScOfOHXGZi4sLZsyYgRdeeAHnzp3D8uXLAQB79uzB9OnTsWbNGuh0OnF7IiIiIqJHYRPNs4IgQCKRoH379rh69SpiY2PRrFkzACWJ9cSJExEbG4tdu3Zh1qxZGDx4MCZOnIhx48ZVqsgLEVmOXC7HokWLAJSUKddP8+4S2Sq5XI43FywAli2zdChEVAU2kVHqW5lDQ0MRGxuLlStXIi8vD0BJwu3h4YGFCxfizz//xP79+wEAixcvRqNGjSwWMxFVjr29Pd599128++67cHZ2Fqf5DATZKnt7e7z11luWDoOIqsimmnoaN26Mn376CYMGDYKjoyPeffddeHl5ASj55d+6dWt4enpaOEoiIiIisjU2lVQDQJ8+fbBt2zaMHDkSycnJGDVqFFq3bo0tW7YgJSUF/v7+lg6RiKpAp9OJRZqaNWuG2NhYcTmRLdLpdIi5fBktLB0IEVWJzSXVADB06FCcOnUKs2bNwrx58yCTyWBnZ4c9e/agQYMGlg6PiKpAo9GgZcuWAErKlOunWaacbJVGo0Gnzp1ZppzIythkUg0A7du3x86dO5GRkYHc3Fz4+fmJXUGIiIiIiEzJZpNqAHB1dYWrq6ulwyAiIiIiG2cTo38QEREREVkSk2oiIiIiIiMxqSYiIiIiMhKTaiIiIiIiI9n0g4pEZP3kcjnmzJkDoKRMuX6aZcrJVsnlcrw2YwawZo2lQyGiKuC3EhHVaPb29li1apU4r5+OioqyVEhEZmVvb48PPviASTWRlWH3DyIiIiIiI7GlmohqNJ1Oh4SEBABAgwYNcOvWLXE5kS3S6XRIjI9HQ0sHQkRVwqSaiGo0jUaDoKAgACVlyvXTLFNOtkqj0aBFSAjLlBNZGXb/ICIiIiIyEpNqIiIiIiIjMakmIiIiIjISk2oiIiIiIiMxqSYiIiIiMhKTaiIiIiIiI3FIPSKq0WQyGaZOnQoAcHBwEKft7OwsGRaR2chkMkx++WXgyy8tHQoRVQGTaiKq0RQKBT777DNxXj/NMuVkqxQKBT7++GMm1URWht0/iIiIiIiMxJZqIqrRBEFAWloaAMDT0xPp6eniciJbJAgC0lJT4W3pQIioSthSTUQ1mlqtho+PD3x8fJCWliZO5+fnWzo0IrNQq9UIDAqydBhEVEVMqomIiIiIjMSkmoiIiIjISCZNqv/9919T7o6IiIiIyCoYnVTn5uZiw4YN6Ny5M9q0aWOKmCyGDz4RERER0aN45NE/jh8/jo0bN2L79u2oV68enn76aYOxZGu6W7du4dKlS8jJyUHnzp3RsGFDSCQS6HQ6SKXsFUNERERElVelpPrOnTvYtGkTNm7ciJycHIwaNQoFBQX49ddf0aJFC3PFaHIXL15E//79ERAQgKioKLRr1w5PPPEE1qxZA6lUWunEuqCgAAUFBeJ8Tk6OOcMmIiIiohqq0k2yQ4cORbNmzXDhwgV88sknSEpKwtq1a80Zm1lkZ2fjxRdfxOjRo3HgwAHEx8dj2LBhOHLkCIYMGQIAYmL9MMuWLYObm5v48vf3N3f4RLWOTCbDuHHjMG7cODg4OIjTLFNOtkomk2HM889bOgwiqqJKJ9W///47Jk2ahMWLF2Pw4MFW+4WWnZ0NjUaDUaNGwc3NDfXr18fMmTPxzjvv4Nq1axg1ahQAVKqlesGCBcjOzhZfiYmJ5g6fqNZRKBTYtGkTNm3aBFdXV3Ha3t7e0qERmYVCocCGDRssHQYRVVGlk+qTJ08iNzcXHTp0wOOPP45PP/1UrHJmTVxcXFBUVIRTp06Jy5ydnREWFoY333wTsbGxWL9+faX2pVAo4OrqavAiIiIiotqn0kl1ly5d8OWXXyI5ORlTpkzBDz/8gHr16kGn0+HAgQPIzc01Z5wm4+TkhJ49e+LgwYO4ePGiuFyhUCA8PByBgYE4evSo5QIkIgOCIEClUkGlUkGn04nTHK2HbJX+nCci61LlYS6USiUmTpyIkydP4uLFi5g9ezaWL18OHx8fhIWFmSNGk1IoFJgzZw7OnTuHJUuW4Pr16+J7Tk5O6NWrF65cuQK1Wm3BKIlIT61Ww9nZGc7OzkhLSxOnWaacbJVarYZP3bqWDoOIqsioseOaNWuGlStX4tatW/j+++9NFZNZ6XQ6tGzZEr/99hv27NmD+fPn48iRI+L7MTExaNCgAWSyRx5tkIiIiIhqGZNkjnZ2dhg+fDiGDx9uit2ZhE6ngyAIBg9U6ofK02q1ePzxx3Hs2DG89NJLmDNnDrRaLQIDA3HkyBEcP36cD0ERERERUaXZZHPs5cuX8cEHH+DOnTto2rQphgwZgsGDB4sJtZ2dHbRaLTp06IDffvsNZ8+exeHDh+Hv74/ly5ejefPmlv4IRERERGRFbC6pjo2NRdeuXTFo0CB06tQJv//+OyIjI3Hw4EF8/PHHsLOzQ2FhIezt7SEIAgICAhAQEIARI0ZYOnQiIiIislI2VY9bEARs2bIFAwYMwPfff49ly5bhxIkTGD58OI4ePYrJkycDgNi1Y+fOnUhJSbFkyERERERkA2wqqZZIJEhKSsKdO3fEZS4uLpgxYwZeeOEFnDt3DsuXLwcA7NmzB9OnT8eaNWsqVT2RiIiIiKgiNpNU68esbd++PbRaLWJjY8X3XFxcMHHiRLRr1w67du1CYWEhBg8ejIkTJ2LixImVqp5IRJZhZ2eH8PBwhIeHw97eXpzmdUu2ys7ODiNq0IP/RFQ5NvOtJJFIAAChoaGIjY3FypUrkZeXB6Ak4fbw8MDChQvx559/Yv/+/QCAxYsXo1GjRhaLmYgezsHBAdu2bcO2bdvg7u4uTisUCkuHRmQWDg4O2Lp1q6XDIKIqsrkHFRs3boyffvoJgwYNgqOjI9599114eXkBAORyOVq3bg1PT08LR0lEREREtsTmkmoA6NOnD7Zt24aRI0ciOTkZo0aNQuvWrbFlyxakpKTA39/f0iESERERkQ2xme4f9xs6dChOnTqF9PR0zJs3D0OHDsUvv/yCPXv2oEGDBpYOj4gqSaVSQSKRQCKRICUlRZzWaDSWDo3ILFQqFZTOzpYOg4iqyCZbqvXat2+PnTt3IiMjA7m5ufDz8xO7ghARERERmYpNJ9UA4OrqCldXV0uHQUREREQ2zGa7fxARERERVRcm1URERERERmJSTURERERkJCbVRERERERGsvkHFYnIutnZ2SE0NBQAYG9vL06zTDnZKjs7Owx46ingXvVfIrIOTKqJqEZzcHDAnj17xHn9dFRUlKVCIjIrBwcH/PLLLwDHqiayKmzqISIiIiIyEpNqIiIiIiIjMakmohpNpVJBqVRCqVQiJSVFnGaZcrJVKpUK3t7elg6DiKqIfaqJqMZTq9XlThPZKjV/NBJZHbZUExEREREZiUk1EREREZGRmFQTERERERmJSTURERERkZGYVBMRERERGYmjfxBRjSaVStGrVy8AgEwmE6clEoklwyIyG6lUih7duwMnT1o6FCKqAibVRFSjOTo64ujRo+K8fpplyslWOTo64o8//mCZciIrw+4fRERERERGYlJNRERERGQkJtVEVKPpSzZ7e3sjJSVFnGaZcrJVKpUKDRs2tHQYRFRF7FNNRDVeWlpaudNEtiotPd3SIRBRFbGlmoiIiIjISEyqiYiIiIiMxKSaiIiIiMhItbZPdXp6OpKTk+Ho6Ah/f3/Y29tbOiQiIiIislK1sqX64sWL6Nu3L5577jm0atUKa9aseaT9FBQUICcnx+BFRERERLVPrUuqr169iieffBIDBgzA9u3bMW/ePCxatAjpj/Ck9bJly+Dm5ia+/P39zRAxUe0mlUrRsWNHdOzYETKZTJxmmXKyVVKpFO3btbN0GERURbWq+4cgCFi3bh169OiBlStXAgDmzZuHv//+G/Hx8UhOToanpyf8/Pwqtb8FCxZg1qxZ4nxOTg4TayITc3R0xJkzZ8R5/TTLlJOtcnR0xIkTJ1imnMjK1KqkWiKRIDc3FxKJBHl5eXB2dsbKlSuxb98+3L59G2lpaejUqRPeeecdtG/f/qH7UygUUCgU1RA5EREREdVkta77R8OGDXHo0CHMmDEDEydOxNKlS/Hjjz/i6NGj2LhxI+7evYv9+/dbOkwiIiIisiK1JqkWBAEA8NZbb+H1119HQEAAkpOTMX36dISHh8PDwwMDBw5E3bp1cfToUcsGS0QitVqNwMBABAYGIi0tTZxmmXKyVWq1GsHBwZYOg4iqqNZ0/5BIJNBqtbCzs8PChQsBAC+//DI8PT0BADqdDlKpFC4uLmjUqJG4LhFZliAIiI+PB1ByneqniWyVIAhISEy0dBhEVEW1JqkGUCZJdnNzw2effYawsDAUFxfjt99+w++//44TJ04woSYiIiKiSqtVSbWeIAiQSCR4++23ceHCBbRt2xZNmjSBXC7HwYMHeduNiIiIiKrE5pLqa9euYcuWLSgsLET9+vXx6quviu/pk2k9d3d37N+/H3v27IGPjw/8/f3h6+tribCJiIiIyIrZVFJ96dIldO3aFU888QTy8/Nx/vx5fP/991i6dCl69OgBmUwm9p0GgPz8fDg4OGDw4MEWjpyIiIiIrJnNjP5RUFCAt956C88++yz++OMPHDhwAFeuXEFhYSHmzJmDP/74wyChnj17Nt5++22oVCoLR05ERERE1s5mkmqFQoG8vDyxGqJEIoGPjw+OHz8OpVKJd955B9evXxfXb9CgAb755huo1WpLhUxElSCRSNCiRQu0aNECUqlUnCayVRKJBMHNm1s6DCKqIptJqnU6HXQ6HaKjowEAMpkMhYWFcHJywv79+5GZmSkOpQcAr7/+Oq5fvw5vb29LhUxEleDk5IRLly7h0qVL8PLyEqcdHR0tHRqRWTg5OSEyMtLSYRBRFdlEUi0IAqRSKRYuXIi9e/fi448/BgDY29tDo9HAwcEBa9euxcmTJxEbGysWgnF3d7dg1ERERERkK2wiqdaP6NGxY0fMnDkTa9euxaeffgoAYmuWg4MDHBwc4OzsbDACCBERERGRsWwiqQaA4uJiuLi4YMKECRgxYgSWLl2KhQsXIicnBxkZGTh27BgcHR3h4OBg6VCJqArUajVCQkIQEhKCtLQ0cZplyslWqdVqdOzY0dJhEFEV2cSQelqtFjKZDHFxcYiKisJrr72GRo0a4a233sLmzZvh6uqK9PR07NmzRyxLTkTWQRAEXL58GUDJsxP6aSJbJQgComNiLB0GEVWR1SfVxcXFYkLdtGlTPP/88xg5ciSmTZuGESNG4Pjx43B2dkbr1q0REBBg6XCJiIiIyAZZdVJdOqFu3749xo4di3Xr1gEoadGqV68ennvuOQtHSURERES2zmqT6vsT6rCwMKxfvx4yWclH0hd5ISIiIiIyN6vMPEv3odYn1F999ZWYUBMRERERVSerTKrt7OwQHx+PkJAQDB8+HBs3bmRCTUREREQWY5WZqFarxXvvvYfRo0dj3bp1sLOzs3RIRGQmEokEDRs2BFDSrUs/TWSrJBIJAvz9gcRES4dCRFVglUm1nZ0dPvzwQ7i5ubHvNJGNc3JyQlxcnDivn46KirJMQERm5uTkhOjoaMDZ2dKhEFEVWGVSDQAeHh6WDoGIiIiICICV9qkmIiIiIqpJmFQTUY2m0WjQqVMndOrUCRkZGeJ0fn6+pUMjMguNRoMePXpYOgwiqiKr7f5BRLWDTqdDZGQkgJLx6fXTgiBYMiwis9HpdIg6d87SYRBRFbGlmoiIiIjISEyqiYiIiIiMxKSaiIiIiMhITKqJiIiIiIzEpJqIiIiIyEgc/YOIajwvL69yp4lslZenJ5CebukwiKgKmFQTUY2mVCqRmpoqzuunWaacbJVSqUR8fDzLlBNZGXb/ICIiIiIyEpNqIiIiIqqRrl27htDQUDg7O8PLywtTp06FSqUyy7bFxcVo3bo1JBIJfvjhhyrHyqSaiGo0jUaD3r17o3fv3sjIyBCnWaacbJVGo8HAgQMtHQaRxWVnZ6Nv377IyMjAtm3bsHr1amzfvh1jx441y7arV6826G5YVexTTUQ1mk6nw7FjxwCUtCLop1mmnGyVTqfDiZMnLR0GkVnFxcUhKCgIN2/eRGBgYLnrrF+/HqmpqYiMjISPjw8AwNHREc888wzOnj2LDh06VLj/qm5769YtLF68GJ9++inGjRv3SJ+JLdVEREREVOPs3bsXffv2FZNiAAgLC4OzszN2795t0m1nzpyJsLAw9OzZ85HjZVJNREREZGZFRUU4dOgQ5s6di06dOsHd3R1yuRy+vr4ICwvDnj17Ktx2/PjxkEgkD3xVtUtcbGws1q5di/Hjx6NVq1aQyWSQSCRYsmRJpbZPTEzE9OnT0bhxYygUCnh5eWHAgAEP/BxVdfnyZQQHBxssk8lkeOyxxxAdHW2ybf/44w/s378fq1atMipedv8gIiIiMrNjx46hf//+AABfX190794dSqUSly9fxq5du7Br1y5MnjwZ69atg0QiKXcf3bp1Q5MmTcp9z87OrkrxfPHFF1i9enXVPsQ9Z86cwcCBA5GRkQE/Pz8MGjQI6enpOHLkCPbv34933nkHixcvNthGEARotVpxXj+t1WpRXFxs8Dn0nz8zMxPu7u5lju/h4YGMjIwHxljZbfPz8zF9+nQsWrQIfn5+iIuLe9jHr1CtTqqLioogl8stHQYRERHZOKlUimeeeQavvfYaevToYfDejz/+iDFjxmDDhg3o1q1bhQ/TvfTSSxg/frxJ4mnZsiXmzJmDdu3aoX379vjggw/w7bffPnS7/Px8PPPMM8jIyMCzzz6Lb775Bo6OjgBKku1BgwbhvffeQ/fu3cUfEQCwefNmTJgwocz+7v+R8M0335jsM1bGBx98AHt7e8yYMcPofdXapPrSpUtYuXIl3nrrLTRt2rTCX4UPUlBQgIKCAnE+JyfHlCESERGRjejbty/69u1b7nvPPvssDhw4gI0bN2LLli2VGt3CWC+99JLBvFRauR7BO3bsQGJiItzd3bFu3ToxoQaATp064Z133sFrr72G9957zyCpHjp0KM6cOSPOJycnIywsDDt37oSfn5+4PCgoSJz28PBAVlZWmRgyMzPRtGnTB8ZZmW3j4+OxcuVKfPfdd+JQe/pcTq1WIzs7G25ubg88Tmm1sk/1v//+ix49ekAmk0EqlT5SQg0Ay5Ytg5ubm/jy9/c3caREBABOTk5wcnIqM01kq5xKJSoPcuvWLbz66qto2rQpHBwc4Obmhm7dumH9+vUGt9r19P1vAWDr1q3o3LkznJ2d4e3tjdGjRyMhIQFAya36Tz/9FG3btoVSqYSXlxfGjx+PlJSUCmPRaDT46KOP0KVLF7i7u8PBwQHNmjXDG2+8gfQKSq7/+++/eOaZZ+Dl5QUnJye0atUKn3zyCXQ6HQIDAyGRSMq9HX/69Gm88cYb6Ny5M3x9fWFvb4+6deti6NChOHjwYKX+djVNu3btAJT0Va7J9Ilxhw4dyu1e8eSTTwIAIiIicOfOHXG5p6cnOnbsKL5atWoFAGjVqpXBck9PT3Gb4ODgMv2ftVotrly5Uqa/9P0qs+3NmzdRUFCA8PBweHh4wMPDA23atAEATJo0CfXr16/Mn0RU61qqs7KyMGnSJIwZMwZr164FAKSlpaGgoAAeHh5V+rJesGABZs2aJc7n5OSUSaxVKhXi4uJw69YtZGVlIT8/Hw4ODnBwcICHhweaN29u8GQqERlSKpUGg/Xrp1mmnGyVUqksGSv3IWXKS/drDQgIwPDhw5GdnY2jR4/i1KlT2LFjB3bu3Al7e/sy2y5YsAAffvghevbsiUGDBuH06dP44YcfEBERgX/++QevvPIKdu7cid69e6NRo0aIiIjA5s2bce7cOZw5c6bMPpOSkjBw4EBcvHgRderUQadOneDi4oKoqCisWrUK27Ztw9GjR9GwYUNxm2PHjmHQoEHQaDRo3Lgx+vfvj/T0dMybNw9//fXXAz/7m2++iSNHjiAkJAQdOnSAUqnE9evXsXv3buzevRuffPIJXnvttSr81S3v6tWrAGDQanu/I0eO4OLFi8jNzYWnpyc6d+6M0NBQKBSK6goTeXl5AGCQ/Jbm5eUFoOSHWVRUFEJDQx/5WKGhoVi8eDFSU1Ph7e0NANi1axfy8vIwePBgo7dt27Ytjhw5YrDdnTt3MHr0aCxcuNCgpb1ShFrm7t27Qrdu3YRbt24JxcXFwvDhw4UuXboILi4uwujRo4V9+/Y98r6zs7MFAEJ2drYgCIJw+vRpYc6cOULfvn2FgIAAwd7eXgAgABA8PDyEjh07CkuXLhVu3Lgh5OXlmeojElWvvDxBAEpeDziP778+HrbesWPHHrje2bNnBQDC2bNnHylsopqkzPlcwXV17NgxAYCQkpIiNGzYUAAgvPLKK0JhYaG4zvXr14XAwEABgPDmm28aHEf/HeTp6SmcP39eXK5Wq4Xu3bsLAIRWrVoJjRs3FuLi4sT3U1NThSZNmggAhK1btxrsU6fTCd26dRMACJMmTRJycnLE94qKioTZs2cLAIQ+ffoYHK9+/foCAGH27NmCVqsV37t06ZJQt25dMdabN2+W+Xvt3btXSEpKKrP81KlTgqurqyCXy4Vbt25V9OcuY9y4ceLxqvI6cuRIpY/xIMnJyYKbm5sAQFizZk2V4vPz8xN+//13o2PQH+P9999/4HpvvfWWAEDo2LFjue//+eefYmyffvpphfu5efNmhf9/9TIzM4X69esLTzzxhLB3717hu+++E+rWrSsMGzbMYL3NmzcLdnZ2wtGjR6u8bUVxff/99w9crzy1rqU6JSUF0dHRyMjIwIIFC6BWq7F06VJcvXoVBw8exNtvvw1XV1d06dLFqOOoVCrs2LED3377Le7evVvm/czMTERGRiItLU28vaZUKo06JhER1Q6//vor4uPjUa9ePXzyyScGD903atQIH374IcLDw7F27VosXLgQDg4OBtu/99574m1uoKQoxqxZs3Dy5ElcvHgRe/bsMWhV9vLywv/93/9h9uzZOHToEMaMGSO+t2/fPkRERKBt27ZYt24dZLL/pRYymQwrV67Evn37cOTIEfz7779o2bIlfv75Z9y+fRsNGzbEsmXLDPrztmjRAgsXLsT06dMr/PyDBg0qd/kTTzyBadOmYdmyZfjtt98wderUSvw1ge7du1dqvfv5+vo+0nalFRcX44UXXkB2djZatWqFKVOmlFmnTZs2WL16Nfr164eAgABoNBr8888/ePfdd3Hq1CmEhYVh//796N27t9HxPEzfvn2xdOlSnD17FufOnRO7reitW7dOnDb2WTN3d3ccPnwYM2bMQHh4OBwcHDBy5Eh8+OGHBuvpdDpotVqDomCV3dakqpyGW7n09HShV69ewsqVK4Xhw4cLUVFR4nt//fWX0KtXL+GTTz55pH3rW9iuXbsmrFmzRqhTp06lfukOGDBA+PXXX9laTdbJzC3VGo1GCA0NFUJDQ4XMzExx+tSpU2ypJptRuqVao9EII5566oEt1S+++KIAQJg1a1a5+9PpdIKHh4cAQDh58qS4XP+9U17r4MWLFwUAgkwmE4qKisq8v3PnTvE7q7Tp06c/tIVz6tSpAgBh/fr1giAIwqRJkwQAwpw5c8pdPzMz84GxCoIgpKWlCZs3bxbmzp0rvPTSS8K4ceOEcePGCV27dhUACDNmzKgwnppE/7fw9PQUYmNjq7StTqcThg0bJgAQ2rRpY1QclW2pFgRB6NmzpwBAaNCggbBz504hKytLuH79ujB79mxBIpEIcrlcACAsX77cqJisTa1rqdb39Zo3bx5kMhnmz58vvvf444/D09MTR48eNaov1vbt2/Hhhx8+dAxFvaioKFy4cAFPPPEEW6uJ7qPVarF3714AQGFhoTj95ptvWjIsIrPRarXYt3//A9dJTk4GYDhSQmkSiQRBQUHIzMzE7du3y7wfEBBQZpnzvT7cfn5+Bq3Nei4uLgBQpsjIjRs3AAALFy7EwoULHxh3amoqgJIHLAFUWJ7a3d0dbm5uyM7OLvf9L7/8Eq+//rrB8xb3s4YRuV577TVs3LgRHh4eOHDgAB577LEqbS+RSLB48WL89ttv+Oeff5CYmFgtgyZs27YNTz/9NCIiIhAWFmbw3syZM3Hy5ElERkaiTp06Zo+lJqlVSbVOp4NUKsWqVauQm5uLDRs24JdffkFQUJD4sKCbmxvq1atn1HHmzZtXpfWzsrJgZ2fHhJqIiKrFg4ZPq+zQano6nQ5ASReKxo0bP3DdkJAQg/kHjb5V0Xtnz57FlClTYGdnhxUrVmDo0KEICAiAk5MTJBIJNmzYgClTphh0BXiYr776CidPnqz0+nrz589H8+bNq7wdAMyePRtr1qyBu7s79u/fX6YbRWWVHgXj1q1b1ZJU+/j44MSJEzh48CAOHz6M9PR01K1bF8OGDUPHjh3FPEo/wocxrl27hhkzZuD48eNwcHDAqFGjsGrVqkrlTFXdtri4GO3bt8fFixfx/fff47nnnqtSrLUqqZZKpWJivXbtWhQUFGDt2rW4desWgoODcefOHezYsQMRERHVGpeLiwsaN27MpJqIiCpFP0KEvpW4PDdv3gSAKg8LVlX6JG7YsGGYM2dOpbbRx1RR9brs7OxyxxgGSlpJBUHAq6++ijfeeKPM+/pRNKri5MmT2Lx5c5W3Gz9+/CMl1W+88Qb+85//wM3NDfv370fHjh2rvA+90sMV6u8mVAeJRIL+/fuXGSHj+vXrSE5OhqenJ9q3b2/UMbKzs9G3b1/Uq1cP27ZtQ0ZGBmbNmoW7d+9i+/btJt929erV4t2UR2Gz41SXNz4n8L9f4HK5HN988w2WL18OQRDwyy+/4O7duzh27BhatGhh1LH79u2Lp556Ct26davU+j179kSjRo2MOiYREdUe+op8P/74Y5nuGEBJgY7MzEy4uLigQ4cOZo1F/9CgPtmtjJ49e4rblC5Rrfff//63wm31XStLP0ipl5+f/9BkqzybNm2CIAhVfj3Kg4Hz58/HqlWr4ObmhgMHDqBTp05V3kdpP/zwAwDA1dUVzZo1M2pfpqB/EHDy5MnlDudYFevXr0dqaip27tyJQYMGYcyYMfjiiy/wyy+/4OzZsybd9tatW1i8eDFWrFjxyPHaXFKtH2nDzs6uwsRap9OJF/6MGTOwZcsWnDp1Clu3bkXr1q2NjsHV1RVubm6Vann29/fH66+/bnQiT0REtcfw4cMREBCApKQkzJo1yyAxvXnzJmbPng0AePXVV8uM/GFqw4YNQ6dOnXD69GlMmDCh3Ja+zMxMrFu3Toxz5MiR8PPzQ1xcHN566y2xCwkAxMTE4L333qvwePruDps3b0Zubq64PD8/H1OnThVb6Guit99+GytWrIC7u3ulE+rz589j586dZX586HQ6bNy4UXy+ZMaMGQajwAAl45E3b94cCxYsMN2HAHD58uUyfdaLi4vxwQcfYP369WjSpAneeusto4+zd+9e9O3b16CeR1hYGJydnbF7926Tbjtz5kyEhYWJP/gehU11/4iJiUHr1q0xcOBA7Ny5U0ys7ezsDNbTt1anpaXBy8sLMpms3IcyzM3NzQ1z584VB64nIiKqDIVCgZ9//hkDBw7EF198gb1796JLly7Izc3F4cOHkZ+fjwEDBmDRokVmj0UqleLXX3/F4MGDsXnzZvz8889o06YNAgICUFhYiBs3buDixYvQarUYP348ZDIZnJycsHXrVgwePBgrV67EL7/8go4dOyIjIwNHjx7FsGHD8PfffyMhIaFMa+eECROwevVqnDt3DkFBQejRowfs7Oxw4sQJaDQavPbaa1i9erXZP3dV7dy5E0uXLgUANGnSBJ999lm563l5eRkM+xYXF4cRI0bAw8MD7du3R926dZGVlYV///1XrIA5evTocv9fJycnIzY2VnywtbSoqCiDIQevX78OoKSFt3TSuWPHjjIFaTZs2ID169ejQ4cOqF+/PgoKCvDXX3/h7t27aNKkCQ4cOGCSvOby5ctlSrbLZDI89thjZaolGrPtH3/8gf379yM2NhYFBQWPHK/NJNXJycmYNGkSOnTogEuXLuGZZ57B9u3bK0ysZ8+ejYSEBHz44Yfl3kJ6FPrW76KiIgAo95YWADg4OCAoKAjjx49HeHg4tFqtVTylTFSu0k/f5+QAFdwh0p/jD7s9rH9fpVIhJyfH4On+0q1S+uVnz54VK3wRWavY2FgAJdXqcnJyIAAQvxVKXVf6814QBHTq1Annz5/HihUr8Pvvv2PHjh1QKBRo164dxo4di5deeqnaGozq1auHv/76C5s2bcKPP/6ICxcu4PTp06hTpw7q1auHV155BWFhYQat5n379sXff/+Nd999F8eOHcOvv/6KRo0aYenSpZgxYwZcXFwglUrLjCDh7u6OyMhILFq0CPv27cPvv/8OT09PPPXUU1i0aNEjPXBYHUqPCBYZGYnIyMhy12vYsKFBUt2mTRvMnDkTkZGRiImJQUREBARBQN26dREeHo4JEyY8UtXCnJwc/P3332WW37p1SxydBUC5SWZoaCji4uIQFRWFyMhIKBQKNGvWDLNnz8b06dPh6OhY5XjKk5mZWW4pdA8Pj4eOsFbZbfPz8zF9+nQsWrRIvHvyyKpx+D6z+u9//ys8/fTTwtGjR4Wff/5ZCAwMFJ5++mnx/fvH3Pz555+FOnXqCLdv3zZZDImJiY9UkYkvvmrTKzExkdcRX3wZ+XrYdWTt9ONxt2rVytKhkIl88803lTq3t23bJm4jk8nKHTe7X79+Qv/+/R94vMpuu3DhQiE4OFisSsqKiijp06VUKtGrVy8UFxdDp9PhjTfeEFusZTIZtFotpFIpBEHAM888gwEDBojjcppCvXr1kJiYCBcXlwcOE2RuOTk58Pf3R2JiIlxdXS0Wx6Ow5tgB647f3LELgoDc3NyHDllZ1evI2v7mjNe8bD3eyl5H1iA1NRV5eXllxtr+999/8fLLLwMo6epBtmHEiBGVqlZderQaDw+PckeByczMRNOmTR+4n8psGx8fj5UrV+K7774T7wLp76qq1WpkZ2fDzc3toTHr2URSrdPp4OTkJA5ALpPJMHToUEgkEsydO9egK8iXX36Jnj17olmzZibvxyyVStGgQQOT7tMYrq6uVvGlUh5rjh2w7vjNGXtl/nF61OvI2v7mjNe8bDneqnzJ12SXLl1Cnz590KJFCzRq1AiOjo64efMmoqKioNPp0L9/f7z66quWDpNMxM3NrcrnbnBwcJn+z1qtFleuXClTdOZRtr158yYKCgoQHh5eZvtJkyZhxowZVepeaBNJdXkD1Ts4OGDw4MGQSCSYM2cOwsPD0aBBA6xZs0bsjG/J1mQiIqLa7LHHHsO0adNw7NgxREREIDc3Fy4uLujatSuef/55vPzyyxYZRIBqjtDQUCxevBipqanw9vYGAOzatQt5eXkYPHiw0du2bdsWR44cMdjuzp07GD16NBYuXFhmDO6HsZmztbi42ODiEwQBjo6OGDx4MIqLizF69Gh4eHggMjKywrKuREREVD3q1auHTz/91NJhUA02ZcoUrF27FsOGDcPChQuRmZmJWbNmiZUb9bZs2YKJEyfi0KFD6NWrV6W3dXd3LzPWuP5BxRYtWojjwVeWTYxTrdVqIZPJEBcXJ1ZE0rdCOzg44ODBg3BycsKJEyeMru5jDRQKBRYtWgSFQmHpUKrMmmMHrDt+a43d2uJmvObFeIlsh7u7Ow4fPgxXV1eEh4fj1VdfxfDhw7F161aD9XQ6HbRarcHoUpXd1pQkglDJ8kc1lL6FOi4uDs2aNcNzzz1nUGp07969mDlzJr7//nuzV5UiIiIiotrJqpPq0gl1+/btMWLECKxfv96gG4hGo0F2djZ8fX0tGCkRERER2TKrTarvT6jDwsLw1VdfGSTUOp2u3IcYiYiIiIhMySqTan2FxAcl1Jag0+mQlJRk8XGqiWqi0uPrPujHLq8joorxOiIyXmWvo6qyytE/7OzsEB8fj5CQEDz77LP48ssvy5Qht4SkpCT4+/tbOgyiGi0xMfGB41DzOiJ6OF5HRMZ72HVUVVbbUj158mRIJBKsW7fO4i3UetnZ2XB3d0dsbCz7cFPtoVIB+upuSUlABUWV9JXjsrKyHlgAQH8dla4wp1KpoFar4eTkZPKiTUQ1UgXXlTHXUW2h1Wpx6tQpAEDXrl1N3uhmlv1X8P/b3J+ltqrsdVRVVplUAyVlJt3c3GpUn+mcnBy4ubkhOzu71v0jRrWYSgU4O5dM5+U9MKmuzPXB64gIFV5XvI5sVCX/HSXTMNf1UTOaeB+Bh4eHpUMgIiIiIgJgxUk1ERER0f2KioqwYcMGAMDkyZMhl8utav+WOhYZz2q7f9REvN1GtRK7fxCZHrt/PDKVSgXne3+7vLw8kz+HYZb9V/D/29yfpbYy1/VRczokExERERFZKSbVRERERERGYlJNRERERGQkJtVEREREREZiUk1EREREZCQm1URERERERuI41WaQmpqKxMREeHt7w8fHx9LhEFkllUqF9PR0AICPjw+HkiKiSlEoFNi9e7c4bW37t9SxyHhMqs0gLS0NcrkcqampTKqJHpFarYZKpQIAKJVKJtVEVCkymQyDBw+22v1b6lhkPCbVZuDl5YX8/Hx4e3tbOhQiq+Xk5ASptKSHGhNqIiKq6ZhUm4G3t3etq2BFZGpKpRJ+fn6WDoOIrExRURG+++47AMCYMWPMUqbcnPu31LHIeEyqzSA1NRV2dnZsXSMyEZVKBZVKxW4gRPRQhYWFmDBhAgBg5MiRJk9Ezb1/Sx2LjMek2gyKi4vFBICIjKdSqXhdEVG1SkhIQFpaWpnlGo1GnD5//jz8/f0REBBQnaFRDcWk2gxkMhm/+IlMSKlUMqEmomqTkJCA4OBgqNXqB67XvXt3ODk5ITo6mok1Mak2B29vb375E5kQu30QUXVKS0uDWq3G1q1bERwcbPCeRqNB9+7dAQBfffUVXnrpJaSlpTGpJibV5qBSqZCfn89EgMgE2J+aiCwlODgY7du3N1imH+oTAJo3b17dIVENxqT6Hp1OJw7fZayEhASo1Wo4OzujYcOGTAaIjJCSkoLc3FzY2dnByckJAIvBEBFRzVMrk+rr169jy5YtSE1Nha+vL95++21IpVIIggCJRGL0/tVqNbRaLTIyMlC/fn32BSWbY9B6XE3H1F9XALuDkG1TqVRQpaSApcOIrEutS6ovXryIfv36oU+fPkhNTUVERATOnz+PX375pcoJdUFBAQoKCsT5nJwcACVFK+Li4uDu7o6CggK4ubmZ9DMQWZp+NI6UlBQoAaO+/Cu6jvSUSiXS0tKg0Wig0Wjg6enJhJpsmkqlEn9AAiV3a5S8O1NpCoUCP/30kzhtzv2be4g7c38WMq1alVTfunULzz33HCZOnIjly5ejoKAAu3btwtKlS3H58mW0aNGiSvtbtmwZFi9eXGZ5RkYGnJycYG9vzxY1skn60TgKCgoMvvwfRUXXEVCSXKSmpkKn00Gr1cLT05MPApPN019felqtlnc8q0Amk2HkyJHVsv+oqCizHef+Y1HNV6uS6kOHDsHX1xczZsyAIAhQKBTo3bs3pk2bhps3b1Y5qV6wYAFmzZolzufk5MDf3x8qlQpXrlyBl5cX7OzsAJT0ASWyFfofiyqVCqlxcUbtq6LrCPhfi11cXBzy8/ORkpKCrKwsALymyHYplUooS53fBQUFEO51uaKaKTo6+qHreHl5cYQQG1erkupmzZrh2WefRb169QCUFGnx8PCAh4eHwe1nPa1WKybF5VEoFOXejikqKoKnpydSUlLg5+eH1NRUJgBksx42juvDVHQdASXJhZ2dHVxcXFBUVITk5GQolUpeU2TT7u9T7eTkhGKFAiqVCg4ODhaNzRoUFxdjx44dAIARI0ZAJjNtqlN6/x06dICTkxNeeOGFh273KONZm/uzkGnVqv87HTp0QJcuXQCUjPahPzldXV2Rn58vrvf1119j/PjxD0yoHyQkJATR0dFwcXGBvb09vL29jQ+eqAZSqVRm7eenVCoRGBgIoOTHqpOTE+RyOa8psmn396lWKpVQ3fuvsd2taoOCggKMGjUKAJCXl2fyRPT+/UdHR5dbebG06OhovPDCC1Uez9rcn4VMy6b/71y/fh0//PADrly5gtDQUISFhYkPFZQe7SM/Px+CIAAAFi1ahPfffx/du3fHY4899kjHDQwMhK+vL1JTU+Ht7c0WNbJZSqUScHEx+3G8vb3FRJrPKZCtu79Pdelz/v4HecnyAgIC2K2DAACmGZi5Brp48SJ69OiBU6dO4c6dOxg9ejR2795tsE5xcTGKiorEETo+/vhjrFq1CpGRkY+cUOvpWxrYB45smVKpNPuPxpSUFNy5cwcqlYrjU1OtUB3XFRGZnk22VF+/fh2DBg3CxIkTsWjRIsjlcowZMwb//PMPwsPDxaHz9K3W9erVw8yZM5GUlITjx4+XqZ5UVSqVComJicjLy0PDhg2N/jxEtZ1arTZ4QJGJNRER1TQ2l1QXFhZi48aNGDZsGN58800xcZZIJIiJicHAgQPRt29f9O7dG48//jgEQUBeXh4SEhJw9uxZtGrVyugYEhISEBUVheTkZLRo0YJdQIgekUqlQlpaGuLj4yGXy5Geng6pVIomTZogMDCQyTURVVlCQkKl+kATVZXNJdX29vZ49tlnoVKpxJLG7733Hn766Se88sorkMlk+Oqrr3DhwgV8/vnncHV1xcKFC9GyZUs0atTIJDGo1WpkZ2cjPj4eubm58PX1xYABA0yyb6KaJsWMld/041QXFhYiOzsbarVaHLMaABNrslnmvK5qs4SEBAQHB1dq1CInJyd4eXlVQ1RkK2wuqQaANm3aiNO3b9/GpUuX8OuvvyI0NBQA0KtXL4wYMQJz585F27ZtMXToUJOUJ9cLCAiAn58frly5ArlcbjCyCJGt0JcqT0xMNMuXv0qlQn5+Puzs7ODp6YmioiKkp6ejsLAQdnZ2UNwbYoxJNdkafa0DJtWml5aWBrVaja1btyI4OPiB63Jcaaoqm0yqS/Pz88OXX34JV1dXcdxpLy8vtGrVCh4eHgBg0oQaKBmpoHnz5rhz5w7S09PF4xDZEn2pckdHR7PsX61Ww9HREQ0aNABQUhFVIpHAw8MDrVu35iggZLNUKhXc3NwsHYbVsre3xzfffCNOlyc4OPiRn5+qzP5NpTqPRcaziaRaPzReeaRSKVzuDfmlH3d6586d8PDwMOs/WhqNBvHx8VAqlcjMzDTbcYiqm6pUZTeZTGa226NOTk6ws7ODk5MT7t69i7Nnz6K4uFhMsvmcAtkqpVKJunXrivMpKSnwCQqyYETWRS6XY/z48Va7f0sdi4xn1UPq6UcDeFhLs/79xMRELFy4EOvXr8fatWvh7u5ulrhUKhWysrKgVCrLrdRIZM30LdR6V65cMctx9MOK+fj4ID4+HoIgICsrCxKJxOgqjkQ12f13YRITEy0YDRFVltW2VJ8/fx6vvvoqPvvsM7Ru3fqh61+4cAHz58/HjRs3cPToUZOM8lERtVqNhg0bIiUlBZ6enmjSpInZjkVU3UoXpoiJiUFxYaHZj9eoUSNotVo0aNAAzZo1Ex9CJrJVKpUK+rRao9Hw+YEqKC4uxr59+wAAAwYMMEuZcnPu31LHIuNZ5f+df/75B507d8bMmTPLJNQ6nQ5SqVTsP63XunVrzJ49G02aNDH72NEqlQrx8fFwdXVF06ZN4e3tjZSUFPYBJZugP4fj4uJgb28PqZn6+ekfVFQqlfDy8kJGRgaAkmEz1Wo1kwyyWSqVyuBujJ+fH8/3KigoKMCQIUMAmK9MuTn3b6ljkfGsrvvHpUuX8MQTT2DBggVYuXIlBEFARkYGbt68CaCkDzXwv/7TX3zxBSIjIwEA/fr1q5ZiLOnp6bh9+zZSUlKQlJQk3i5ndUWyFSkpKcjNzUVhYSHq1KljlmOo1Wrk5OQgLi4OaWlpKCwsRFJSEvLy8litlGya/vrSKygoMHiWgYhqJqtKqtPT0zF8+HA0b94cixcvBgBMmjQJTz31FLp164ZevXrh/PnzEAQBQMk/TNOmTcOSJUuqtW+zk5MTiouL8ddff+HYsWO4cuUKZDIZWxnIpui7YFy9etVs+8/MzERKSgpu3LiB2NhYpKSk4NatW8jKymKSQTZLrVaLd2YA4MaNG8jNzeX5TlTDWVVS7enpiYEDB0KpVOLdd99F586dkZycjClTpuDzzz9HUVERhg8fjuvXrwMoGR0gJiYGK1asgEKhqLY4AwIC4OjoCKlUivT0dGRlZbG0MtkUHx8f+Pr6wtHREbdu3TLLMZRKpZhYq9VquLu7o6ioSByiUj9ONZGtk0ql4sPvRFRzWU1SrdPpAABr165F586dsW7dOvj4+GDTpk14+eWXMXz4cJw6dQrOzs5YsmQJgJIO/o899hiaNWtWrbEqlUq4urrCzs4O+fn5cHBw4Jc/2RT98wEajcasY6c6OTmhXr16qFOnDjQaDQAgMzMTjo6OHKqSbJaTk5NBtyqdTofmzZszqSaq4Wp8j3eVSgWdTgdBEODq6goA+Oijj1CvXj0EBQWJY9XqH0xs3ry5wRi6luLr6wtfX1/xgRM+ZEK2RqVSIS8vD4VmHP1Df31nZ2dDo9FAp9PBy8sLTk5OLKpENkupVCKj1HXl4ODA7w8iK1CjW6ovX76Mp59+Gr169UJwcDC+++47aLVaAMDs2bMxZMgQcQxqOzs7sQhMixYtAEDsW13d9ONUx8fHIzk5GbGxsRaJg8gcVCoVUlJSkJqaivPnzyMhIcEsx0lNTYVKpcKFCxewe/duxMTEQCKRwMvLC97e3nxOgWySfvSoa9euicvkcjnvdhJZgRrbUn358mX07NkTY8eORceOHXH27FlMmDABISEhaNu2LQDDkp3FxcVYvHgxIiIisGzZMgCmLz9eWWq1GpmZmcjKykJWVhaCgoL4DyLZDP1oNteuXcPdu3eRdeeOWY5TXFyM+Ph4XLhwAbdv30ZRURF8fX3FbiCsqEi2KCUlBYmJiYiLixOXnT9/Hl4NG/KcryR7e3t8+umn4rS17d9SxyLj1cikOiMjA6+//jrGjBmD//znPwCA559/HlFRUfj666+xZs0ag9LkBw4cwNq1a3HmzBns3bvX4sVWnJycIJVKUVRUBEEQcOfOHaSlpSGIZWbJBpQu/uLs7IyMe887mJpMJoOTkxO8vb1RVFQEtVqNpKQksZWcCQbZKkdHRwQGBorzcrlcbLl2cHCwUFTWQy6XY9q0aVa7f0sdi4xXI5PqoqIiZGVlITw8HMD/CroEBQWJwwzpE2pBEBAUFIQWLVpg5cqVaN68ucXi1lMqlahbty58fX2RkJBg8JAVkbXTP6TYpEkTpKenI7VUi5opeXt7Iz8/H35+fvD29sbdu3dRUFAAe3t7eHt7m+WYRJam/7Go8fISl+l/xBYUFCAvL88icRHRw9XIpLpu3brYunUrmjZtCqDkIUSpVIr69esjPj7eYF2NRoMmTZpg6dKlBhUULS0/Px8FBQWQSCSQSCRwdHS0dEhEJuXt7Y2MjAwUFRWJy8zxQK5Wq0VWVhYyMjLg6+uLRo0asZWabJb+4fb8Ug0x+lGkFAoF3NzcLBidddBqtThx4gQAoEePHibPDcy9f0sdi4xXI5NqAGJCrdPpIJfLAZS0SqekpIjrLFu2DPb29njttddqZOlOjUYj9q8mskUajcagSIVKpYKpHx1MS0uDRCKBTqcTu4BwNB2yRSqVCjExMSgsLMTdGzfQ4d5yfZePkJAQ5OTkWC5AK5Gfn48+ffoAKCntbep/K8y9f0sdi4xX8zLR+0ilUoP+0/oy5O+88w6WLFmCc+fO1ciEGigZvSA1NRVpaWns/kE2R6VSITs7WxxDHihJgJUmKnR09uxZJCQkICUlRTyOTqeDm5sbk2qySSkpKdBqtbhx4waKsrLE5YWFhahTp474PhHVTDUzG72PPqmWyWTw9/fHhx9+iJUrVyIyMhJt2rSxdHjlSkxMRHp6OvLy8nDjxg0kJydbOiQik1Kr1ahXrx6u/fOPuCwvL89kCW90dDT+/fdfpKenIyMjAzKZDAUFBYiPj4dUKoVKpWKlUrI5np6eyMnJQWapblX6W/7FxcVsoCGqwawiqda3Tsvlcnz55ZdwdXXFyZMn0b59ewtHVrGMjIySfnH5+SgqKkJiYqKlQyIyKW9vb9SvXx8BAQHiMjs7O5MluXl5edBoNEhKSkJhYSFyc3Ph6emJ+Ph4yOVyuLu7iw9NEtkC/Y9EFxcXnC3VUu3s7AwnJydxRBwiqplqdPGX+w0YMAAAcOrUKXTs2NHC0TxYnTp14OjoCJlMBq1WC4VCYemQiEzKx8cHDg4OSE1NFZf5+/ubLMn19/dHu3btoFQqxQd+1Wo1bt++LSYeTKjJliiVSvj4+MDHx8eg77RWq4WTkxPvzBDVcFaVVHfs2BG5ublixcSazN/fH3Z2dmK3FSbVZGtUKhXOnTuHtLQ0cZkp78iEhISgWbNmcHd3h1arRU5ODlQqFRQKBVq3bo2goCAmGGSTUlJScObMGXFeKpVCrVZbMCIiqgyrSqoBWM2XqIODAwoLC5Gfn4+0tDRcuXKFVRXJpqSkpCAnJ8cgqTblswPe3t5wdHSERCJBXl4e8vLykJCQgOvXr+PKlSsGIwER2QqVSoXTp0/j5s2b4rKYmBj8+++//A4hquGsok+1NcrPz0fWvT5xBQUFyMzM5IgFZHM8PDzg6uoqzpuy2ptarca5c+eQmJgoPpuQl5eH27dviyPrcLxqsjUpKSm4cuUKXFxcxGXe3t5ITk6GSqViRcVKkMvlWLlypThtbfu31LHIeEyqzcjBwQEFBQUQBAFZWVlMqMnmaLVa3L17V5xPSEgw2b6dnJyQmZmJ4uJiCIIAqVQKuVwOrVaLzMxMPrBFNikxMREpKSm4deuWuOzChQtorVQiNTUVderUsWB01sHe3h5z586tkfuPjo4ud7lUo0Hbe9OJiYnwv1cd2tyfhUyLSbWZtGzZEk2aNMHZs2cBlNy+Y0s12RK1Wo3U1FTk5uaKyyr6wngUSqUS7dq1w/79+6HT6aBQKODg4IAmTZrA09MTarWa1xTZDJVKBZVKheTkZOTm5ho8ACyXy5Gfn4/c3Fy2VFspLy8vODk54YUXXij3fScA+s497du3x9mYGIORlcg6MKk2E29vbzRs2BAXL15EUVERVCoV4uPjebuabIaTkxNu375t8ACVXC43WaKrUqng4eEBZ2dncYxqZ2dnyOVyFBUVQaFQMKkmm6FSqZCTkyN2/yguLhbfu3v3Lp548knY2dnxDk0laLVaREVFAShJUM1Rpryq+w8ICEB0dLTBMyilSTUaoHt3AIBao0FaWhoCAgLM/lnItJhUAwYVG00lJiYGly9fRmFhIYD/jTNKZCuUSiWkUqk4jjxQ0uUpLi4OISEhRu8/NTUVly5dQnp6OtRqNQRBQG5uLgoKCuDu7g5XV1cm1GQzlPe6d+h0OkgkEtjb24vvubi4wM/PD15eXhaM0Hrk5+ejc+fOAMxXpvxR9h8QEFBx63MFD6Ga+7OQadXapPrKlSvYsWMH5s2bZ/KEGihJqkvfvgPApJpshkqlQmpqKnJyciAIgri8oKDApEN/paenIzs7W+xX7eDgAI1GIz6jwC8YshVKpRLe3t5iNxB9gwwAXL58GZs3b8aBAwcMWrCJqGaplUn1hQsX0L9/fwwePBjR0dEIDg4GYLoW6379+iEtLQ3Z2dnisjt37uDKlSsICgoyev9ElqZSqfDPP/8gLS0NOp1OXF5YWIj69etj5MiR+Ouvvwy2Kb1eZXh7e6Np06YQBAFFRUXQarVIS0uDp6cnoqOjcfnyZXTq1Mkkn4eoppDJZNDpdAZJ9Z27d3Hj998tGBURVUatS6qTk5MxfPhwjBkzBv/5z38M3qtqQl1QUICCggJxXl8BKzIyssy6mZmZ+PPPP8WqkETWTKlUIikpCYBhsnz37l3s2LEDP//8c6X3VdF1pG/x9vf3x40bN8RlGRkZ0Gg0OHfuHBo2bMjnFMimpKSkoKCgwOAB4PHjxqHw3nBqhYWF2LJli6XCI6IHqHVJ9bVr19CoUSP85z//gVarxfTp03Hr1i0kJyfjtddeQ9++fVG/fv1K7WvZsmVYvHhxmeUrVqyAo6MjAGDHjh04cuQIdDodi1WQzVAqlejcuTN27twJmUwG3GtVu3r1Kn47eBAA8Oqrr6JHjx7iNmq1GuPHjy+zr4quoyZNmpR7bKlUiqKiInh4eHCsarI5MpkM+fn5Bt2qpk2bBuFe98G8vDwm1UQ1VK1Lqm/fvo07d+4gLy8PYWFhsLOzQ9++fXHx4kUsWbIEMTExmDNnDjw8PB66rwULFmDWrFnifE5ODvz9/dGrVy84OzsDAHx9fXHkyBHY2dnBzc3NbJ+LqLr5+/ujbdu2iC51ZyYhIQE5OTlo0qQJXnrpJYMn1fPy8srdT0XXUXkkEglcXV0RGBiIunXrwtvb20SfhqhmcHBwKEmqLR0IEVVZrUuqPT09UVBQgL/++gseHh74/PPPUbduXQDAypUrsWbNGowePbpSSbVCoYBCoXjgOi1atEDXrl1x6tSpcruFEFkrtVot9v/U03fdePHFFys99FNF19Hhw4fFH6e//fYbli5dCkEQYG9vD0dHRzRv3pyt1GRTVCoVXF1dkZmZCdbOI7I+tS6p7t+/P+rUqYOnn34a9erVM3jvjTfewPr167Fz5060bNnSZMd8+eWXcerUKRw7dgzJycnw8/Mz2b6JLMXJyamkKEWp/tA6QYC7uzsGDx5skv3rR8wJDw/Hpk2bxDtNMpkMqgqGoCKyZv/88w+KiorgW6cOkJFh6XCsklwux6JFi8Rpa9u/pY5FxpM+fBXrlZCQgK1bt2L58uXi4OkA8PHHHyM4OBgpKSm4efOmuLygoAANGzZEo0aNTBpHx44d0bZtWxQVFZV5OJLIWimVSuTn55dZPnLkSJNXfZPL5fi///s/ACVduOLi4ky6f6KaQKlU4s8//wQAhIaGWjga62Vvb493330X7777rsF439ayf0sdi4xns0n1xYsX0adPH3z22Wf4+uuv0aVLF/x+b0iidu3a4Z133oGvry+ef/55/PDDDzh48CCWLl2KmJgYcaB1U5FIJJg8eTIA4LPPPkMGWx/IRjRv3hxy2f9ueMns7PDcc8+Z5VhDhgxBYGAgdDodjh8/zjGqyeakpKSII90MHTrUwtFYn4SEBERFRT3wFR0dbekwyYbZZPePmzdvYsiQIXj++efx5ptvQi6X46233sLMmTPRqVMneHl5YeDAgWjevDlmzpyJt956CxKJBG5ubtizZ4/JW6oBoEePHmjWrBliY2Oxdu1a8XYOkbVSKpXo2LEj3N3dgXuld/v27Ss+o2BqMpkMU6dOxRtvvIFLly4hKyuLfarJJugLvnz33XfQarVo06YNaxpUUUJCAoKDgytVfMrJycmo6pQ6nU5MzoODgw2qyppadR6LjGdz/3eKioqwYcMGdO7cGQsXLoSLiwscHBwQGhqKoqIi8YS0s7ND48aNsWvXLuzbtw9Hjx7FgQMH0K5dO7PEJZFI8PLLLwMAVq9ebTAGKZG1SkpKMqjwZq5War2BAweiadOmyM3NxYoVK9ivmmyCSqVCUVERfvzxRwDAiBEjLByR9UlLS4NarcbWrVtx8uRJcfnJkydx9uxZg1d0dHTF5cIrQaPRoGXLlmjZsiU0Go0pwq8RxyLj2VxSLZfL0aJFCzRp0sSgLHi7du2g0WiQlJQErVYLAOI4oE2aNEGDBg1Qp04ds8bWv39/BAYGIjMzExs2bDDrsYjMTaVS4ezZswY/EFu1amXWY0qlUkybNg0A8MMPPyAiIoKJNVk9pVKJ8+fP4+bNm3B0dMTAgQMtHZLVCg4ORtu2bcX5tm3bon379gYvYxJqogexye4fL774YpllOp0OUqkUUqlUHOorKioKzZo1E4ftMpXp06eXFMS4p02bNpg+fTrs7OwwadIkLFy4EB999BGmTZtm8ge6iKqLUqmEIAgovvcjtTRBEMq0qlTmtmxp+/fvN7g+HnvsMTRp0gT9+vVDixYtcPnyZWzcuBFt27Zl/2qyakqlUqxC+tRTT5V8J5W6XjQaDfQDV1b1OiKi6mOTSbWeIAiQSCTQarUoLCyETCYTE+h58+bhq6++QmxsrMmT6vvHo/7rr7/w+OOPo3PnzhgyZAg+++wzJCcn47vvvsOkSZNMemyi6lLR6B9FRUUYM2YMLl++bNT+33nnHYN5e3t77N+/H15eXpg+fTqmTp2K3377DatWrTLqOESWptFo8NNPPwEAnn76aQAQ76gCQM9evcBUmqjms4nuH6XLuZYmkUgAlPSftre3R3FxMXQ6Hd555x189tln+P333416WKEiU6ZMwYwZMzBjxgxxTOq9e/cCKOmeMnz4cABARESEyY9NVJ2uXLlSZtnp06eNTqiBksJJrVq1EruUFBYW4uC9Eug9evRAQEAACgoKEBMTY/SxiCzp33//hVqtRp06ddC+fXsAZRtniKjms+qW6qysLLi7u4vJ84PI5XK4ublh+vTpOHDgAE6dOoUOHTqYJa4nnnhC7M/t6uqKJUuW4MiRI3jnnXcglUrF0UWuXr1qluMTVZdbt26VWab/Afnkk09i7Nix4nKNRoMpU6ZUet9vvPGGeB399ttv+O6773Dw4EE899xzkEgkaNq0KRISEvDPP//gqaeeMvKTEFnOuXPnAJQ836P/Pjt06BAm3Hv/qy+/RPG9qqNVvY6IqPpYbUv1+fPnMXToUFy4cOGh6+p0OqSnpyMmJgZHjx7F6dOnzZZQ369FixZwdHREenq6GGvDhg0BlN/KR2QtVCoVkpKSDJYVFhbi0KFDAEpakx0cHMRXeaXIK0s/dvzp06eRnZ0NAOKPU1O0ihNZ0sWLFwEAjRs3BlDSherIkSPi+wqFwiTXERGZl1W2VP/zzz/o3LkzZs6cidatWxu8p38gUavVig8k6luHP/jgAwwePNjsIxSUJpPJ0L59e0RERODw4cNo27atmFSnpKQgJycHrq6u1RYPkanExMTg7t27cCy1LCIiAnl5efD09ESzZs1Mdiw/Pz/4+/sjMTERx44dQ1hYmDiOr75YBpG1unbtGoD/JdVnzpxB1r0fj1R1crkcc+bMEaetbf+WOhYZ75GT6sjISPz0009ISEhAYWGhwXu//PKL0YFV5NKlS3jiiSewYMECLF68GIIgIDMzE9nZ2QgKCjIYhxoA1q1bh44dO6Jjx46YP3++2eJ6kE6dOiEiIgKHDh3C66+/DmdnZ9SpUwcZGRm4evVqtbWaE5lSbGwsBEGAq4sLcG9Yvf379wMAunbtavIiBZ07d0ZiYiIOHjxokFTzjg9Zu9jYWAD/S6r37dtnyXCsnr29vVkfYDb3/i11LDLeI33r/fDDD+jatSuio6OxY8cOFBUV4dKlSzh8+DDc3NxMHaMoPT0dw4cPR/PmzbF48WIAwKRJk/DUU0+hW7du6NWrF86fPy8+uJiSkoKpU6diyZIlKCgoMFtcD9O2bVvY2dkhPj5ebFULDAwEwH7VZL30Q3vp77wAwNGjRwEA3bp1M/nx9F1ATpw4AbVaLSbVd+7cQVZWlsmPR1QdVCoV4uLiAJQk1UVFRWIXKiKyLo+UVH/wwQf4+OOPsWvXLtjb22P16tWIiYnBqFGjzDqouqenJwYOHAilUol3330XnTt3RnJyMqZMmYLPP/8cRUVFGD58OK5fvw4A8PHxQUxMDFasWGHRfmhOTk5ilxN9Pzn934mtbGSt7ty5A8AwqS7WauHn52eWEsuBgYHw9vZGUVERIiIi4OzsLJYp17f0EVkb/R0fDw8P1KlTB6dPn0ZWVhZcTDzUa22i0+kQFxeHuLg46HS6h29Qw/ZvqWOR8R4pqb5+/ToGDx4MoOTWhEqlgkQiweuvv262SoH6k2nt2rXo3Lkz1q1bBx8fH2zatAkvv/wyhg8fjlOnTsHZ2RlLliwBABQXF+Oxxx4zad/OR9WpUycAwOHDhwH8LxFhSzVZK30iWzqpBkpaqSszIk9VSSQSsbVa35KnT945rB5ZK/2DtvquH/ouVPrvDKo6jUaDoKAgBAUFmaW0t7n3b6ljkfEeKan28PAQSxPXr18f//77L4CSIe5MXe1JpVIhNzcXeXl54rKPPvoIc+fOxcSJE8WWKv1A+c2bNxfLFpeuamhpHTt2BABcuHABKSkpTKrJ6umT6vvvTnXt2tVsx9Qn1cePH0dRUZE4AgiTarJWly5dAvC/rh/6sdj15zoRWY9HSqp79uyJAwcOAABGjhyJ1157DS+//DJGjx6Nfv36mSy4y5cv4+mnn0avXr0QHByM7777TkyeZ8+ejSFDhhgUeNFXUGzRogWAiovCWIKHhweaNm0KoKTfqT4RYVJN1kgQBDGp1j8fAAAB/v5o0KCB2Y7brFkzuLm5IScnB2fOnBFbqqOjo812TCJzKt1SfebMGWRnZ8PV1RXNmze3cGREVFWP1JT76aefiuWJ33rrLcjlcpw6dQrPPPMM3n77bZMEdvnyZfTs2RNjx45Fx44dcfbsWUyYMAEhISFo27YtgJKuJ3rFxcVYvHgxIiIisGzZMgAwyy1oY4SEhODq1auIiYnBgAEDAAAZGRkoLCw0+CxENZ1KpUJOTg4AiFVDAaBdu3YAgKSkJMTHxxtsY4qHhaVSKUJCQnDq1CnExMSILdX3j5dNZC0SExMBlNz11Y9X3bp1a3EEK6DkWZzCe8Op3T/aFhHVHI+UVNepU0eclkqlJh+qLiMjA6+//jrGjBmD//znPwCA559/HlFRUfj666+xZs0asVUaAA4cOIC1a9fizJkz2Lt3L5o0aWLSeKrqlVdeMUjoQ0JCMHfuXPj6+gIo+UfUudRDKFlZWWI3FiJrkJ6eDn8AfjIZ3O+NXAAAjysUcIqJwXeLF+OOVotEI45RmetI/wM7MzPTiCMRWU5qaiqAku/V69evwx9AT2dnOJR6+Dbqm2+g702bUv0hElElPVJSrW+hup9EIoFCoTC61bWoqAhZWVkIDw8H8L+CLkFBQcjIyBCPBZTchg4KCkKLFi2wcuXKGnHLTN+KrxcZGYnk5GQxGbh16xbs7Ozg7OyMvLw8JtVkdXL+/RexAByLi4Fx48Tlr23bBmzbhsEANAD61a+PpHvPNuh0OrFVrjLKu47u3LljkFTrh/DU/7tAZE0EQTBIqlXR0SXX1R9/AH/8Ia4XUWqbVAD8tiCqmR4pqXZ3d39g14oGDRpg/PjxWLRo0SMVgKhbty62bt0q9kHWarWQSqWoX79+mVvKGo0GTZo0wdKlSw1ul1nSlClTxB8W27dvR1JSEq5fvy729U5OTkZxcTFcXFzEpJrImuTFxRlUUiyPI4Bxgwcj+V73kIKCAixfvrzSx5g8ebLBdZScnIzr16+LSXVCQoJYjTQrK0v88U1kLfLy8sRuUS4uLtAkJj70umKRcqKa65GS6k2bNuGtt97C+PHjxSeUT58+jc2bN+Ptt99GamoqPvzwQygUCrz55puPFJg+odbpdGJpTkEQkJLyv5tfy5Ytg729PV577bUaNdKHt1IJh3vjYjfx80NWUhJuxcaiZ4cOcJPJUFRcjJSbN+GjVCIbQO6dO8C9EUuIrEHuvda1h5EVFUF+rw+otop9QX2cncXrqGm9eshOTkZiTAzaDBkCJ/x/e3ceFlXZ/gH8OwzDwIzsqyAoCCruIOKaWppLapmapaFlVmppuZX2updlkdpipS32K9O3MvNtcXldyjLF1xVcABVcAAVFQLYZtmHu3x94jjOCyjacOXh/rotLOHM85xbnOXOf5zzP/QD5GRnQAtAAgNGI/IwMuLi41OgcjEkpKyUFGgD2ajXy0tNhazBIHVKjYGtri5deekn8Xm7Hl+pcrO4UVIsSGf3798fkyZMxZswYs+2bNm3C559/jj/++APfffcd3n777XopdSWMn16wYAGOHz+O7du3Y9GiRVi2bBliY2PRqVOnOp+jPuTn58PZ2Rl5AJykDoYxK5MPwBkQqxvccT9uR4zdUY3b0T32ayyOHz+OLl264NixYwgPD5c6nJrT6YCbc620AL7YsAGhoaF3/SseHh4WXXCvMbNU+6jVbU9MTAzWrl1baXtYWBgOHjwIAOjduzdSU1PrFt1NQlJta2sLf39/rFixAtHR0Th69KjVJNSMMcYYY3WlcXBAVFTUvffTaJCYmMiJtRWpVVLt7++PdevWVRofuW7dOvj7+wOoqA7g6upa9wgBcZykSqXCl19+CScnJ+zfv99q70bfnzVLfGxNRFi1ahVKSkvx9rJl+Ouvv7B7zx5MGD8eOTk52LptG956803MmjVL4qgZq75VEyZg1pYt99xv3cSJuHpzDHRxSQlws5pPddypHb3z9tv4+eefcez4ccyZPRu//PILks+fx2+//lqvdfIZs7TvvvsOU6ZORc8ePeDo6Ihru3aZTUpktUNEyMrKAlDRm1vf5XUtffzjx4/j+s2F9IhInHdlOp8tMTERUVFRyMrK4qTaitQqqV6xYgWeeOIJ7NixQ1xK9ejRozhz5gw2b94MADhy5AiefPLJ+osUwKBBg7Bw4ULExMSIk/6skcHODmUmFVBc/Pxw8eJFnElLg2uzZtADOH/1Kry9vaEHKhqPVitZvIzVVFY15wAYVCqxLRhqONLs9nbk7OuLS5cu4UxaGlz8/KA/fhwXrl2Dnasr9ACyi4u5HTFZycjPhx6Ag4cH4s+eBb9764derxcrahUWFkJbz9cFSx/f398f/jePqdPpxBK8ljgXq1+1mir/6KOP4syZMxgyZAhycnKQk5ODIUOG4MyZMxg2bBgAYOrUqWKN6foSERGBgoICq06oq+Lr6wsAOH/+PLy9vQFUlANzdHQEwDV2mfzk5eU1+DmFdpScnCy2I9MKINnZ2Q0eE2N1IZTTc3FxwSWTeu+MMXmq9VTSwMDAGpXHqi9yvEszrasrrKR47do18e6zoKBAstgYq43LxcUoAu5a/qvM1hZ6jabezims3JiWliZWHbp69ao4medO9fMZs1bCzamtrS1KS0uRo1CgTKmE6i5VQMqUSqC8vKFCZIzVQK2T6n/++Qeff/45Lly4gJ9++gl+fn747rvvEBgYiN69e9dnjLLn4eEBoGLRF6Hk140bN6C+OV60sLBQqtAYq5ULBgNaA5j25JP49ccfxXGgvQC4+fpi2LBh0Gs0yLu5OEt9ENpRenq62I6ysrKguZm4czticiO8Z41GIwAg38UFS4cNw/bvvkOQjw82X70KoGJuguFmadkcpRJYs0aagBljd1Wr4R8///wzBg0aBAcHBxw/flwsXp+Xl4d33nmnXgNsDNzd3QFUjI0iInGRGsPN3ggd16hmMqPT6ZAGIMnREXEm2+MAXPX1RUbTpvWaUAO32lFBQYFYrzUnJwf29vYAOKlm8iO8Z4XPAhcXF8QXFCAWwLWbT2YA4KqPDzKaNkVG06bIr+d2xRirP7VKqpctW4a1a9fiyy+/FBdmAYBevXrh+PHj9RZcY6FSqcTllDNMFqgoKysDwMkAkx/hRlB/c4a6KSH5rW+m7Sg/P7/SCqrcjpjcCO/Z0psLI7m4uCAjIwNAxbLljDF5qVVSffbsWfTp06fSdmdnZ15y+w6qenRdXFwMgHuqmfwI79mqEllLJdXArXaUkZEhJtjlN8eX8twEJjdC+xE+C0yTaku2I8aYZdRqTLWPjw+Sk5PRokULs+379+9HUFBQfcTV6Hh4eOD8+fO4cuWKmFQXFRUB4B42Ji9EJPZQVzU50N3dHUVFRZWqcZTWcJnyqpi2I1dXV+Tk5PATHyZbwo2g0J6cnZ2r7Km+du0aSm8+Fa6PdtTY2dra4plnnhG/l9vxpToXq7ta/Q+98MILePXVV/H1119DoVAgPT0dBw8exJw5c7Bw4cL6jtHihBUbLcm0p1q4WAoXUu6pZnJSVFQEullz+vbSejYKBRwdHfHxxx9bJMk1bUfC4lLCnA5OqpncCO9Z4U+VSoWysjLY2NiI1aEAYN3XX6PyQCt2J2q1Gt98841sjy/VuVjd1SqpnjdvHoxGI/r37w+9Xo8+ffpArVZjzpw5mD59en3HaBHl5eXimExLJ9TArWTgypUrYm++0MvHSTWTE9PkNScnx+w1Nzc3ZGdno7CwEAqFAg4Ot4ruEZH4dKa2TNtRhw4dANx6dM5JNZMb4T0r9FgLQ5nc3NzM2pa9Wg3jzc8pIhJvJBlj1qVWSbVCocD8+fPx2muvITk5GYWFhWjbtq3ZnbU1O3PmDN5//33o9Xo0adIEixYtQrNmzWqcXJeUlJhd3IQkOTMzE3YmK8G5urqKyYBpfWqhl0+v18NoNIrLsTNmzYSbQLVaXWmIh5ubm/j42sPDA/369RNfKysrwy+//FLpeLVtR7169TKLh8dUMzkhIjGpFuYiCe3A3d0d165dE/d95JFHUHLz0f+d2hG7xXSImkajscgy5ZY8vlTnYnVXpwE6dnZ2slvd8OzZs4iMjMTw4cOhUqkQGxuLTp06YcWKFXj88cfFR8rVsXz5cixdurTS9q+//trsZ2dnZ0ybNg1qtdoseTBdSVGn04krLDJmzYQk1t7eHnl5eTBd3sXV1RVXb9bWFeYO3Et125GTkxOmT58utiPhJlRIpjmpZnJSWloqltITPguEJNvNzc0sqWY1I3SYAZZbpryhlg5vyHOxuqtVUl1cXIzVq1dj7969yMzMFAvXC6y1rB4R4cMPP8SgQYOwceNGcfvUqVOxYMEC6PV6PPPMM9VObt944w3MmjVL/Dk/Px/+/v6ws7MT7yZLSkqQl5eH/Px8MeEQLqSmSUBxcTEn1UwWhEfUt5e0AyoS7Rs3e5qFxY3u5U7tSK1WQ6FQiI+78/PzzdqRcN0RenHqOrSEsYZkMFk1UZh8KGyzt7dHMQ8LZEx2apVUT5o0Cbt27cLo0aMRGRkpm8cRCoUCOp1OHOdZVlYGlUqFNWvWQK1WY8mSJWjVqhUGDhxYrcmLarW6ysThkUceEet379y5E/n5+bhx44Z4hylUK9DpdLC1tYXBYOAxckw2hGRamKxoqqysTFyMRXif38ud2tGQIUPu2o6EBERIprkNMTkxvSlVqVQoLy8XqzuUlZVV+6aUMWY9apVUb926Fdu3bxfHNMqJq6srduzYAaDiQlZSUgK1Wo0PP/wQV65cwbRp0xAfH2+2qE1daLXaSsmAabUCOzs7TqqZrAjJgNBjbaqsrAzq224e60NV7Ujo3RN6qrkNMTkxLY+mUqlQXFxsllQ3uXlzyhiTj1rNjPPz85PtUIW5c+fCaDRi7NixACp6yYSerjfffBMFBQU4ePBgvZ1PSABMkwEhCSgtLRUvolx7lMmFaU/17UNAatNTXR1VtSOh6ofQnoxGo9kjdcas2e091abbysrKzCbpMsbkoVZJ9cqVKzF37lykpKTUdzz1KjU1FRs2bMC7776LY8eOAaioSDB//nycPHkSkyZNAgBxOIhKpYJGoxGTgvogJAC5ubniZAPTpZ2Fiyn3sjG5MO2pvv2JjsFgsGhSbdqOhJth0/bE7YjJhUKhECfbCu3ItKe6Pj+HGGMNo1bDPyIiIlBcXIygoCBoNJpKH6y3166VwqlTpzBixAh4eXkhOzsbixYtwi+//IJHHnkEjz/+OPR6PT7++GMMGDAAn3zyCcrLy7Fp0yaUl5ejWbNm9RaHaQ9bq1atAFSU0hMqGAgJCicDTC6E96zRaKwYC3qzxxgwHwtqqZ5qoR0JVUhMx3YXFxfz7HgmG0qlEkajUUymTXuqOalmTH5qlVSPHTsWV65cwTvvvANvb2+rm6h48eJFDBs2DOPGjcO//vUvqFQqzJ8/HzNmzEBkZCQ8PDzw7LPPonXr1pg/fz569+4NV1dXGI1GbNmyBb6+vvUWS1U9bHl5edBoNGZJNQ//YHIhvGcNBgM0Gg1MR1ZbeviHaTvKz8+HUqlEeXk5bGxsYDQa+eaUyYpSqRQnzAO3FiLj4R91o1QqMXr0aPF7uR1fqnOxuqtVUh0TE4ODBw+iU6dO9R1PnZWVleGLL75AZGQkFi5cCI2mooruI488gv/85z/i4zatVosBAwZgwIABOHDgAJycnODp6QkfH596jUdIBoqKisTeiLy8PGi1Wty4cYN7qpns3N5TXWzymmlSXZ83indqRxqNBgUFBeKkY25HTE6E9zL3VNcve3t7/PTTT7I9vlTnYnVXq6S6TZs2VlsTVqVSiQvSCAk1AISFhaGoqAjp6elwdnaGUqkUy+ZZsoqJSqWCnZ0dSktLxWoJ+fn5cHd3BwCzetaMyYFpUn17b1ppaanZ8A/TiYN1mUR4p3bk6emJgoICvjllsiS8b4WkWvg8MG1HQEXbMZh8zxizTrVKqt99913Mnj0bb7/9Njp06FBpTLWTk1O9BFdb48ePr7RNWAbcxsZGvJAdP34crVu3tvjy6k2aNEFOTo54I0JEYjLCSTWTm6qqFghMJyqWl5fjP//5T72dt6p2JCQeQlJSXFx8x7/PmLW5PakWnqTeXqf6t99/h77yX2eMWZlaVf8YPHgwDh48iIceegheXl5wdXWFq6srXFxcarTMt6UJE5jKy8vF8nVCAj137lwMHDiwQT6ETcecC9/f/idjcmGaVN8+xs9gMMDBwaFeJ/sKqmpHd/qZMTkQ2o/wp/A+FtoRqx2dTgeFQiEu+Ca340t1LlZ3teqp3rt3b33HUSd3Wv1Q2KZUKsVFVoxGIxYtWoRPP/0Uf/75Jzw8PCwen5C4q9VqMdEXHuEJP5suBMCYNbtXT7VCocBzzz1XaaJiSUkJVq1aVevz3q0dCUNC6mvRJsYawu1JtUBoR4I5s2ej7ObTzbq2I8aY5dQqk+vbty9yc3Oxbt06JCYmAgDatm2LSZMmwdnZuV4DvJvc3Fy4uLhUq5dKpVLB2dkZ06ZNw+7duxETE4MuXbpYPEYiEpMBIXFWqVRiwiEkB5wMMLkQHlEDVScDQMUN7e3jrata1ry67tSOhMmQRqNR3MaYXAht6fae6ttvSO3s7KC42Z7q0o4YY5ZVq+EfR48eRXBwMD744APk5OQgJycHH3zwAVq2bInjx4/Xd4xViouLw/Dhw3Hy5Ml77ms0GpGdnY0zZ87gr7/+wuHDhxskoQYqLo5CL5pw4WzSpIk4hpqTaiY3pon07U9Y6rOMnimDwVBlOxISbeE1LkPG5ORuwz8YY/JTq6R65syZGD58OC5duoQtW7Zgy5YtYm3oGTNm1HOIlZ04cQKRkZHo0aMHOnbsaPaa0GMlfMgCFb0BQUFBeOeddxq8FKDpI2sh4TBNqrmHjcnN3XqqjUTie7o+CZMT1Wq1mHBotVqxp1rYxu2IycntPdVCJ4vRaLRIO2KMWVate6rnzp1r1ktla2uL119/HUePHq234KoSHx+PHj164I033kB0dDSICDk5Obh48SKAyheptWvXijHNmzcPHTp0sGh8txOSakdHR/F7rVZbKanmMdVMLkyT6qret6Y3tPVFaDtNmjQRE2ytViveqPKYaiZHwueU0KZMhzJybzVj8lOrpNrJyQmpqamVtqelpcHR0bHOQd1JdnY2RowYgTZt2mDp0qUAgEmTJmHgwIHo1asX+vbti7i4OPFuPzMzEy+99BKWLVsmWck6IQG4PRngnmomV3er/gFYJhkwTaqF76taHIPbEZMTIZkW/jQdL81JNWPyU6vu0SeffBKTJk3CihUr0LNnTwDAgQMH8Nprr2Hs2LH1GqApd3d3DB48GHFxcViyZAm2b98Od3d3TJ48GZ6enoiOjsaIESOwZ88eBAcHw8vLC2fOnDGrZ9vQquqp1mg04gWTk2omN6a9aQ2dVDs6Ooo3p0LJMYVCISYjPKaaycntY6qNRiOUSiXKy8st8sTnfqFUKvHII4+I38vt+FKdi9VdrZLqFStWQKFQYMKECWZjGadOnYp33323XgMUCIu3rF69GrNnz8batWsRERGBdevWwdvbGwAwYsQItG/fHsuWLcM333wDg8GAVq1aWSSe6qqqp9q0h40fWzO5USgUsLGxEROA21lisqJpOzKdpwBUtCfhdW5HTE5u76k2GAyws7NDUVER91TXgb29PbZt2ybb40t1LlZ3tUqq7ezs8NFHH2H58uU4f/48AKBly5Zmy4LXF51OB6PRCCISV2pcuXIlfH19ERgYCC8vLwAVyalSqUSbNm3EAunWME7ZtIctOzsbwK2kWqFQ8AQrJkt3S6otPfzjxo0bAG71SnNSzeTq9jHV5eXlUKlUKCoqslglHcaY5dRqTLVAo9GgQ4cO6NChg0US6oSEBIwcORJ9+/ZFaGgoNm7cKPbszp49G8OGDTNb4EVYBKZt27YArKOeZ2FhIQDAxcUF+fn5AG71sDVp0kS8cPJjayYnWq0WAMQVSk1lZGTU+/lM21FBQQGAW21GuPbY2dnxyopMVoR2JHS0ZGVlwdPTEwBw9epVyeJijNWO9F25d5CQkIA+ffpgwoQJiIiIwLFjxzBx4kS0a9cOnTt3BmCeiBoMBixduhQHDhzA8uXLAUi3dPEvv/xSaZuXl5fYwyaMBXV1dRUnfFripoQxS/H09EReXl6VK5ImJSXVS9nK//73v2IbFnqiTduRcHPq5OSEK1euiAkKY3IhDF0Ubk5TU1PRrVs3nD9/HhcuXJAyNFnT6XTiU+zMzMx6vzZY+vg1OZewAN/deHh4ICAgwCLxMXNWmVTn5ORg5syZePrpp8XlWMeNG4fjx4/j66+/xscff2y2NPnu3buxevVqHDlyBNu3b0dwcLCU4Vfi5eUFFxcXMRkQHlG7uLiISbWQaDMmB56enkhOThaHZJk6f/68OAeiLoQhHwInJye4urqK7ch0ERiAb0yZ/AhJtdBBlJ6ejvbt22PLli2cVJtITU1FVlbWXfe5PbnU6/WWDMnix7/XuTw8PKDRaBAVFXXPv6/RaJCYmMiJdQOwyqS6rKwMubm5GD16NIBbkxQDAwORk5MD4FYvNBEhMDAQbdu2RXR0NNq0aSNZ3ILp06ebVRtxcHBAYWEhysvLzRIN04SEk2omJ8IjarVaDdPnQVqNBtf1ely5cgX+/v51Osczzzxj9jTK3d0dxcXFMBgMUCgUYuUcoeeGe6qZ3AhJdVFRkXjDaG9vD61Wi+Kbc4Pud6mpqQgNDa1WEqvRaKp8etYYBQQEIDExsVo3G1FRUcjKyuKkugFYZVLt7e2NDRs2ICQkBADEZNTPzw8pKSlm+xYVFSE4OBhvv/221ZSb0Wq1lUr4Cb1rnp6e4vhQoaa3ra0tT7BisiIk1YWFhWjevDlws1126NABfx46hKSkpDon1U2bNq3UjjIzMwFU9NII7Ui4IeWkmsmNkFTn5OSgVatWOHToENLS0tCpUyfExcRIHJ11yMrKgl6vx4YNGxAaGnrXfYVhDrr75IYkICCAE2UrY5VJNQAxoTYajWLCSUTihyoALF++HHZ2dnj11VetotLH3Qg97N7e3uKEReGxNfdSM7kRkuqcnBy0bNlSTKo7duyIPw8dQnJyMh566KF6P69wc+rl5YW8vDwAtyZ58fAPJjfCWNns7Gy0b98ehw4dQkpKCjp37sxJ9W1CQ0MRHh4udRiM3VXdBj02ABsbG7MqHsLwiUWLFmH+/PkYMGCA1SfUgHkywEk1kzshqb5x44bZHIaOHTsCqKgAIvQk1yehHXl7e1eqAsI91UxuhJ7q7OxscU2F1NRUcTI+Y0xerD6pBm6VxrO1tYW/vz9WrFiB6OhoHD16tF6qDDQEIRnw8fERkwEhmeYeNiY3pj3Vpkm1s7MzAgMDAUCsYV+fquqpFp5kcTticlNVUp2SkgJnZ2e0aN5cytAYY7Vg/V28uNU7rVKp8OWXX8LJyQn79++X1aOgqnqqhcfW3FPN5EZ4bH17Ul1eXo6wsDBcvHgR586dMxsDWVpaWufz5ubmAjAfRiXMpeCeaiY3QlJdWFiIZs2aQaFQoKCgADdu3KjoMLo5rKq0tBTCUjD10Y4aOxsbG/Tt21f8Xm7Hl+pcrO5kkVQLBg0ahIULFyImJkZc4EUOiMhsTLWQGAiTsDipZnJj2lPt6+srbk9PT0dYWBi2bNmChIQEJCQk1Ns5iUhcldTd3V184iN80HBPNZMbFxcXqFQqlJWVoaCgAIGBgbhw4QIuXrxYkVT/9hsAYMXKlWi4Am7y5+DggL/++ku2x5fqXKzuZHXbExERgYKCAlkl1EBFSZuioiKoVCrk5+cjLy8PKpVKXM7Zx8dH4ggZqxlhiEdWVpZ4kwgA+/btQ0hICIKCgur9nBcuXEBhYSGUSiWuX78OIoKbm5tYUso0uWdMDhQKhTjsIyEhAREREQCAP//8s2ICMGNMVmTVUw3I7xEvEWH//v0AgOHDh2PXrl0AgBEjRojlAe9VJogxa+Pq6or27dvj9OnTiIuLQ++b2/fu3YvhTz2F5cuXo6SkxOzvFBUVYfLkybU+58GDBwEAAwcOxL59+wAAjz/+OP7++28AQJcuXWp9bMak0q9fP8THx+PIkSN4+umnsWnTJhw9etSs0tVXX34Jw80nm3VtR4wxy5FVT7UcXbhwARkZGVAqlejatSuOHj0KAIiKisK5c+cAwOpWgGSsOnr3rkilT5w4IW4rKS3FH3/8AYVCAXt7e7Ov22tO10RmZqY48bFr1644ceIEFAoFhg0bJq48x0k1kyOh9OShQ4fQsmVL9OnTB0SE//73v+I+arW6XtrR/UKn08HT0xOenp4WqVlt6eNLdS5Wd7LrqZaD//znP+LkqYyMDAAVvWsHDhwAEaFnz54IDAxEUlISAE4GmDw98MADWLt2LeLi4sy279ixA4888kidS12ePn1arOwRHx8PAOjWrZuYxPfs2ROFhYUwGo1o2rQpD/9gstS3b18oFAqcP38eWVlZePbZZ7Fv3z7xaQyrnXutNGjtx5fqXKxuOKm2gLNnz5r9bGNjg4cffhj/+te/AFT0Uqenp0On00GlUnFNUiZLDzzwAADz97uHuztSs7Nx8OBB8fXa2rp1a6VtQ4YMwapVqwAAY8aMEZNtvjFlcuXu7l6xgmJcHI4cOYLBgwejbdu2uFSPk3wZYw2Dk2oLiIqKEhekAICgoCCcPn0axcXFaNGiBXr37i3O5m3RogUvUc5kyd/fHwEBAchKTRW3PfHEE1i5di22bt2K3r17Q6FQ1Pr47dq1E5/4AEDr1q2Rm5uL/Px8eHl5oW/fvli0aBEATqqZvD344IOIi4vD4cOHMWTIEDzzzDNYOneu1GExxmqIk2oLGDBggFl5L6PRiE8//RQA8PTTT8PGxkYcTy3M9mZMjh544AH8Z+NG8edRo0Zh9ddf4+LFi4iPj0f79u1rfezXXnutUpm8pUuXAgBGjhwJW1tbsWQfJ9VMzh588EF88MEHOHz4MICK4YJrV64ETCYrMsas332bVF+9ehUFBQUICQmx+Lni4uKQkZEBe3t7PPbYYwAgJtV1SToYk9rtSbWzszNGjRqF77//Hv/3f/9nVv5SKCFZW1euXEF8fDwUCgVGjRoFvV7PkxRZo9CnTx/Y2NggJSUFV69ehY+PD5566ing448BAHPmzIH+5lMfo9EoZaiMsbu4L5Pq1NRUtG/fHg8++CCio6PRunXrWh2npKTErGyYsMLb7bZt2wYAePLJJ8WeNyGpLiwshE6nk12pQMYAVDluevz48fjhhx+QlpaGtLS0ex6juu1oz549AComdjVt2hSxsbEwGo3w8vLiSYpM1pydnREWFoZjx47h8OHDePTRRzFixAgxqb6WmcmLvzAmA/dtUq1Wq/HHH3/glVdewSeffIKWLVvCxsYGRFTtcaDLly8XH0ff7VynTp2CQqHAuHHjAAB6vV6sUX358mVOqplstWnTBm6ursCNG+K2gIAAfPrppzh58qTZvqWlpfj6668rHaM67ai0tFSsRz1mzBgAtyqChIeH1+nfwJg16N+/v1lS3aRJE/G1r9etQ7m9PYCKz48XXnhBqjBlwcbGRhxaaallyi15fKnOxeruvkyqAwMD8fjjj+P1119Hr169MG3aNHz55ZcICAhAWloaAgICqnWcN954A7NmzRJ/zs/Ph7+/P/744w9xouKRI0cAVIyz9vPzAwD8888/ICIAFaVyOKFmcmVjY4OwsDDgzz/Ntvfp0wd9+vQx21ZYWFhlUn2ndmTqf//7HwoLC+Hr64tevXoBAGJiYsRzMSZ33bt3BwAkJydXeq1Dhw6gm085CwsLGzQuOXJwcBA/e+V4fKnOxeruvkyqvb29ceDAASiVSuzbtw89e/bEtGnToFKpUFZWhs2bN0OlUt2zx1qtVldZiP+7776rtO25554Tvzct6u/g4CAWdOfkmslR8+bN6/T379SOPvvsM7HWtVDTfdSoUVAqldDr9Th06BCAipVKGZM7YRGwVJNqOowxebnvkury8nLY2tqiefPmOHjwIMaNG4fz58+jadOmMBqN2LRpk1k5vNp46KGHzBa+6Ny5Mzp06ACgopfBtKi/ra2tOIaUk2omRy1atLDIcYVKCAKlUomRI0cCqOi5Likpgb+/P9q1a2eR8zPWkIKCggBUPKnJy8uDC5daZUx27rukWqh726VLF3Fc85w5c+Dk5ISSkhKsW7cOrVu3Rps2bWp9jnfeecdsPJypvXv3mk3KUqvVKCkp4YSayVZde6rvZObMmWY92O3atYOnpycAiHXehw4dWqda2IxZC61Wi6ZNmyIjIwOpqalwadlS6pBkS6/Xi5WHEhISKpXmtPbjS3UuVneNOqnOzc2Fg4OD2QezMBHRy8sLhw4dwpQpU7Bt2zYcO3YMRITmzZtDrVbj3//+t0UWZdmxY4fZzzz0g8mdpZLqJ598ssqbU6PRKE5a5FJ6rDFp2bIlMjIykJaWho6cVNcaEYmdZsL8JTkdX6pzsbprtEl1QkICevfujZdffhlLliwxW5kNqCgFtmTJEjg7O2P79u1o1qwZAODChQsoKSmxSEKdl5cnTq4y3ZadnV3v52KsoZgm1WVlZRa/qJw+fRrZ2dmws7NDYGCghc/GWMNp2bIl9u/fX61SlI1BamoqsrKy7rpPYmJiA0XDWN01yqQ6PT0dEyZMgL+/P1asWAGFQoHFixdDqVSKj4oDAwOxZMkS9O/fX3y0YjAYLDY+FAB2795daQGMK1euwN3d3WLnZMzShCEZAJCRkQF/Z2eLnk8Y+tG0aVOL3PwyJhWh8tT9kFSnpqYiNDQUev29K3BrNBp4eHg0QFSM1U2jS6qNRiP279+PwMBALFq0CHFxcZg4cSIAiIk1EcHZ2Rkvv/yyWd1H08mFliBU/RCqjCiVSoSEhJglJYzJjemY5vT0dPjXYT5CdQhJdbdu3eDg4GDRczHWkIRFjO6HpDorKwt6vR4bNmxAaGjoXff18PCodqlbxqTU6JJqGxsbhIeHw8nJCR06dKio70kklrRbtGiRmDw35ASnrKwssZqBo6Mj8vPzodFoEBgYiOvXr8PLy6vBYmHMUtLT0y1+/HPnzsHGxgYPP/ywxcZzMyYF4Unp/ZBUC0JDQ3kBJ9ZoNLqkGqio9ynU/DQajZgwYQIUCoVZj7XRaMQPP/yA9u3bVyxeYWEbN26E0WhE+/btUVBQAJ1Oh/Lycuj1ely+fJnLgrFG4fLlyxY9vjDRNyAgAJ06deKbUdaoeHt7AwAyMzNRVFQkcTSMsZpqlEm1KWF4x/jx4wEAEydOBBHh2rVr+PHHHystpWwJGzduxJdffgkAGDlyJP755x9cu3YNZWVlOHnyJF588UWLx8BYQzh48CCmW+jYly5dwtq1awFU1PTl0lKssXFxcYG7uzuys7Nx8OBBcG2b2lEoFOJcKUs8kbb08aU6F6u7RpFUC2Xy7mX8+PEgIjz77LNwdnbGn3/+aZHHxz/88INYxu/atWv45ptvAACvv/46+vfvj+TkZKhUKhiNRpSVlfEjbNZoJJ8/j+TkZPFJUX0pKyvDvHnzUFRUBCcnJ7Rs2ZKr5rBGx8vLC926dcP27duxa9cuTJM6IJnSaDSIj4+X7fGlOherO1kn1bm5uXBxcan23VtpaSn+97//wcnJCTExMfecHFFbH374YaVts2fPxqxZs3Dt2jWUlJRAoVDAYDBAqVQiJSWFH2OzRmPHjh2YPr3u/dWmN6cJCQk4ffo0nJyc4OLiwo/GWaOk1WoxYMAAbN++Hfv375c6HMZYDck2qY6Li8P06dPx6aefomPHjvfcn4jwzz//4Ndff8Xu3bstllADwOjRo81KffXq1QsvvfQS9Ho9srKyYGdnB3t7e9jY2KBJkybIycmxWCyMNbQdO3Zg2rRpdX5UWdXNqVarhYeHB+zt7eHv71+n4zNmbXQ6HcLDw+Hp6Qnd9etSh8MYqyFZJtUnTpxAZGQkZsyYUSmhNhqNsLGxQXl5udmCL8K4pNjYWIv3Cq9btw5OTk6Vtmu1Wmg0GuTl5aGwsBAKhQJ2dnbiwjOMyZ29Wo3U1FTEx8ejffv2dTrWiBEjkJubi6ysLOTl5QGoKEdpMBhgb2+PrKwsXvyFNSo6nQ5qtRrh4eH4Z+dOqcORLb1ej65duwIAjhw5YpFlyi15fKnOxerO5t67WJf4+Hj06NEDb7zxBqKjo0FEyMnJwcWLFwHcmpgoJNRr167FsWPHAFQsFiH1MIusrCzcuHEDNjY2UCgU4uNtYblyxuRs6NChAG5V6aiLb7/9FkuXLsW0adPQunVr2NnZwdbWFm5ubrCxsUF6ejoyMzO57bBGQ6vVorS0lBc6qSMiQkJCAhISEiy2TLkljy/VuVjdyaqnOjs7GyNGjECbNm2wdOlSAMCkSZNw8uRJpKenIyQkBB999BE6deoEhUKBzMxMvPTSS3j00Ufx448/igmslDQaDby9vXH69GkoFArk5OQgMzMTnp6e0Gq1UofHWJ088cQT+G7LFvz3v//F7NmzzRZXqg0HBwdkZGQgJydHfPrk4+MDb29vuLq6wmAwQKfTcdthsqfT6aDT6eDu7g4/Pz84OToCBQVSh8Uaieos986L7NSdrJJqd3d3DB48GHFxcViyZAm2b98Od3d3TJ48GZ6enoiOjsaIESOwZ88eBAcHw8vLC2fOnAERWUVCDVQU9+/atSvi4+ORm5sLtVqN/Px8TgpYo/Dwww/D0dER165dQ2xsLLp0qVtRMI1GI47Ntre3R2BgILp06YLevXvD29sbtra23HZYo6DT6WAwGAAAbdu2RVBQEHDiBADg888/R+nNeTqlpaWSxcjkx8PDAxqNBlFRUffcV6PRIDExkRPrOpBNUi2MlV69ejVmz56NtWvXIiIiAuvWrRML5o8YMQLt27fHsmXL8M0338BgMKBVq1YSR25Oq9XC0dFRTBRKSkpgb2+PzMxMeHl5cYLAZM3e3h6PPfYYNmzYgK1bt5pNCNbr9TU+nqenJ0pKSuDo6IiysjKUlZXh0qVL8PT0hLe3N7RaLbcZ1ihotVrodDpoNBqEhYXh4YcfFpPqr9atQ81bD2MVC2UlJiYiKyvrrvslJiYiKioKWVlZnFTXgdUn1TqdDkajEUQkTv5buXIlfH19ERgYKI6RFh4Nt2nTRhxjKSxHbm0CAgLg6uoKIoJWq0V6ejqcnZ05QWCNQlRUFDZs2ICffvoJP/30U52O5eXlhb59+6KgoABpaWm4ceMGMjIycPHiRYSFhXGbYY2G8F7OzMxEWlpaRU/1TUGBgQjp3Bl+fn4oLS3FF198IWGkTG4CAgI4UW4gVj1RMSEhASNHjkTfvn0RGhqKjRs3ory8HEBF3edhw4aJPb5KpVJcBEZYfchaB/W7uroiNDQUrq6ucHJyQnFxMZRKJScHrFHo3r17vS7+0qxZM4SFhcHFxQXe3t5QKpXw9PREbm4utxnW6Oh0OpSXl+P8+fPitv79+2PGjBlYvXo13n//fQmjY4zdjXV25aIioe7Tpw8mTJiAiIgIHDt2DBMnTkS7du3QuXNnAICdnZ24v8FgwNKlS3HgwAEsX74cgPUu6VlUVITs7GwYDAYYDAb4+vqibdu2nCCwRqGkpASbN2/GpUuX8M8//yAmJgZOTk4YMmQIZsyYUePjXb58GefOnUNZWRl0Oh2CgoLQrFkztGnThtsMa5QuXbqElJQU8eeuXbuiTZs2EkYkLwqFQlyp2FLLlFvy+FKdi9WdVSbVOTk5mDlzJp5++mmsWrUKADBu3DgcP34cX3/9NT7++GOzpcl3796N1atX48iRI9i+fXu9L5Fc34qLi1FaWgqdTof8/Hxcu3aNKxiwRkOr1aJp06YIDg5GUlISXFxc6vQkJiUlBZmZmTh37hy8vb2h1+vh6+sreXlMxizBy8sLZWVllRYQA4DMzEwUcEWQe9JoNLh06ZJsjy/VuVjdWeXwj7KyMuTm5mL06NEAKiYpAkBgYKC4+qCQUBMRAgMD0bZtW+zduxdhYWHSBF0D9vb20Gg0cHFxgZOTE2xsbHCdV89ijYRWqxUn3arVajg6OsLPzw9ubm61Op4w5MPX1xdlZWVwc3PjZcpZo6XVatGrVy94enqabTcYDLh+/bpYIYQxZn2ssqfa29sbGzZsQEhICICKSYg2Njbw8/MzeyQGVAylCA4Oxttvv222gqI1c3NzQ2hoKJRKJUJCQuDp6cmrJLFGycvLC8HBwWjevHmtnyAFBwcjMDAQTk5OsLW1hbu7ez1HyZj10Ol08PT0xEMPPQR8/DGAiso5Gk9PeHp6ck81Y1bMKnuqAYgJtdFoFB+DEREyMzPFfZYvX441a9bAYDDIJqEGKupGlpaWws3NDXZ2dmjXrh0P/WCNTmZmJoqLixEYGIg2bdpU6nmrDp1Oh+zsbPj7+yMgIACenp5o2rQpioqKeCVF1igJwwLt7e3FbUVFReIToNq0o/tNUVERunbtiq5du1rkqZaljy/VuVjdWWVPtSkbGxuz8dPCCm2LFi3CsmXLEBsba7Wl8+5Eq9UiKCgIsbGx4ixvOQxbYawmrl+/jrKyMhw6dAiXL1+uVRIsVEK4du0a3Nzc4OPjg4KCAqjVamRmZiIwMNACkTMmHa1Wi9jYWKQmJqL3zW2XL1+Gm78/v9+ryWg04ujRo+L3cju+VOdidSeLbFRIqm1tbeHv748VK1YgOjoaR48eRadOnaQOr8a0Wi18fX1x6dIllJWViWUCGWtMNBoNMjMzYTQakZaWBgcHhxofQ6vVorS0FA4ODrh69ao4lITbDGushKeWN27cELeZPqFljFkvqx3+YUronVapVPjyyy+xbNky7N+/H+Hh4RJHVjtarRaurq5QqVTIz8/H5cuXpQ6JMYto0aIFmjRpArVabVbNoLq0Wi2aNWsGnU6HkpISnDhxAnq9Ho6Ojlz9gzVarVq1glqtFn9u2rQpAPCQJ8asnCySasGgQYMAADExMYiIiJA4mrpxcHCAWq2GVquFra0tXyxZo9S0aVMMGjQIw4YNQ6tWrWp1jBYtWqBTp06wtbXFjRs3kJSUBAA8D4E1WoGBgWjXrp34c/v27aFWq/lzgjErJ4vhH4KIiAgUFBQ0ig9TjUaDdu3aISsri8fJsUbJy8sLer1efBJT06odOp0OxcXF0Gq16N27NxwcHHD+/PlaDSNhTE50Oh1cXV3Fnz09PZF582mNHIY+paamIisr6677JCYmNlA0jDUcWSXVQOPpndLr9cjLy4NarYaXlxc/ymaNjlarhUajgZ2dHTQaDTw8PGr09/V6PYgI169fF0tO+vr6olmzZtxeWKOWmZlptmKwVquFFhW1qvV6vXSBVUNqaipCQ0OrFWdtrguMWTPZJdWNSXl5OcrLy5GdnY0WLVpIHQ5j9S4rKwtnzpyBUqlE9+7da/R3NRoNcnJyoFarce7cOeTn50Or1cLT07PR3FwzVhW9Xo/09HR0uPlzZmYmtF5e0Ol0Vr+mQVZWFvR6PTZs2IDQ0NC77uvh4YGAgACLxGHpZL0hbwb4xkM+OKmWSPPmzZGTk4OUlBQUFhZyeTDWaBUVFaGwsBDx8fE1+ntarRZOTk7Q6XRwc3NDdnY2rl+/zuNKWaN3e+J87tw5hN1cpfTq1asSRVUzoaGhkhUT0Gq1Fl2l2NLHl+pcQPWG5VjyZkjuOKmWiJeXF8LCwuDi4mL1j/MYq63mzZsjISEBAGo1FlSr1YpfRUVFUCqV3F5Yo+fl5YUck95JZ2dn8WaSlylnluDh4QGNRoOoqKh77qvRaJCYmMiJdRU4qZaYjY0Nj6lmjZaXlxeGDBmClJSUOi9c4O7uDr1ezyvKsUZPq9WiTZs24s/e3t7ikCdeppxZQkBAABITE6s1wTQqKgpZWVmcVFeBk2rAbMXGhqLT6cREQ6PR8BhR1mgJN435+fm1PoawsiIv08vuF6afCV5eXsDNn/mm8t6KioowZMgQAMCOHTvqvWKQpY8v1bkCAgI4Ua6j+zapLigoQGFhIZydnSWZ+KHT6WA0GvlRNmv0dDqdmBTXllarRW5uLpRKJa5fv85Pdth9RZioyJ0v1WM0GvH333+L38vt+FKdi9WdrBZ/qS/x8fEYOnQoBg4ciDZt2mDXrl0AKnqsG4pWq0WTJk3QpEkTvlCyRu3SpUuIj49HampqrY+h1Wrh7++P8vJyq69+wFh9MF2avLy8nCfoMiYD911SfebMGfTr1w/h4eFYsWIFunXrhilTpqC8vBwKhaJGiXVJSQny8/PNvqpLqOHLq2Sxxk6v16O8vPyOT2Vq0o5cXFwsFCVj1sV0bKtSqeTOF8Zk4L4a/mEwGLBs2TIMGzYMH374IQDAz88PixcvxvXr12Fvby9+Vcfy5cuxdOlSC0bMmPw1b95cbF9V4XbEWGWmtYlNx1QzxqzXfdVTbWNjg+vXr6NZs2bitu+//x67du3Cgw8+iM6dO2Pp0qXIyMio1vHeeOMN5OXliV9paWk1isfLyws+Pj48PpQ1al5eXmjXrt0dJ1hVtx1xe2H3E36fMyY/91VPtY2NDVxdXfHdd9/Bzc0NFy5cwFdffYUvvvgC3bt3x2+//YbPPvsMvXv3xtChQ+95PLVaDbVaXet4hPq7jDVmOp0OmZmZdywFVt12xO2F3a8uXboE0mg40WbMyt1XPdUAsHHjRkRERCAhIQF//fUXFi9ejPHjxyMkJASzZ8+GRqPBb7/9JnWYjDUaOp1OrLbDGKu5goICFBQU8PybGtBoNBad1Gzp40t1LlY3jbqnOjk5Gb///jsyMjLw4IMPolOnTvD19cXmzZthNBrxyCOPoHXr1gAqJkvZ2NjA399f3MYYqzutVgtHR0epw2BMthwdHUE31zOoS2nK+4VWq7XoDYiljy/VuVjdNdqe6tOnTyMyMhJbtmzBvn378Pjjj2POnDnYunUrgIqhIDY2Nvjoo49ARMjOzsZ7772H2NhYDB8+XOLoGWs8tFotAgMD0aJFC6lDYUyWWrRogcDAQB7+xJiVa5Q91UVFRXjjjTcQFRWFDz74AEqlEv/973/xwQcfYOXKlSgqKsITTzyB+fPnY8KECdBqtWjdujXy8/Oxfft2hISESP1PYKzRuNeYasbY3V26dAmeLVpInlSnpqZWaxlrxu5XjTKptrOzw5UrV9C9e3colUoAwODBg+Hi4oLly5fj66+/RsuWLdGrVy/ExMRgy5Yt8PPzQ+fOnXmJTsbqGY+pZqxuCgoKoNHpJE2qU1NTERoaWq1VgDUajVlJwIZWXFyMUaNGAQB+/vnnapfJtZbjS3WumqjPmycPD49Gk3s1uqSaiFBSUoKmTZuKd9Tl5eVQKpXo3r075syZgylTpmDTpk0IDw+Ht7c3pk6dKnHUjDVePKaasbpxdHSUvJc6KysLer0eGzZsQGho6F33lTpJKi8vx/bt28Xv5XZ8qc5VHR4eHtBoNIiKiqq3Y2o0GiQmJjaKxLrRJdUKhQIajQbDhw/Hq6++iiFDhmDgwIEwGo2wsbHBAw88gGnTpmHu3LmYPXv2HWvnMsbqhzCmuiYrjjLGbmnRooXVLP4SGhqK8PBwqcNgEgkICEBiYuI9hwFVV2JiIqKiopCVlcVJtTUiIigUCkyZMgXHjh3D6NGjsWPHDvTq1UvcJzg4GC1atBCHhjDGGGOMsXsLCAhoFAmwJTSKpFpIpAGIfwLAu+++i6KiIgwcOBBr1qxBnz594O/vj507d4rVPxhjjDHGGKsrWSfVubm5cHFxMUukTbm7u+PLL7+Ev78/Zs6ciSZNmsDLywsXL17E7t274eLi0rABM8YYY4wxM9WZ+Cj1WP3qkG1SHRcXh+nTp+PTTz9Fx44d77ifg4MDli9fjkcffRRXrlxBaWkpevbsyTVzGWOMMcYkVJOJj3KY0CjLpPrEiROIjIzEjBkzKiXUwoREoeKHMDSkR48eFo+LiACAJ2Sx+4vpal/5+cAdZqhfvXoVwK12cifcjhiDWbu6fv48NJ6e0Gq1Yruobjs6cOBAvVQOOXv2LACgsLDQ6tum6QqE+fn59V41wyLHv8N11NL/Fqm5uLjg8OHDyM7Ovut+Z8+exYsvvoidO3fWy6rXwu/1Xu2oxkhmTp8+TQ4ODrRo0SIiIjIajZSdnU0XLlyocv81a9bQ0aNHGyS2tLQ0AsBf/MVfd/lKS0vjdsRf/FXHL25H/MVfdf+6VzuqKQVRfafplpOdnY3u3bvD0dERx48fBwA899xzOHnyJNLT0xESEoKPPvoInTp1gkKhQGZmJnx8fPDoo4/ixx9/hFqttmh8RqMR6enpcHR0vOM474aQn58Pf39/pKWlwcnJSbI4akPOsQPyjt/SsRMRCgoK4Ovre9dJwjVtR3L7nXO8ltXY47VUO6pLTFKw9hg5vrqxls+jmpLV8A93d3cMHjwYcXFxWLJkCbZv3w53d3dMnjwZnp6eiI6OxogRI7Bnzx4EBwfDy8sLZ86cARFZPKEGABsbGzRr1szi56kuJycnq2ws1SHn2AF5x2/J2J2dne+5T23bkdx+5xyvZTXmeC3Zjmobk1SsPUaOr26k/jyqKdnUlDMajQCA1atXIzIyEmvXroWXlxe++eYbvPDCCxgxYgRiYmLQpEkTLFu2DABgMBjQqlWrehl/wxhjjDHG2J1YfU+1TqeD0WgEEYl3KytXroSvry8CAwPh5eUF4NZS5G3atBEHoNvaWv0/jzHGGGOMNQJW3VOdkJCAkSNHom/fvggNDcXGjRvFma+zZ8/GsGHDxLFippU+2rZtCwD1P6tTJtRqNRYvXtwgQ17qm5xjB+Qdv1xjl1vcHK9lcbx1Z40x3c7aY+T46sba47sTq52omJCQgD59+mDChAmIiIjAsWPHsHr1ahw+fBidO3eutL/BYMDSpUuxbt067Nu3D8HBwQ0fNGOMMcYYuy9ZZVKdk5ODsWPHok2bNvjoo4/E7Q8++CA6dOiAjz/+2Gxp8t27d2P16tU4cuQItm/fjrCwMKlCZ4wxxhhj9yGrHP5RVlaG3NxcjB49GsCtSYqBgYHIyckBADGhJiIEBgaibdu22Lt3LyfUjDHGGGOswVllTzUAJCUlISQkBEBFkq1SqbBw4UKkpKRg/fr14n56vR4ajUacqMgYY4wxxlhDs8qeagBiQm00GqFSqQBU9EpnZmaK+yxfvhxr1qyBwWDghJoxxhhjjEnG6mvO2djYmI2fFla+WbRoEZYtW4bY2FgunccYY4wxxiRltT3VpoQRKra2tvD398eKFSsQHR2No0ePolOnThJHZ71uH9kjjE2XC6F8ohyVlJSI31vpCCvG6sTa39dyvP7J8Zpn7e8DObD236G1x2dNZJFUC73TKpUKX375JZYtW4b9+/cjPDxc4sisV1JSEl5//XW89NJLiI6OBoB6Xd/e0pKTk/HFF1/gypUrUodSYwkJCRg1ahT++OMPABWTavmi1PDKysqkDqHasrOzcfr0aZw/fx6lpaVSh1PJ5cuXsXPnTvz0009ISUkBUPG+ttZEVY7XPzld8woKCpCRkQG9Xi8+RbYmV69eRVJSktRh3JXpDZQ1/g7PnTuH9957D4B1xnc7a7kWWfdV5jaDBg0CAMTExCAiIkLiaKzXqVOn0LNnT6SkpODs2bP44YcfsHbtWvF1a0/wTp48icjISCQnJ0Ov1wOwngZzL0SE6Oho7N+/Hx9++KGsEuuLFy/ik08+wSuvvIKdO3ciOztb6pBqLT4+Hs8//zzOnTtn9b/3U6dO4aGHHsJTTz0llgy1JqdOnUJERAQWLlyIsWPHYvTo0XjllVcAVCSq1tY25Xj9k9M1Lz4+HkOHDsXAgQPRpk0b7Nq1C4D1/F5TU1PRqlUrzJkzB2fPnpU6nCqdOXMGL774IsaOHYsXXngBaWlpVvP7Ayrejw888ADOnj2LxMREcbu1xHj+/HksXrwYL730Et58800YjUZxqLDUZJVUR0REoKCgQFwxkVWWlZWFqKgoPPfcc9i0aRO2bNkCHx8fFBUViftYcw9TRkYGRo8ejUmTJmHlypXihFXhg8baKRQKaLVatGnTBiqVCu+++y52794tvmatTp06hQceeAC//vor/vnnH4wZM0assmOt75U7OX36NB544AHY2trCxsbGqn/vSUlJGDBgAAYNGoSff/4Zc+fOxeLFi63mhiYvLw/jx4/H2LFjsXv3bqSkpOCxxx7D3r17MWzYMADWlVjL8fonp2vemTNn0K9fP4SHh2PFihXo1q0bpkyZgvLycqvpOEhNTYVarcYff/yBV155BUlJSeL/tzXEd/bsWURGRqK4uBgqlQqxsbHo1KkT/u///g83btyQOjxkZGRgxIgRePrpp/H1118jNDRUfM0arqWnTp1Cjx49cObMGZw5cwY///yzWH7ZGuIDsUbl2LFj1KZNG0pOTha3TZw4kUaOHEnjxo2jqVOnitvLy8ulCPGu9uzZQ927dyeDwUAGg4FeeeUVGjhwID3wwAP08ccfSx1etfz73/+md999lw4dOkSDBg2igQMHUmxsLL333nuUkpIidXiVXLp0iUJCQuiNN96gsrIyIiJ67733yMPDg7KzsyWOrmZu3LhBkZGRNG3aNHHb9evX6fLly6TT6SSMrDKj0UizZs2iUaNGiduKiopoyJAhdOzYMTp16hSlp6dLGCFRSkoKtWrVimJiYsRtBQUFtGnTJmrdujU98cQTEkZXmRyvf3K55pWVldHTTz9Nzz77rLjt1KlTNHLkSMrIyKAbN25QUVGRhBFWuHz5Mr3wwguUlJREXl5eNHDgQPG6K/X112g00pQpU2j06NFm26dMmUJNmzal1atXU35+vkTRVdi3bx/179+fiIgMBgNNmTKFhg0bRl26dKH169fT5cuXJYstLS2N2rZtS3PnziUiouLiYvrpp5+oc+fOFB8fL1lcpmTVU83uTavVQq/XY8OGDTAYDHjrrbfw3XffISQkBF5eXvjzzz/xwAMPALDOMYZpaWlQKpVQKpUYMGAAkpKSEBkZicjISMyYMQNz5syROsR7cnR0xG+//YbIyEi89tpr0Gq1GDZsGObNmwe1Wg3AOnpMgIpxfb/88gvCw8MxY8YM8T0RFRUFZ2dnXLt2TeIIa6a0tBQqlQrz5s1DeXk5Hn/8cQwfPhyhoaF4/vnnxUfV1kChUKCgoAAKhQKFhYUAgOjoaOzcuRMTJ07EoEGDMHXqVBw/flyyGB0dHVFWVoaYmBhxW5MmTfDoo4/iX//6F86ePYvPP/9csvhuJ8frn1yueTY2Nrh+/TqaNWsmbvv++++xa9cuPPjgg+jcuTOWLl2KjIwMCaMEvL29ceDAASiVSuzbtw9Hjx7FtGnTMGrUKEybNg2lpaWSXX8VCgV0Oh0cHBwA3Jr3sWbNGowZMwZLlizBwYMHAUj3GXHlyhVcvXoVhYWFePjhh5GcnIyePXuiVatWWLZsGT777DPJetT/+OMP+Pj44JVXXgERQa1Wo1+/fkhPT8fFixcliakSaXN6Vt/y8vLo9ddfJz8/P3r44YfJ1taWfv75Z/H1P//8k3x8fOivv/6SMMo7i4mJIRcXF1qxYgU98sgjZnfFW7ZsIRsbG9q6dauEEd7b2bNnqVu3buLPAwYMII1GQ927d6d//vlHwsiq9uOPP9KyZcvMtuXl5ZGXlxft2bNHoqhq59SpU+Tm5kYnT56k8ePH08CBA+mPP/6gtWvX0ujRo6lr16508OBBqcMULVu2jFxdXWnixIk0ceJEsrOzo59++olycnJox44d1L17d1q+fLlk8RUXF9MzzzxDgwcPppMnT5q9ptPp6NFHH6WnnnpKougqk+P1T07XvCeffJKaN29Oq1atomnTppG9vT2tX7+ezp07RytWrKCgoCBJYzUYDERENGTIENq4cSMRVTy9sre3Jzs7O/rll18ki03wyiuvUEhIiPhzcXGx+P3o0aMpJCSESktLpQiNiIh27dpFwcHBtHv3bho5ciRdvXpVfO29994jPz8/OnXqlCSxHTx4kD7//HPx57KyMjIYDNS6dWuzdi4Q3g8NiZPqRig/P58uXLhAf//9N7Vv356uX78uvnb06FEKDg6mY8eOSRhh1YxGI924cYMmTJhAYWFh1LFjR7PXCgoKKCwsjFavXi1hlPdWXl5Offr0odTUVBo/fjz5+vrSZ599RiNGjKCuXbvS33//LXWId2Q0GomoImEKDg6mffv2ia9t376dzp07J1Vo1ZKdnU19+/al6OhoGjFiBB0/flx87X//+x/17duXPvzwQwkjrCD8nomI3nzzTVq8eDENHjyYZs2aZbbfY489RoMGDWro8MycOnWKvL29acyYMWbDKoiIVq5cSeHh4VY1tEZO1z+5XfMMBgONGjWKnn/+eWrfvn2lG7727dvTiy++KFF0tyxYsIDeeecdIiKaNGkSeXl5kbOzMw0fPpwSExMlje3KlSvUsmVLs5tRvV5PREQJCQnk4+Mj+WdEZGQkOTo6UuvWrc2SaiKioKAgevvttyWJy/Rmw3T4VteuXcWbKCKidevWSTa8yzqef7F65ejoiMDAQPj5+UGtVpvN3v3111/RpEkT+Pn5SRhh1RQKBVxcXDB8+HAUFhbi1KlT4uN6hUKBJk2awNXVFfb29hJHemdEBIPBACJCjx498Ndff2Hbtm2YOnUqXnjhBQQFBaFFixZShymi2x4xCpONbGxsoNFooNFoAADz5s3DM888Y9W/ewBwc3ND165dMXfuXGzbts2sPF23bt3g7u6Ov/76S7oAb1IoFGJJrYULF2LJkiVo1qwZ3N3dAdyaHOro6Ii2bdtKVr/YaDSiffv2+PXXX7Ft2zbMmzcPe/fuFV8/c+YMmjVrZlULcMnp+mfN17zk5GR88MEHeP3117Fjxw6kp6dDqVRi8+bN+Pzzz+Hn54fWrVsDqKjLX1ZWBn9/f3GbpeXm5pqtBwDcup55eXkhPj4eU6ZMwbZt23Ds2DGcOnUKW7duxcKFCxus3GZqaio2bNiAd999F8eOHQMAeHh4YP78+Th58iQmTZoEAOJwEJVKBY1G02D/36bxmQ4z++CDDxAaGorMzEyzYRUlJSVo3rw5goKCGiS+8+fP4+2338YzzzyDH3/8EQaDQXzNtNpHcXGx+P3ixYvx/PPPIzk5uUFirESSVJ7VWXl5eaVHG7ffmV27do0iIiLo4YcfpjFjxtBzzz1Hrq6uFBsb24CRVp9p792PP/5I7dq1o2bNmtH//d//0d9//03z5s2jpk2b0oULFySM8hbTeG+3YcMG6tatGx09etRse2FhoaXDqhcFBQUUEBBABw4coMWLF5NGo6HDhw9LHdZdmb7/J0+eTAqFgl5//XW6du2auH3ixIk0f/58KcK7p9mzZ5Ovry+dOnWKYmNjacmSJeTu7k4JCQkWP/fdrifC9qNHj1Lnzp0pPDycOnXqRI899hg5OTlRXFycxeOrSbwCOVz/rPWad+rUKXJ1daXevXtTt27dSK1W09ixY+n3338X9xkyZAj17duXjEYjXblyhd566y3y8fFpkKdZ8fHx5OrqSgsWLDB7Hwi/z9jYWHJzc6PAwECzp1UXL16kM2fOWDw+IqKTJ09SUFAQde/enUJCQkilUtG2bduIqGJIyieffEKtWrWi/v37U2JiIp0+fZoWLVpEzZs3pytXrkgS3/bt24mooud869atFBoaSoGBgfT999/T7t27aeHChdS0aVM6f/58g8TXtGlTeuSRR2jgwIGkUCho06ZNZvuUlpZSaWkptWrVin7//XdatWoVOTg4SPokipNqGYqPj6enn36a+vfvT1OmTDEbwyZcYISLS0JCAk2ZMoUGDx5MkydPbpAP6Hu52zgn0w/GP/74gyZPnkz29vbUoUMH6tixo9kFUgqFhYWUn59PeXl5d92vtLSUbty4If58twS8oWRnZ1NiYiKdO3eOSkpK7rpvYWEhderUiXr27ElqtbrSzYG1Et4/paWl9Oyzz5KDgwONGzeO3nrrLXr55ZfJxcXFamaJC4T3xo0bN+jhhx8mpVJJrVu3pvbt2zdIAlid64nwZ0pKCm3ZsoWmTZtG7733niSP0uV4/ZPTNU+v19OwYcNo+vTpYtw7duyggQMHUr9+/cTEZv/+/RQUFEQODg7UuXNnCgoKapBYr1y5Ql26dKGOHTuSvb09LVy4sNLvNzc3lz7++GOzti5UNmoIFy5coICAAJo3bx7l5+dTUVERzZo1i0JCQsThSIWFhbR7926KjIwkd3d3Cg4OpqCgoAZJCO8UX6tWrcT4DAYDJScn07BhwygoKIhatmxJ4eHhDfJ/nJycTH5+fjR//nxxyMe4ceNo/vz5VX6W9uvXj1q2bEkODg505MgRi8d3N5xUy8yZM2fI2dmZnnrqKZo3bx516tSJIiIiaMaMGeI+QsIkXKyF8VpSTn4QnD17llasWHHXUmG3X/yuXLlCmZmZlJOTY+nw7io+Pp4GDhxIYWFh5OvrSxs2bCAi84T5Xr1nUjl16hSFhYVRhw4dSK1W01tvvXXHD3qj0UjXr18nX19fcnNzoxMnTjRwtPdW3QkoH330EY0dO5bCwsJo9OjRDf5vSUpKooULF9LcuXMrlUcT3je3f0hs3bqVDh8+TBkZGRaPrybXE2u4MZTj9U9u1zyDwUBhYWGVJi8fPHiQHn30URo8eLCY+F29epU+++wz+vXXXxukXF15eTn9+OOPNHr0aDp58iStX7+elEqlWWItvE+luvaWlpbSvHnzaPTo0WZzDfbs2UOBgYFVlindv38/nTx5skHafG3iS0pKorS0tAYpsVpSUkJvvPEGvfTSS2bxPf300zRq1CgaOHAgvfvuu/S///2PiCr+vyMiIkilUlWaTC0FTqplxGg00r/+9S8aM2aMuC0/P5+WLVtGnTt3phdeeMFs/19++cXs0bfUH4pJSUnk5uZGCoWC3njjDbMJRIKqYrSGxDQ+Pp7c3d1p5syZtHHjRpo1axapVKo79iT++uuvlJmZ2bBB3oEQ+5w5cyg+Pp5WrFhBCoWCUlNTxX2q+h1/8MEHdPr06YYM9Z5MJ83cKbEuLy83ex+VlZVRUVGR2Sz7hnD69GlycnKiQYMGUd++fcnZ2Zl69OhBf/75p5hEmf7eG7rGb12vJw1Njtc/uV3zjEYj6XQ6euSRR8QbFdN2tm/fPrM6wVJISkqiHTt2iD9/++23YmJtenMi5efd+vXrad68eWbbsrOzycfHh06dOlXpBqChWXt8cXFxdODAAfHnpUuXkkqlounTp9PMmTMpODiYxo0bR7m5uWQ0GunXX39tkCEp1cFJtcw8++yz1KdPH7Nt+fn5tGLFCoqIiBBnY2/dupWaNWtG8+fPt4qktLCwkJ577jl69tln6dNPPyWFQkGvvfZalR8yRETR0dH05ptvNnCUVcvOzqaBAwfSK6+8Yra9X79+NH36dCIyv/j8/vvvVvO7v379OvXp04deffVVcZvRaKTBgwdTTEwMxcbGUlpamvjaRx99RF999ZUEkd5bYmIiqVQqGj58uLjtbj3Wd3pvNYTi4mJ67LHHxESvtLSUrl27Rl26dKHw8HD6/fffzd4bs2bNotmzZzf4mHu5XU/kFK+cr3lr1qwhOzs72rlzJxGZJ/mfffYZOTo6WkWngRDX7T3WpaWltH79esmHC5oSnv6ZDkk5evQoFRQUSBjVLdYa3+XLl2nMmDHieHSiihtmhUIhdmpJ3WFoipNqmRDeNB9//DH16tWr0mSLnJwceuGFF6hnz57i489FixZZzd2bXq+nTz/9lH744QciqpiUc6cPmezsbHryySepW7duVrGi39WrVykyMlIsLydcyCdOnEhPP/10lX9n4cKFVvG7z8rKonfeecds8tCbb75JCoWCOnfuTM2aNaNBgwbRP//8Q9nZ2dS1a1caPHjwPceMN7T09HTq2bMnde/enYKCgmjkyJHia1Ul1rNmzaLRo0fTpUuXGjJMM/3796dFixYR0a3H+zqdjh544AEKCwsz+z9ZtWoVubm5NViiIrfridziJZLnNc80OXn++efJ0dGR9u/fb7bPrl27qEOHDlZxbTYlJNYLFiygF154gZycnCRt/0S3fp8Gg4GuXLlCAQEB4jCZ119/ndzc3CS9+bf2+IgqPm+FzyPhWr9//37q2LGj5P+/VeGkWmaSk5PJw8ODnnvuOfEOUmgYqamppFAozGZoW5Pbe+F++OEHUigUNGfOHMrKyiKiikZz48YNys7OlnyJZlOmCZAwNnPBggU0fvx4s/1MJydaC9Nlb7///ntSKBT0448/UnZ2Nv3999/UtWtXWrx4MRFVzLiWeinfqvz73/+mkSNH0l9//UWbN2+mFi1amCXWt49J3bx5M7m5uTXILPqqlJeX04MPPmi2jLeQ7BUVFVGLFi3oySefNPs7Urx35HY9kVu8crjm3amXLysri55++mnSaDT07bff0sWLF8lgMNDs2bOpU6dODfZ+rUkv5LfffksKhYJcXFwadHJ1dWIUeoIvXrxICxcuJK1WS4cOHWqA6OQf3+2vv/7669S3b1+r/LzlpFqG/vzzT1Kr1fTyyy+b3UVmZGRQp06dKCYmRsLo7s1gMIiNREjyXnvtNbpy5QrNmDGDRowY0eDjX6vL9DHo/PnzzRbmeOedd2jlypUNOsu8pi5dulRpdvnQoUNp6NChVvUI7XY6nY5+/fVXIqpIoDdt2lQpsRbeV8L/kVSPLYXf459//klarZZWrVolviZMmvv999/Jz8+Pzpw5I/nvXW7XE7nFS2Sd17zqJCR6vZ7mzZtHbm5uFBAQQBEREeTu7t4gwypqmjCVlJTQ1KlTydnZucGqvNQkxtzcXAoNDaWhQ4eSnZ1dgyT9jSk+ooob5wULFpCzs7NVTEqsCifVMvXbb7+RWq2mkSNH0g8//EAJCQliTVPTMbLWyjT5+eGHH0ilUlHr1q3J1tbWqsbBVUX4cJw/fz4NGTKEiCqGeygUCklq9tZWeXk5FRUV0ZNPPinZClnVUdWY2KKiIvrpp58qJdZffPGFODRA6mQ1Pz+f5s+fT4GBgZVWxNu9eze1bNnSbElqKcnteiK3eIms65oXGxtLvXv3rnY1nJiYGPrpp59o48aNdPHiRcsGRzWPz2g00p49e8jX17fB6unXJMby8nI6f/48KRQK0mq1DfI50ZjiIyI6ceIEDRkyhFq3bm1VteZvx0m1jB07doz69u1LzZs3p5YtW1KrVq2sPiE1ZTQaxcTnoYceIjc3N6u9+zQlfDAuXryYXnzxRXr//fdJrVZbzdLHNbFw4UIKCAiw+uXHq6LX68WhIKNGjaJXX32VFAqFVSwOJDytSE5OplmzZpGPjw8tWLCA8vLyKDs7mxYsWEDt27cXhwBYA7ldT+QWL5F1XPPi4uJIpVLRa6+9Vum12xf8keLGtCbxmUpPT2+w6jS1jXH58uUN8v/dWOPbs2ePVY6jNsVJtczl5eXRxYsX6eTJk5JPKKgNg8FAM2fOJIVCYZX1kO9m2bJlpFAoyNnZWfKC8zW1adMmevnllxvsUW5d3T6kRviwLyoqEsepurm5WcWNjfBhcPHiRdq0aROlpKTQJ598Qs7OzuTv70/t2rUjHx8fq4j1dnK7nsgtXiJpr3mnT58mBwcHcQKt0Wik7OzsO96IrlmzpkHHJlt7fES1i7EhPx8aW3yfffaZrD5fOalmkjIYDPTVV19Z9eOcOzly5AgpFAqrW6GvOk6fPk1jxoyxihU278U0Sf3mm28qvf7888+TVqu1iv8HIfm/ePEi2dra0oQJE8TXrly5Qt9//z39/vvvVjkZlDUMqa55WVlZFBwcTGFhYeK2iRMnUpcuXahp06bUp08fio2NFW9Yr127RgqFgh577LEGGe9t7fHJIUaOT3oKIiIwJiEigkKhkDqMWtHpdNBqtVKHUStlZWVQqVRSh3FXBoMBtra2uHTpElq3bo2nnnoK3377rfj69u3bMWPGDHz//ffo0qWLhJGaxxoeHo7HH38ca9euhUqlgtFohI2NjaTxMesh1TVv+vTpiIuLQ//+/bF9+3a4u7tj5MiR8PT0RHR0NNLT07Fnzx4EBwcDAM6dOwciQuvWrTk+mcTI8UlMsnSeMcbuwrTX19XVlZ577rlKw0D0en2DLO17L7fH+swzz1h1FRh2f7l9oSFvb28aOnSo2QqlRETt2rWjZ555hogqD7m6n+Mjsv4YOT7rYCt1Us8YY7e7vdf30Ucfxeeffw5b21uXLKPRCAcHBzg4OEgYKVBeXl4p1q+++sosVsakoNPpYDQaQURwcnICAKxcuRK+vr4IDAyEl5cXgIr3sFKpRJs2baDT6QCgQd6/1h6fHGLk+KwLP49kjFmV6iap1jKcQqlUIiUlBe3atcOIESOwbt06WX4YsMYlISEBI0eORN++fREaGoqNGzeivLwcADB79mwMGzZMHIKiVCrFISlt27YFUDFE5X6OTw4xcnxWSILeccYYu6tLly6RRqOhiRMnVllayZoYDAZ67rnnaNKkSbJ8XMkan/j4eHJ3d6eZM2fSxo0badasWaRSqe44ObKsrIwWLFhATZs2paSkpPs+PjnEyPFZJ06qGWNWRY5Jak5OTpWL1DDW0LKzs2ngwIH0yiuvmG3v168fTZ8+nYjM60/v2rWLhg8fTj4+Pg1SXtPa45NDjByf9bKO56eMMXaTUqnEihUr8MUXX8hmGIWrq6vVDEdh97eysjLk5uZi9OjRACrmHgBAYGAgcnJyAEB85E5ECAwMRNu2bbF3716EhYXd9/HJIUaOz3pxST3GGGOsEUlKSkJISAiAW6UzFy5ciJSUFKxfv17cT6/XQ6PRiJPEOD75xMjxWSfuWmGMMcYaESGZMRqNYi16IkJmZqa4z/Lly7FmzRoYDIYGT2asPT45xMjxWSd5PFtljDHGWI3Y2NiYLTQjDFFatGgRli1bhtjYWEmHWFl7fEJM1hwjx2dduKeaMcYYa6SEEZ62trbw9/fHihUrEB0djaNHj6JTp04SR2f98QHWHyPHZz0az+0BY4wxxswIPYMqlQpffvklnJycsH//foSHh0scWQVrjw+w/hg5PuvBPdWMMcZYIzdo0CAAQExMDCIiIiSOpjJrjw+w/hg5Pulx9Q/GGGPsPqDT6aDVaqUO446sPT7A+mPk+KTFSTVjjDHGGGN1xMM/GGOMMcYYqyNOqhljjDHGGKsjTqoZY4wxxhirI06qGWOMMcYYqyNOqhljjDHGGKsjTqoZY4wxxhirI06qGWOMMcYYqyNOqhljjDHGGKsjTqoZY4wxxhirI06qGWOMMcYYqyNOqhljjDHGGKsjTqoZY4wxxhirI06qWaPQokULfPjhh3d8/dKlS1AoFIiLi2uwmBhjjDF2/+Ckmt0X/P39kZGRgfbt24vbUlNTMXToUGg0Gnh5eeG1116DwWCQMErGrMvBgwehVCoxdOhQqUNhjDGrx0k1uy8olUr4+PjA1tYWAFBeXo6hQ4eitLQUMTEx+Pbbb/HNN99g0aJFEkfKmPVYt24dpk+fjn379iE9PV3qcBhjzKpxUs1koV+/fpg2bRqmTZsGZ2dneHh4YOHChSAicR+9Xo/nnnsOjo6OCAgIwBdffCG+dvvwj127diEhIQEbNmxA586dMWTIELz11lv49NNPUVpa2tD/PMasTmFhIX788UdMnToVQ4cOxTfffCN1SIw1iJKSErzyyivw8vKCvb09evfujSNHjgAA/vrrLygUCuzcuRNhYWFwcHDAQw89hMzMTOzYsQOhoaFwcnLCuHHjoNfrxWMajUYsX74cgYGBcHBwQKdOnbB582az8/72228ICQmBvb09HnzwQXz77bdQKBTIzc0FAGRnZ2Ps2LHw8/ODRqNBhw4d8P333zfY74XdGyfVTDa+/fZb2Nra4vDhw/joo4+watUqfPXVV+LrK1euREREBGJjY/HSSy9h6tSpOHv2bJXHOnjwIDp06ABvb29x26BBg5Cfn4/4+HiL/1sYs3abNm1CmzZt0Lp1a0RFReHrr782u4llrLF6/fXX8fPPP+Pbb7/F8ePHERwcjEGDBiEnJ0fcZ8mSJfjkk08QExODtLQ0jBkzBh9++CH+/e9/Y9u2bdi1axdWr14t7r98+XKsX78ea9euRXx8PGbOnImoqCj8/fffAICLFy9i9OjRGDFiBE6cOIHJkydj/vz5ZnEVFxejS5cu2LZtG06fPo0XX3wR48ePx+HDhxvmF8PujRiTgb59+1JoaCgZjUZx29y5cyk0NJSIiJo3b05RUVHia0ajkby8vGjNmjVERHTx4kUCQLGxsURE9MILL9DAgQPNzqHT6QgAbd++3cL/GsasX8+ePenDDz8kIqKysjLy8PCgvXv3ShsUYxZWWFhIKpWKNm7cKG4rLS0lX19fio6Opr179xIA2rNnj/j68uXLCQCdP39e3DZ58mQaNGgQEREVFxeTRqOhmJgYs3NNmjSJxo4dS0QVn2ft27c3e33+/PkEgG7cuHHHeIcOHUqzZ8+u9b+X1S/uqWay0b17dygUCvHnHj16ICkpCeXl5QCAjh07iq8pFAr4+PggMzOzweNkTO7Onj2Lw4cPY+zYsQAAW1tbPPnkk1i3bp3EkTFmWefPn0dZWRl69eolblOpVIiMjERiYqK4zfTzxtvbGxqNBkFBQWbbhM+f5ORk6PV6PPzww2jSpIn4tX79epw/fx5ARZvr2rWrWSyRkZFmP5eXl+Ott95Chw4d4ObmhiZNmmDnzp1ITU2tv18AqxNbqQNgrL6oVCqznxUKBYxGY5X7+vj4VHpkdu3aNfE1xu5n69atg8FggK+vr7iNiKBWq/HJJ5/A2dlZwugYk57p541Cobjr509hYSEAYNu2bfDz8zPbT61WV/uc77//Pj766CN8+OGH6NChA7RaLWbMmMHzgKwI91Qz2Th06JDZz//73/8QEhICpVJZ42P16NEDp06dMuvJ3r17N5ycnNC2bds6x8qYXBkMBqxfvx4rV65EXFyc+HXixAn4+vryxCjWqLVs2RJ2dnY4cOCAuK2srAxHjhyp9WdD27ZtoVarkZqaiuDgYLMvf39/AEDr1q1x9OhRs78nTI4UHDhwAI899hiioqLQqVMnBAUF4dy5c7WKiVkG91Qz2UhNTcWsWbMwefJkHD9+HKtXr8bKlStrdayBAweibdu2GD9+PKKjo3H16lUsWLAAL7/8co16DhhrbLZu3YobN25g0qRJlXqkR40ahXXr1mHKlCkSRceYZWm1WkydOhWvvfYa3NzcEBAQgOjoaOj1ekyaNAknTpyo8TEdHR0xZ84czJw5E0ajEb1790ZeXh4OHDgAJycnPPPMM5g8eTJWrVqFuXPnYtKkSYiLixMr7gjDHkNCQrB582bExMTA1dUVq1atwrVr17gjyIpwTzWTjQkTJqCoqAiRkZF4+eWX8eqrr+LFF1+s1bGUSiW2bt0KpVKJHj16ICoqChMmTMCbb75Zz1EzJi/r1q3DgAEDqhziMWrUKBw9ehQnT56UIDLGGsa7776LUaNGYfz48QgPD0dycjJ27twJV1fXWh/zrbfewsKFC7F8+XKEhoZi8ODB2LZtGwIDAwEAgYGB2Lx5M7Zs2YKOHTtizZo1YvUPoaNnwYIFCA8Px6BBg9CvXz/4+PhgxIgRdf73svqjIOIaScz69evXD507d77rUuSMMcZYY/H2229j7dq1SEtLkzoUVk08/IMxxhhjTGKfffYZunbtCnd3dxw4cADvv/8+pk2bJnVYrAY4qWaMMcYYk1hSUhKWLVuGnJwcBAQEYPbs2XjjjTekDovVAA//YIwxxhhjrI54oiJjjDHGGGN1xEk1Y4wxxhhjdcRJNWOMMcYYY3XESTVjjDHGGGN1xEk1Y4wxxhhjdcRJNWOMMcYYY3XESTVjjDHGGGN1xEk1Y4wxxhhjdcRJNWOMMcYYY3X0/9qqFVfi2ci4AAAAAElFTkSuQmCC",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"y_pred, weights = engine.predict(x_obs, x_err=np.array([0.1]*50),y_true=y_true, n_samples=10000, corner=True, corner_reweight=True,seed=0)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"@webio": {
"lastCommId": null,
"lastKernelId": null
},
"kernelspec": {
"display_name": "py311-arm",
"language": "python",
"name": "py311-arm"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.11.4"
}
},
"nbformat": 4,
"nbformat_minor": 4
}