mooc_ecmwf_bocquet_brajard_part2.ipynb

{"cells":[{"cell_type":"code","execution_count":null,"id":"a53802f5-6ba5-45ee-8404-efd7065a4133","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":17,"resources":{"http://localhost:8080/custom.css":{"data":"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","headers":[["content-length","1449"],["content-type","text/html; charset=utf-8"]],"ok":false,"status":404,"status_text":""}}},"id":"a53802f5-6ba5-45ee-8404-efd7065a4133","outputId":"e11e2926-fde4-4a19-bc6e-9e28b95678f3"},"outputs":[{"data":{"text/html":["<link rel=\"stylesheet\" href=\"custom.css\">\n"],"text/plain":["<IPython.core.display.HTML object>"]},"metadata":{},"output_type":"display_data"}],"source":["%%HTML\n","<link rel=\"stylesheet\" href=\"custom.css\">"]},{"cell_type":"markdown","id":"0gdhCwxKgz59","metadata":{"id":"0gdhCwxKgz59"},"source":["## <span style=\"color:blue\"> Before you start, if you run on colab </span>...\n","On colab, each notebook runs on its own environment, so you first need to run the following cell in order to download all the repository and to run the notebook part1.\n","\n","**Advice:** If it is not done, don't forget to change the Runtime type into \"GPU\" so to speed up the computations.\n","\n","## <span style=\"color:blue\"> Before you start, if you run on kaggle </span>...\n","You first need to **turn internet on** (available in the Settings in the right panel). To do that you need to have a verified kaggle account.\n","On kaggle, each notebook runs on its own environment, so you first need to run the following cell in order to download all the repository and to run the notebook part1.\n","\n","\n","**Warning:** The cell below re-run the notebook part 1, it can take a couple of minutes.\n"]},{"cell_type":"code","execution_count":null,"id":"BI6pdXhCg0Ye","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":1000,"referenced_widgets":["85a3c4eb85ed4672a8ee46a037372300","4a2e89f1f6bf4eedbca0ad25d5a04de0","63ab33992e6747789c6aa30d7fd3ca7b","ddb4dc44dafc4f7ab81ee0be4bd36d29","12762e6f19e44be9b4824d54b97acd15","2403d9c27b92416595809f8851b7e786","69b622f0565f42e8a16844d3d6fef28b","125c84f09fcb4666b79236f179d19827","ec55d4f2783d4264a14534273656a4a5","b0e7cc4350bf4ed3ac85b424eabab9f8","3f80dd8cc17842b0b542dffa7d0b1f8b","835d070e8f344cfd92b79773ba7023b5","ffb331998a1d41cd8132263b449c68b9","5c276d02f53b4cb88de8b336dc660cfd","23ad55be58a045ea82d6aa18519d6739","10bd83a61fce44e9b951d9f72cee9c1e","320eba1b9795450db713503c124ef4c8","35b724c3fb1f4517ae9e33f151b578e9","0ca41fb9aee44015b0e1112daa393634","da7f410b46fa4574adf21b5b7c1be74d","21bd22f8052b49b98acbb98a922d7d52","09ba5482356046c1905faafcf7f77c9d","2b2a5f4feec34418a815cbaddac8c9e6","19c75562bce7421e8da187e97c4b08c3","e48505496f2142ecac7cc2f43105dfe2","8153b322cbc9499a9444429bb39e797c","0dfef0c43c87493d803b6849ecf7fa11","98ccbf5dfc0e48d29b21891e60ab6c4c","6fa25cae71b9497da2a408cbb2e44ccf","2265c62e2c654e25a51d7bba6d6744f9","6887d263554c4ce2a3e629693d29825b","f966d3ecfee44e5cacdedcd292c4c6b0","d6c6b760b1cc472dbd2df015c2c3efa9","ea90e17d28754b61970391b99450da70","9de6ad16dac748058be999dde9ffd367","979bb47d6beb4f3ba39930094adabb8e","f0c6e5f86cd4441e8c299e4bd01515fc","9315aadffe264b1a89bddc22ed186249","ab611e839b6e45c985e821d7af6d17fb","89b15c11862e4bfda72fcff7ea1e92d1","cf8df982c4134fe39681a3d9906fadc8","3acf1cac85e9490483a4ef865289b917","ada68873de254113a3534c1154e0347b","d6dc2ba5f0a14b248c56e3fe88e94c3c","5e313cac6dd94aac9ec6e55789cc5136","9fb61e3b0e6949aea6f4a6b7154f4ce0","95a89ac1b6924365b4bccff08a482f49","f8cab4150ab040aeb033f64745faad74","46df280ab0e1494aada1d25fd0c6cd93","a4df1248688a4522b24d39639de0c3dd","d25593e56677436da0eb4d72a3c18f4e","db781734b9984ade9318d80c0a70eb5d","608345d3c72442538d04f01665b690e9","675634702dbb4e31ba310b6cfb154ab6","c0e73a04cfd24537921109d7d648dbf5"],"resources":{"http://localhost:8080/custom.css":{"data":"CjwhRE9DVFlQRSBodG1sPgo8aHRtbCBsYW5nPWVuPgogIDxtZXRhIGNoYXJzZXQ9dXRmLTg+CiAgPG1ldGEgbmFtZT12aWV3cG9ydCBjb250ZW50PSJpbml0aWFsLXNjYWxlPTEsIG1pbmltdW0tc2NhbGU9MSwgd2lkdGg9ZGV2aWNlLXdpZHRoIj4KICA8dGl0bGU+RXJyb3IgNDA0IChOb3QgRm91bmQpISExPC90aXRsZT4KICA8c3R5bGU+CiAgICAqe21hcmdpbjowO3BhZGRpbmc6MH1odG1sLGNvZGV7Zm9udDoxNXB4LzIycHggYXJpYWwsc2Fucy1zZXJpZn1odG1se2JhY2tncm91bmQ6I2ZmZjtjb2xvcjojMjIyO3BhZGRpbmc6MTVweH1ib2R5e21hcmdpbjo3JSBhdXRvIDA7bWF4LXdpZHRoOjM5MHB4O21pbi1oZWlnaHQ6MTgwcHg7cGFkZGluZzozMHB4IDAgMTVweH0qID4gYm9keXtiYWNrZ3JvdW5kOnVybCgvL3d3dy5nb29nbGUuY29tL2ltYWdlcy9lcnJvcnMvcm9ib3QucG5nKSAxMDAlIDVweCBuby1yZXBlYXQ7cGFkZGluZy1yaWdodDoyMDVweH1we21hcmdpbjoxMXB4IDAgMjJweDtvdmVyZmxvdzpoaWRkZW59aW5ze2NvbG9yOiM3Nzc7dGV4dC1kZWNvcmF0aW9uOm5vbmV9YSBpbWd7Ym9yZGVyOjB9QG1lZGlhIHNjcmVlbiBhbmQgKG1heC13aWR0aDo3NzJweCl7Ym9keXtiYWNrZ3JvdW5kOm5vbmU7bWFyZ2luLXRvcDowO21heC13aWR0aDpub25lO3BhZGRpbmctcmlnaHQ6MH19I2xvZ297YmFja2dyb3VuZDp1cmwoLy93d3cuZ29vZ2xlLmNvbS9pbWFnZXMvbG9nb3MvZXJyb3JwYWdlL2Vycm9yX2xvZ28tMTUweDU0LnBuZykgbm8tcmVwZWF0O21hcmdpbi1sZWZ0Oi01cHh9QG1lZGlhIG9ubHkgc2NyZWVuIGFuZCAobWluLXJlc29sdXRpb246MTkyZHBpKXsjbG9nb3tiYWNrZ3JvdW5kOnVybCgvL3d3dy5nb29nbGUuY29tL2ltYWdlcy9sb2dvcy9lcnJvcnBhZ2UvZXJyb3JfbG9nby0xNTB4NTQtMngucG5nKSBuby1yZXBlYXQgMCUgMCUvMTAwJSAxMDAlOy1tb3otYm9yZGVyLWltYWdlOnVybCgvL3d3dy5nb29nbGUuY29tL2ltYWdlcy9sb2dvcy9lcnJvcnBhZ2UvZXJyb3JfbG9nby0xNTB4NTQtMngucG5nKSAwfX1AbWVkaWEgb25seSBzY3JlZW4gYW5kICgtd2Via2l0LW1pbi1kZXZpY2UtcGl4ZWwtcmF0aW86Mil7I2xvZ297YmFja2dyb3VuZDp1cmwoLy93d3cuZ29vZ2xlLmNvbS9pbWFnZXMvbG9nb3MvZXJyb3JwYWdlL2Vycm9yX2xvZ28tMTUweDU0LTJ4LnBuZykgbm8tcmVwZWF0Oy13ZWJraXQtYmFja2dyb3VuZC1zaXplOjEwMCUgMTAwJX19I2xvZ297ZGlzcGxheTppbmxpbmUtYmxvY2s7aGVpZ2h0OjU0cHg7d2lkdGg6MTUwcHh9CiAgPC9zdHlsZT4KICA8YSBocmVmPS8vd3d3Lmdvb2dsZS5jb20vPjxzcGFuIGlkPWxvZ28gYXJpYS1sYWJlbD1Hb29nbGU+PC9zcGFuPjwvYT4KICA8cD48Yj40MDQuPC9iPiA8aW5zPlRoYXTigJlzIGFuIGVycm9yLjwvaW5zPgogIDxwPiAgPGlucz5UaGF04oCZcyBhbGwgd2Uga25vdy48L2lucz4K","headers":[["content-length","1449"],["content-type","text/html; charset=utf-8"]],"ok":false,"status":404,"status_text":""}}},"id":"BI6pdXhCg0Ye","outputId":"362b3fd9-93ee-43f3-a936-e2e391fff94b"},"outputs":[{"name":"stdout","output_type":"stream","text":["Kaggle detected\n"]}],"source":["# First run configuration for colab & kaggle\n","import sys, os\n","IN_COLAB = 'google.colab' in sys.modules\n","IN_KAGGLE = os.environ.get('KAGGLE_KERNEL_RUN_TYPE','Localhost') != ''\n","\n","if IN_COLAB or IN_KAGGLE:\n","    print('Colab detected' if IN_COLAB else 'Kaggle detected')\n","    \n","    if os.path.isdir('2022-mooc-ecmwf-bocquet-brajard'):\n","      %cd 2022-mooc-ecmwf-bocquet-brajard/\n","    if not os.path.isfile('mooc_ecmwf_bocquet_brajard_part1.ipynb'):\n","      \n","      # Clone the git repository\n","      !git clone https://github.com/marcbocquet/2022-mooc-ecmwf-bocquet-brajard.git\n","\n","      # Make the repository as working directory\n","      %cd 2022-mooc-ecmwf-bocquet-brajard/\n","\n","    if not os.path.isfile('Output/xt.npy'):\n","\n","      # Run the notebook part 1\n","      %run mooc_ecmwf_bocquet_brajard_part1.ipynb"]},{"cell_type":"markdown","id":"923d2d94","metadata":{"id":"923d2d94"},"source":["# <span style=\"color:maroon\"> Data assimilation, machine learning, and dynamical systems - Part II  </span>\n","#### Marc Bocquet¹ [marc.bocquet@enpc.fr](mailto:marc.bocquet@enpc.fr) and Julien Brajard² [julien.brajard@nersc.no](mailto:julien.brajard@nersc.no)\n","#### (1) CEREA, École des Ponts et EdF R&D, Île-de-France, France\n","#### (2) Nansen Center (NERSC), Bergen, Norway\n","\n","During this session, we will discover some connections between data assimilation and deep learning when applied to dynamical systems."]},{"cell_type":"markdown","id":"225f4f08-bba5-4b5d-b0c2-10c8b4d9cda8","metadata":{"id":"225f4f08-bba5-4b5d-b0c2-10c8b4d9cda8"},"source":["# <span style=\"color:teal\"> Second step: Learn the model error through machine learning.</span>\n","\n","In this notebook, we will use the output of data assimilation to train a neural network into predicting the model error. The procedure is the following:\n","1. The **analysis** $x^{\\rm a}_k$ is used as the best possible estimate from the truth at time step $t_k$.\n","2. The physical model is used to produced a **forecast** $x^{\\rm f}_{k}$ using the analysis $x^{\\rm a}_{k-1}$ as initial conditions.\n","3. For each time $t_k$, the difference between the forecast and the analysis is used to estimate the **model error** $\\mathrm{err}_k = x^{\\rm a}_k - x^{\\rm f}_k$. Note that this value is called **analysis increment** in the context of data assimilation.\n","4. A **neural network** taking $x^{\\rm a}_{k-1}$ as an input is trained, supervised by the data $\\mathrm{err}_k$\n","<p align = \"center\">\n","<img src=\"https://github.com/marcbocquet/2022-mooc-ecmwf-bocquet-brajard/blob/colab/Figures/scheme-err.png?raw=1\" alt=\"Model error estimation\" width=\"600\"/>\n","</p>\n","\n","## <span style=\"color:blue\"> I. Check the result of data assimilation </span>\n","Before training the neural network, as it is recommended for *any* machine learning process, we have a look at the data.\n"]},{"cell_type":"code","execution_count":null,"id":"2a567a4f-5cfc-4dbc-b2e7-9827fc7efd99","metadata":{"id":"2a567a4f-5cfc-4dbc-b2e7-9827fc7efd99"},"outputs":[],"source":["# Import standard modules\n","import pathlib\n","import json\n","import seaborn as sns\n","from matplotlib import pyplot as plt\n","import numpy as np\n","from tqdm.notebook import trange, tqdm\n","from pickle import dump\n","\n","\n","from tensorflow.keras.layers import Input, Conv1D, BatchNormalization\n","from tensorflow.keras import regularizers\n","from tensorflow.keras import Model\n","import tensorflow as tf\n","from utils import tqdm_callback\n","\n","from sklearn.preprocessing import StandardScaler\n","from sklearn.metrics import r2_score, mean_squared_error\n","\n","from model.neuralnet import Periodic1DPadding\n","\n","# Load dataset lengths\n","with open('./Nt.cfg', 'r') as f:\n","    Nt = json.load(f)\n","\n","# Directory for results\n","outputdir = pathlib.Path(\"./Output\")\n","\n","# For plot customisation\n","sns.set_context('notebook')\n","sns.set_style('darkgrid')\n","plt.rc('axes', linewidth=1)\n","plt.rc('axes', edgecolor='k')\n","plt.rc('figure', dpi=100)\n","palette = sns.color_palette('deep')"]},{"cell_type":"markdown","id":"3c2d5483-6222-4fcb-9a75-b6351f8d5e5c","metadata":{"id":"3c2d5483-6222-4fcb-9a75-b6351f8d5e5c"},"source":["First, we load the result of the data assimilation, the so-called **analysis**. We also load the truth to plot some diagnostics."]},{"cell_type":"code","execution_count":null,"id":"2d15ac0c-0fdf-43c1-bd02-973518a375ee","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"2d15ac0c-0fdf-43c1-bd02-973518a375ee","outputId":"db42e3f9-2d98-4fed-ae6d-fe49fbd3458c"},"outputs":[{"name":"stdout","output_type":"stream","text":["Shape (number of samples, number of features) of the truth: (5000, 40)\n","Shape (number of samples, number of features) of the analysis: (5000, 40)\n"]}],"source":["# Load the analysis:\n","xa = np.load(outputdir / 'xa.npy')\n","\n","# Load the slice of the truth corresponding to the analysis\n","xt = np.load(outputdir / 'xt.npy')[:Nt['train']]\n","\n","print(f'Shape (number of samples, number of features) of the truth: {xt.shape}')\n","print(f'Shape (number of samples, number of features) of the analysis: {xa.shape}')"]},{"cell_type":"markdown","id":"f7e4b421-f901-477c-90de-de9bb237d0b2","metadata":{"id":"f7e4b421-f901-477c-90de-de9bb237d0b2"},"source":["Here we define the physical model so as to compute the forecast:"]},{"cell_type":"code","execution_count":null,"id":"84d7713e-d6a3-4e8e-a5ff-7c380034dd6b","metadata":{"id":"84d7713e-d6a3-4e8e-a5ff-7c380034dd6b"},"outputs":[],"source":["# Check that the parameters of L96 are the same as in the first notebook.\n","from model.lorenz96 import l96\n","Nx = 40\n","dt = 0.05\n","F = 8\n","phi = l96(Nx, dt, F)"]},{"cell_type":"markdown","id":"7d65b102-7814-4186-bc86-f598c3818515","metadata":{"id":"7d65b102-7814-4186-bc86-f598c3818515"},"source":["In the standard configuration, the best estimate we have from the system is the analysis. \n","\n","Here, as we know the truth, we can choose to use the true value of the L96 variable instead of the analysis.  Then, the data are **perfect**: there is no error due to the observation error, the observation sampling, and the data assimilation process. It is an ideal case, but it cannot be achieved for most realistic systems.\n","\n","If you want to do the test with perfect data, you just have to uncomment the line ```learning = 'truth'```. It is expected to be much easier to learn the model error as there is not additional noise.\n","\n"]},{"cell_type":"code","execution_count":null,"id":"03950505-f3d7-4e2c-b3a6-dcf3982528a9","metadata":{"id":"03950505-f3d7-4e2c-b3a6-dcf3982528a9"},"outputs":[],"source":["learning = 'analysis'\n","#learning = 'truth'\n","if learning == 'analysis':\n","    x = xa\n","else:\n","     x = xt #learn on the true error"]},{"cell_type":"code","execution_count":null,"id":"5fd3c94f-2112-4f09-99e7-d3984dfbaf6d","metadata":{"id":"5fd3c94f-2112-4f09-99e7-d3984dfbaf6d"},"outputs":[],"source":["# Compute the forecast of the physical model\n","xf = x + phi(x)\n","\n","# Compute the model error\n","err = x[1:] - xf[:-1]"]},{"cell_type":"markdown","id":"98480fc0-2f68-4b17-9029-5cb4efcb6d54","metadata":{"id":"98480fc0-2f68-4b17-9029-5cb4efcb6d54"},"source":["Now, we will compute and plot a diagnostic of the data assimilation. This diagnostic use the truth to evaluate the results, let's highlight that in a \"real-world\" case, the truth is not known and such diagnostics cannot be computed the same way. In such cases the truth is not known, observations can also be used, which come with additional challenges out of the scope of this module."]},{"cell_type":"code","execution_count":null,"id":"eefa853f-d86c-4d36-a112-84d86e673a71","metadata":{"id":"eefa853f-d86c-4d36-a112-84d86e673a71"},"outputs":[],"source":["# One utility function to compute scatter plots\n","def scatter_plot(x,y,ax=None):\n","    x, y = x.ravel(), y.ravel()\n","    ax = sns.regplot(x=x, y=y, ax=ax)\n","    r2 = r2_score(x, y)\n","    mse = mean_squared_error(x, y, squared=False)\n","    ax.set_title(f'R2={r2:.2f}, RMSE={mse:.2e}')\n","    return ax"]},{"cell_type":"markdown","id":"ebf0def8-6a83-4ebb-84d7-458daf41e712","metadata":{"id":"ebf0def8-6a83-4ebb-84d7-458daf41e712"},"source":["The RMSE in time is defined as\n","$$\n","\\mathrm{RMSE}(k) = \\sqrt{\\frac{1}{N_x} \\sum_{n=1}^{N_x} \\left( x^{\\rm t}_n - x^{\\rm a}_n \\right)^2},\n","$$\n","where $k$ is the index for the time, $n$ is the space index, $x^{\\rm t}_n$ is the true value and $x^{\\rm a}_n$ is the analysis (the result of the data assimilation)."]},{"cell_type":"code","execution_count":null,"id":"794e9760-f6fc-49c8-ae91-8827687a9e37","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":413},"id":"794e9760-f6fc-49c8-ae91-8827687a9e37","outputId":"22a937e7-d248-4cac-b177-64e0735a110a"},"outputs":[{"data":{"image/png":"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","text/plain":["<Figure size 640x480 with 1 Axes>"]},"metadata":{},"output_type":"display_data"}],"source":["rmse_time = np.sqrt(np.mean(np.square(xt-xa),axis=1))\n","plt.plot(dt*np.arange(rmse_time.shape[0]),rmse_time)\n","plt.plot(dt*np.arange(rmse_time.shape[0]), rmse_time.mean()*np.ones(rmse_time.shape[0]), ':k' )\n","plt.xlabel('Time');\n","plt.ylabel('Analysis RMSE');\n","plt.title(f'RMSE (mean) = {rmse_time.mean():.3e}');"]},{"cell_type":"markdown","id":"179b7fb3-c16d-4087-928a-a36ff6e5689f","metadata":{"id":"179b7fb3-c16d-4087-928a-a36ff6e5689f"},"source":["The dotted line is the mean of the RMSE, we should retrieve the same value as in part one of this session (around 0.2). We observe that, despite a couple of peaks, the RMSE is stable in time, which is a necessary condition. Indeed, in the following, we assume that the analysis is stationary, which means that its statistical properties do not depend on the time (for example there is no trend in the error).\n","\n","To have a sense of the accuracy of the analysis compared with the truth, we can also represent the analysis in the form of a scatter plot.\n"]},{"cell_type":"code","execution_count":null,"id":"7fc75043-6488-4b60-b028-59aa2ce4365c","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":413},"id":"7fc75043-6488-4b60-b028-59aa2ce4365c","outputId":"2648c54b-11d4-4dc2-df53-175355e35f9f"},"outputs":[{"data":{"image/png":"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","text/plain":["<Figure size 640x480 with 1 Axes>"]},"metadata":{},"output_type":"display_data"}],"source":["ax = scatter_plot (xt[:,20], xa[:,20])\n","ax.set_xlabel('Truth')\n","ax.set_ylabel('Analysis');\n","plt.savefig('figures/scatter_analysis')"]},{"cell_type":"markdown","id":"210e5b2f-0d47-4493-8ac6-7d279620ee2a","metadata":{"id":"210e5b2f-0d47-4493-8ac6-7d279620ee2a"},"source":["It appears clearly that the analysis is close to the true value, so the assumption that \"The analysis is the best estimate we have of the truth\" is reasonable.\n","Note that the RMSE can be a bit different as it is the total RMSE for the variable 20, chosen arbitrary and not the mean RMSE in time for all the variables. "]},{"cell_type":"markdown","id":"9202b765-6f46-4ff6-b481-bff4e95b9ea5","metadata":{"id":"9202b765-6f46-4ff6-b481-bff4e95b9ea5"},"source":["## <span style=\"color:blue\"> II. Build the neural network.</span>\n","It is now time to define the neural network. We are using the Keras API of TensorFlow https://www.tensorflow.org/api_docs/python/tf/keras, the approach is of course also applicable using any deep-learning framework such as Pytorch https://pytorch.org/"]},{"cell_type":"markdown","id":"198f78f7-01b9-415c-83f8-81df8e72c42e","metadata":{"id":"198f78f7-01b9-415c-83f8-81df8e72c42e"},"source":["We initialize random generators so the notebooks are fully reproducible. Several aspects are based on random processes:\n","- The initialization of the weights of the neural network\n","- The batch-sized chunk presented to the neural during the training are randomly shuffled at each epoch.\n"]},{"cell_type":"code","execution_count":null,"id":"7c9d2802-7a16-4184-85ed-1e7df1b2eca2","metadata":{"id":"7c9d2802-7a16-4184-85ed-1e7df1b2eca2"},"outputs":[],"source":["import random, os\n","seed = 1980\n","os.environ['PYTHONHASHSEED'] = str(seed)\n","random.seed(seed)\n","tf.random.set_seed(seed)\n","np.random.seed(seed)"]},{"cell_type":"markdown","id":"18a76687-a309-40ba-9d0d-b9aac9160504","metadata":{"id":"18a76687-a309-40ba-9d0d-b9aac9160504"},"source":["We define the function that builds the neural net model as the form of a convolutional neural network. The function takes several input arguments:\n","- `archi`: specify the hidden layers of the neural network as a form of a list of triplets. Each element of the list represents a hidden layer. A triplet is in the form of  `(nunits, kernelsize, activation)` where:\n","    - `nunits` is the number of units (neurons) on the layer\n","    - `kernelsize` is the kernel size of the convolution \n","    - `activation` is the activation function\n","- `Nx`: is the spatial size of the problem (in our case: $N_x=40$)\n","- `reg`: is the L2 weight regularization of the output layer (default=1e-5). Can be set to zeros if you don't want to use regularization\n","- `batchlayer`: Possibility to add batchnormalization before hidden layer (default is None). It is represented as a set of indexes. For example, specify `{0,2}` to have batchnormalization before the first and third hidden layers."]},{"cell_type":"code","execution_count":null,"id":"64045b57-030a-4e49-aa6d-012ae661b73a","metadata":{"id":"64045b57-030a-4e49-aa6d-012ae661b73a"},"outputs":[],"source":["def buildmodel(archi, Nx=40, reg=1e-5, batchlayer={}):\n","\n","    # Input Layer:\n","    inputs = Input(shape=(Nx,1))\n","    \n","    # Pre-processing layer to account for the periodicity of the space:\n","    border = int(np.sum(np.array([kern//2 for nfil,kern,activ in archi])))\n","    x = Periodic1DPadding(padding_size=border)(inputs)\n","    \n","    # BatchNormalization layer\n","    x = BatchNormalization()(x)\n","    \n","    # Definition of the hidden layers\n","    for i, (nfil, kern, activ) in enumerate(archi):\n","        if i in batchlayer:\n","            x = BatchNormalization()(x)\n","        x = Conv1D(nfil, kern, activation=activ)(x)\n","        \n","    # Ouput layer (with optional regularization)\n","    output= Conv1D(1,1,activation='linear', kernel_regularizer=regularizers.l2(reg))(x)\n","    return Model(inputs,output)\n"]},{"cell_type":"code","execution_count":null,"id":"5c3ed55e-2ca4-4bc2-934d-6c81f2d24a3e","metadata":{"id":"5c3ed55e-2ca4-4bc2-934d-6c81f2d24a3e"},"outputs":[],"source":["#We propose here a default neural network setting but feel free to play with the neural net characteristics\n","\n","archi = [(40, 7, 'tanh'),\n","         (20, 5, 'tanh'),\n","         (20, 1, 'tanh')\n","        ]\n","model = buildmodel(archi)"]},{"cell_type":"code","execution_count":null,"id":"7b32c90a-88d7-460a-b295-50613e1eba3c","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":754},"id":"7b32c90a-88d7-460a-b295-50613e1eba3c","outputId":"68542fc9-78ad-4c48-fb3a-4251d7ea6a8f"},"outputs":[{"data":{"image/png":"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","text/plain":["<IPython.core.display.Image object>"]},"execution_count":14,"metadata":{},"output_type":"execute_result"}],"source":["# We can plot the architecture of our model\n","tf.keras.utils.plot_model(model, show_shapes=True)"]},{"cell_type":"code","execution_count":null,"id":"68bb83b7-c479-4bf5-877f-b7d1c70b2d7d","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"68bb83b7-c479-4bf5-877f-b7d1c70b2d7d","outputId":"3c265662-ee34-49a7-8698-6100805d9431"},"outputs":[{"name":"stdout","output_type":"stream","text":["Model: \"model\"\n","_________________________________________________________________\n"," Layer (type)                Output Shape              Param #   \n","=================================================================\n"," input_1 (InputLayer)        [(None, 40, 1)]           0         \n","                                                                 \n"," periodic1d_padding (Periodi  (None, 50, 1)            0         \n"," c1DPadding)                                                     \n","                                                                 \n"," batch_normalization (BatchN  (None, 50, 1)            4         \n"," ormalization)                                                   \n","                                                                 \n"," conv1d (Conv1D)             (None, 44, 40)            320       \n","                                                                 \n"," conv1d_1 (Conv1D)           (None, 40, 20)            4020      \n","                                                                 \n"," conv1d_2 (Conv1D)           (None, 40, 20)            420       \n","                                                                 \n"," conv1d_3 (Conv1D)           (None, 40, 1)             21        \n","                                                                 \n","=================================================================\n","Total params: 4,785\n","Trainable params: 4,783\n","Non-trainable params: 2\n","_________________________________________________________________\n"]}],"source":["# We can also have a summary of the model\n","model.summary()"]},{"cell_type":"markdown","id":"d20a5c4b-1ea6-480d-9f71-02791fd07b3b","metadata":{"id":"d20a5c4b-1ea6-480d-9f71-02791fd07b3b"},"source":["## <span style=\"color:blue\"> III. Train the neural network.</span>\n","In this section, the neural network defined previously is trained. The dataset is divided into two parts: the training set used to optimize the weights, and the validation set to follow the evolution of the metric and avoid overfitting.\n","\n","The training/validation is defined with two parameters:\n","- `frac_train` (default: 0.8): Fraction of the data selected as training data, the rest is used for validation\n","- `valend` (default: True): Flag to indicate if the validation is taken at the end of the analysis (if True) or at the beginning of the analysis (if False)"]},{"cell_type":"code","execution_count":null,"id":"781ff538-f95a-4e51-b2a5-a48e9ef24873","metadata":{"id":"781ff538-f95a-4e51-b2a5-a48e9ef24873"},"outputs":[],"source":["frac_train = .8\n","valend = True"]},{"cell_type":"code","execution_count":null,"id":"d8d401b0-21ce-43f3-900b-b94e9ea92be5","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"d8d401b0-21ce-43f3-900b-b94e9ea92be5","outputId":"75ae02ff-5add-4fd6-fee3-5793b6f8343c"},"outputs":[{"name":"stdout","output_type":"stream","text":["Total number of samples: 4999\n","Number of training samples: 3999\n","Number of validation samples: 1000\n","Number of input features: 40\n"]}],"source":["# Split in train/validation\n","X = x[:-1]\n","y = err\n","\n","itrain = int(frac_train*X.shape[0])\n","if valend == True:\n","    Xtrain, ytrain = X[:itrain], y[:itrain]\n","    Xval, yval = X[itrain:], y[itrain:]\n","else:\n","    Xtrain, ytrain = X[-itrain:], y[-itrain:]\n","    Xval, yval = X[:-itrain], y[:-itrain]\n","print(f'Total number of samples: {X.shape[0]}')\n","print(f'Number of training samples: {Xtrain.shape[0]}')\n","print(f'Number of validation samples: {Xval.shape[0]}')\n","print(f'Number of input features: {Xval.shape[1]}')"]},{"cell_type":"markdown","id":"86d3954a-00f9-46de-8f04-7015ecd7e36c","metadata":{"id":"86d3954a-00f9-46de-8f04-7015ecd7e36c"},"source":["The data presented as an input or output are standardised to facilitate the training. Important: the scaling factors (mean and standard deviation) are computed on the training set only, and the same values will be applied for validation and testing."]},{"cell_type":"code","execution_count":null,"id":"c958c329-fd2f-49f7-aaea-10061fc3bdbc","metadata":{"id":"c958c329-fd2f-49f7-aaea-10061fc3bdbc"},"outputs":[],"source":["scaler_x = StandardScaler().fit(Xtrain)\n","scaler_y = StandardScaler().fit(ytrain)\n"]},{"cell_type":"markdown","id":"738695aa-fb23-420d-b08d-3b44bae0cf01","metadata":{"id":"738695aa-fb23-420d-b08d-3b44bae0cf01"},"source":["The model is compiled as required by the keras API. The loss function optimized is the mean square error and the chosen optimiser is `Adam` (see https://www.tensorflow.org/api_docs/python/tf/keras/Model#compile for more details)"]},{"cell_type":"code","execution_count":null,"id":"8b2598dc-19ad-4782-92f8-de4ffb183a71","metadata":{"id":"8b2598dc-19ad-4782-92f8-de4ffb183a71"},"outputs":[],"source":["model.compile(loss='mse', optimizer='adam')"]},{"cell_type":"markdown","id":"c8e1b8d6-721b-4c93-ba97-8aa748b8d37b","metadata":{"id":"c8e1b8d6-721b-4c93-ba97-8aa748b8d37b"},"source":["To fit the model to the data, we specify the number of epochs. The evolution of the metrics during the training is saved in the `history` variable."]},{"cell_type":"code","execution_count":null,"id":"56dec98b-e95d-4b93-9744-7115da6dae7b","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":49,"referenced_widgets":["747d3c6ccb794e67bfc262ce031f9fc5","973fca2dd29b4e0881bace37d1eca805","24eb5bf591a3440b92ef49cba0b848ff","38b06265149143a09648875f5a435782","cb7abe9177294838930daa210455aef3","9ded00ca570047ed9d9b6f7f120982ff","0c091879d3ba48edb9b762aff8c5cd92","013c2ca16a9f43bbbde8c9f746837eda","670eb83f4a3047f38456fd21ab792418","2233d06a6bc949c1b21917875d699b85","82a0a5bcffbd431cb12359259e78b76e","39a241de90394f989e139eba6d3dfefa"]},"id":"56dec98b-e95d-4b93-9744-7115da6dae7b","outputId":"b700975b-02a8-4d23-ab58-7ab465a0f6bc"},"outputs":[{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"39a241de90394f989e139eba6d3dfefa","version_major":2,"version_minor":0},"text/plain":["NN training:   0%|          | 0/500 [00:00<?, ?it/s]"]},"metadata":{},"output_type":"display_data"},{"name":"stderr","output_type":"stream","text":["2023-03-01 11:58:19.025420: W tensorflow/tsl/platform/profile_utils/cpu_utils.cc:128] Failed to get CPU frequency: 0 Hz\n"]}],"source":["num_epochs = 500\n","\n","# We define a callback to follow the evolution of the training\n","callbacks = [ tqdm_callback(num_epochs, 'NN training') ]\n","\n","history = model.fit(scaler_x.transform(Xtrain), scaler_y.transform(ytrain), \n","                    epochs=num_epochs, \n","                    batch_size=256, \n","                    validation_data = (scaler_x.transform(Xval), scaler_y.transform(yval)),\n","                    verbose=0, callbacks=callbacks)"]},{"cell_type":"code","execution_count":null,"id":"3ba0cf8e-7d88-44ee-b452-27343387a17a","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":367},"id":"3ba0cf8e-7d88-44ee-b452-27343387a17a","outputId":"51958fb6-cdf4-456a-cffb-87fdc29d641c"},"outputs":[{"data":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAAkMAAAGgCAYAAACpAWMqAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy88F64QAAAACXBIWXMAAA9hAAAPYQGoP6dpAABkB0lEQVR4nO3dd3RU1d7G8e+kdwgghF5CkxZCl16lYwNRygVBsAGCBSwIIqioIIgCoqIgioKAgAIivUjvvScQAkkgCaSQnnn/mDcDY0I1mZkkz2ct10rOOXPOb/bkMs/de599DEaj0YiIiIhIPuVg6wJEREREbElhSERERPI1hSERERHJ1xSGREREJF9TGBIREZF8TWFIRERE8jWFIREREcnXnGxdgL27evUqq1evply5cri7u9u6HBEREbkHCQkJBAcH0759e4oUKXLHYxWG7mL16tX06dPH1mWIiIjIA/jpp5/o3bv3HY9RGLqLcuXKATBz5rdUqlQlW8/t6OiAj487MTEJpKWlZ+u55Sa1s3Wona1HbW0damfryKl2Pn36JC+9NMj8PX4nCkN3kTE0VqlSFWrVqp2t53ZycsDX15Po6HhSU/U/tJyidrYOtbP1qK2tQ+1sHTndzvcyxUUTqEVERCRfUxgSERGRfE1hSERERPI1hSERERHJ1zSBWkRErCI9PZ20tFRbl3HP0tMNJCY6kpycRFqa0dbl5FkP0s6Ojk44OGRff47CkIiI5Cij0UhMTBQJCXG2LuW+Xb3qQHq67iTLaQ/Szu7uXvj4FMJgMPzn6ysMiYhIjsoIQl5evri4uGbLl5e1ODoa1CtkBffTzkajkeTkJOLiogEoUKDwf76+wpCIiOSY9PQ0cxDy8vKxdTn3zcnJQWsMWcH9trOLiysAcXHReHv7/uchM02gFhGRHJOWlgbc/PISyS4Zf1PZMQ9NYUhERHJcbhoak9whO/+mFIZEREQkX1MYEhERkXxNYciGjEbdoSAikluEhYWxdu3q/3SOffv20LRpPS5fvnRPx69c+QdNm9b7T9e8myFDBvPhh+/n6DXsncKQjURFRfLttzPYvn27rUsREZF78OGHY9m587/9m12zZgDLlv1F0aLF7un4Nm3asWzZX//pmnJ3CkM2cvVqBImJCZw+fdrWpYiIyD3Ijt58Z2dnChcugqOj4z0d7+rqRuHCRf7zdeXOtM6QjTg7uwCQnJxs40pERKzPaDSSmmq7R3M4OTnd191IQ4YM5sCBfRw4sI/9+/eyaNEfdO/elebNW7Jr1w6ioqKYMOETKlWqwqxZX7Ft21YiI69SoEBBmjdvydChI3B1dWPfvj0MG/Yiv/22nOLFS9C9e1cef/wpjh8/yq5dO3BxcaF9+068/PKrODk5sXLlH3z00Ti2bt0DQNOm9Rg58l3Wr1/DoUMH8fHx4amnnqZv3+fMtf7991/Mnfsdly9fwt+/Iu3adWTatMnmc9xNcHAQM2dO4/DhQ6SlpVK/fiOGDBmBn58fACEhF5gy5TOOHj1EerqRmjVr8corw/H3rwjA9u3/8N13XxMcfA53dw8eeaQJQ4e+ho+P/a4zpTBkIy4uzoDCkIjkP0ajkd9/X0BY2L3Nm8kJfn4leOKJnvcciD766DNGjhxB0aLFGDFipHn70qWL+eSTKXh7e1OhQkXGjHmLiIhwxo//hEKFCnHkyCE+/vgDypQpx9NPP5vlub///hteemkYL744lF27tjN16iQqV65Khw6dszx+xowvGDFiJK+9NorVq1cya9Z0atWqTUBAIP/8s4UPPxzLCy8MoWnT5uzbt5tp06bcc7uEhV3mxRefo169hkyb9jXJyUlMn/4FQ4cOZu7cX/Dw8GTs2HeoWLES3303j9TUVKZPn8o777zBggVLuXbtGu+++yZDhoygceOm/98WY5kx4wveeuu9e67D2hSGbCSjZygpKcnGlYiIyN34+BTAyckJV1dXfH19zdsbNWpC/foNzb/Xr9+QWrUCqVSpMgDFi5dgyZKFnD17+ykRDRs+Qo8ezwBQunQZ/vxzGYcOHbhtGOrYsSvt23cCYNCgl1iy5DcOHTpAQEAgv/wyj5Yt29CrV18AypQpS0hICAsW/HxP73PJkt9wd/dgzJjxuLiYvqcmTPiE7t27snr1Kp54ojuXLl2kQYNGFC9eAicnJ95+ewznzweTnp7OlSvhJCcnU6yYH35+xfHzK84nn3xuXnzTXikM2YiGyUQkvzIYDDzxRM9cNUx2O6VKlbb4/YknerB162b+/nsVFy+GcO7cGS5dCqVUqTK3PUfZsuUtfvf09Lpj25QtW+5fx3uSkpICwMmTJxg8+GWL/QEBgfcchs6dO0PVqg+bgxCAr28hypQpaw50gwa9zLRpk1m6dBF16tSnYcNHaN26HQ4ODlSqVIW2bdszapSpF61+/YY0btyUJk2a39P1bUUTqG3E2fnmMJlusReR/MZgMODs7Gyz/7Jr9WJX15uPGTEajYwaNYKpUz/DwcGBVq3a8MknU6hZM+CO58j4PrjVnb4Xbg0q/z7e0dERo/HBn6VmOk3mtklPT8fJydR/8tRTT7N06SqGDXsdd3c3Zs2aTp8+3YmKigTg/fc/ZP78RTzzTG+ioiJ5//13ee21IQ9ckzWoZ8hGXFxcCA8Px9fX9/+7D5VLRUTs2d0C1KlTJ9m+/R9mzZpD9eo1AEhNTSU0NIQSJUpao0QqVqzE0aOHLbYdO3bknl/v7+/P33//RXJysjl0RUVFEhISwhNPmALPnDnf0adPfzp16kqnTl25ciWCJ57oxP79+yhWzI/16/9m2LDX/3+eVC/+/nsVH3zwHtHRUfj6FsrW95td9A1sIzt37mDmzJn89ddfpKRoqExExN65u3tw+fIlIiLCs9xfuHBhHB0dWb9+DZcuhXLixDHee+8tIiMjrfbvfJ8+/dm4cT2//voTFy+GsGrVnyxa9Os9v/6xx54iPj6eDz54jzNnTnP8+FHee+8tChYsSJs27fHxKcC2bVv55JMPOX36JKGhF/n990U4OztTterDeHp6smTJb8yYMY2LF0M4e/YMa9euplSpMhQoUDDn3vh/pDBkIxnjuyEhIZo3JCKSCzz++FMEBZ2lX79ns5wQXKTIQ7z77jj++Wczffr0YPToUTz00EP07NmL48ePWWVKRKNGjXnzzbdZsuQ3+vZ9mmXLlvDYY09lORSXlZIlS/Hll7OIjY3lhRf6M2LEEAoXLsKMGbPx9vbGycmJSZOm4eBg4NVXX6Zv357s3bubTz+dSsmSpShfvgIffvgZ+/bt4bnnevHyywNxdHRi8uRpODjYb+QwGDVh5Y727dtH3bp1Wbt2M7Vq1c62854/H0z9+rVwdHTkyJGTFC5cNNvOLZacnBzw9fUkOjqe1NQHH0uXO1M7W09uauuUlGQiIy9TuHBx840juYmTk4Pdt/Gt9u/fS+HChSlTppx5248/fs+ffy5j4cJltivsLh6kne/2t3Xo0AHatm3O3r17qVOnzh3PZb8xLY8rVao0jo6OpKWlceHCeVuXIyIiecCuXTsYMWII+/btISwsjK1bN7Fw4S/mW/Ela5pAbSOOjo489NBDhIWFcfbsGQID69u6JBERyeWee24QCQkJjB8/hmvXoilatBg9e/aiV6//2bo0u6YwZEPFivkRFhbGuXNnbV2KiIjkAS4uLgwf/gbDh79h61JyFQ2T2VCJEiUA03NgRERExDYUhmyoaFHTpOmIiAgbVyIiIpJ/KQzZUMbiU1FRUTauREREJP9SGLKhIkWKABAdrTAkIiJiKwpDNlSkyEMAREdfs20hIiIi+ZjCkA1lzBm6cSOepKQkG1cjIiKSPykM2VDx4iXMy5NfvXrFxtWIiEhO6t69K7NnzwJg5co/aNq03h2Pb9q0HitX/nHP5w8LC2Pt2tVZXi8n7Nu3h6ZN63H58qUcu4a1aJ0hG/L29sbT05PY2FiuXr1CyZKlbF2SiIhYQZs27WjY8JFsPeeHH47Fz684bdu2B+Dbb3/E1dU1W6+RVykM2ZCX180wpNvrRUTyD1dXN1xd3bL1nP9+1Kivr2+2nj8vUxiyIU9PLzw9PQG4dCnUxtWIiFiP0Wjkxo0bNru+h4cHBoPhno//8MP3CQ4O4ttv55q3RUSE0717Vz7//Cvq1WvAihXLWbToV86fP4+Dg4GqVasxdOhrVKlSNdP5Vq78g48+GsfWrXvM5/r880/Yu3cP3t7evPzyMIvjjUYjv/wyjxUrlnPpUiguLi4EBAQyfPiblChRkiFDBnPgwD4OHNjH/v17WbToD7p370rHjl0YOPAFALZt28qcOd8RFHQWDw9P2rZtz+DBL5t7j5o2rcfIke+yfv0aDh06iI+PD0899TR9+z53T22UlpbGokW/snTpYsLDwyhWzI9nn+1Lt25PmI+ZP38eS5cu4sqVCIoUeYjOnbvRr99AABITE5k69TO2bdtKXFwsZcuWo3//52nRovU9f04PSmHIhhwdHSlYsCAAly5dtG0xIiJWYjQa6dLlUXbv3mmzGho0aMQff6y+50DUsWMXhg17kYsXQyhVqjQAf/+9iiJFHqJOnXps2rSBSZM+ZuTIdwkMrEtkZCRTp37GxIkf8MMP8+947tTUVF5/fSheXl589dU3pKQkM3nyRItjFi6cz48//sDo0eOoWLESly6F8umnH/Lll1P4+ONJfPTRZ4wcOYKiRYsxYsTITNfYvHkjo0ePZMCAwYwe/T4hISFMnjyRsLBLfPjhZ+bjZsz4ghEjRvLaa6NYvXols2ZNp1at2gQEBN61jb76aip//bWCESNG8vDD1di1awdTpnxKcnIS3bs/w9atm/nxx+8ZP/5jSpcux9Gjh5gwYSzFi5egc+cufPvtTM6ePc1nn32Bt7c3f/yxlDFj3ubXX3+nePES9/IxPTCFIRsrVMi08OKVKxomE5H84356ZexBYGBdSpQoyd9/r2LAgMGAKQx16NAZBwcHChQowKhRo+nQoTMAfn7F6dr1cSZN+viu5967dzdBQedYsGCpee7oO++M5bnnepuPKVmyNKNHv0/Tps3N52/Vqi3r1v0NgI9PAZycnHB1dc1yeGzevB9o3rwl/fs/D0CZMuUwGo289dZrBAcHUa5ceQA6duxqfsL9oEEvsWTJbxw6dOCuYSg+Po7ff/+NoUNH8OijHQAoXboMly6F8uOPP/DUUz25dOkirq4uFC9eEj8/P/z8/ChSpCjFivkBpk4BDw8vSpYshZeXF88//yIBAYF4e/vctQ3/K4UhG8voGbp27ZpN6xARsRaDwcAff6zOVcNkBoOBDh068/fffzFgwGBOnz7JuXNnmTDhUwBq165DcHAQc+Z8x8WLIVy4cJ4zZ06Rnp5+13OfPXsGb28fi5toKlWqgpvbzTlFTZs25+jRI8yePYuQkAucPx/EuXNneeihovdU/7lzZ2jXrr3Fttq16/z/9U+bw1DZsuUsjvH09CQlJeWu5z9/PpjU1FRq1ar9r2sEsmDBz0RHR/Hoo51YsWI5zzzzBP7+lahfvyEtWrTCz88Uhnr37seoUSPo2rUd1avXpEGDRrRp8yheXl739B7/C4UhG8tI8DEx121ciYiI9RgMBvOcydyiU6eu/PDDtxw/fpR169ZQs2YtypQpC8DatasZP34M7dp14OGHq9Oly2OcO3eWzz//5J7O/e/JzwBOTje/on/+eS6zZ39Dp05dCQys+//DTpssbqW/8/kBLMNfenpapuu4uLjcU21Znz+ra6Sbr+HjU4AffpjPkSOH2L17Jzt3bufXX39i4MAXGDToBWrUqMWSJSvYvXsne/bs4s8/lzF79iwmT/6SevUa3NP7fFBaZ8jGMnqG4uLibFuIiIjckZ9fcQID67Jhw1rWr19Dhw5dzPvmzfuBrl0fZ/TocTz11NPUrl2H0FDTXNC7hYnKlasQFxfLuXNnzdsuXDhv8b0wd+73DBgwiDfeeIvHHnuSGjVqEhJy3uLcd+rp8vf359Ch/RbbDh40/V62bPl7ePd3VrZsORwdHbO8RuHChfH29mH16pX8/vsiatWqzcCBL/DNN3Po2vVx81Df7NmzOHToAE2btmD48Df55ZcllCxZio0b1//n+u5GPUM2pjAkIpJ7dOrUlc8//5TU1FTatHnUvL1o0WIcPnyQkydP4OXlxdatm1iyZCEAycnJd1zvp06delSrVoMJE8bw2mtv4eTkyJQpn5kX5c04/+7dO2nSpDmOjg789ddKNm3aQKFChc3HuLt7cPnyJSIiwilatJjFNZ599n+MHfs2c+Z8R+vW7QgJucCUKZ/RuHEz8xDZf+Hl5UW3bk/y3Xez8PYuQLVq1dm5czu//76IwYNfwWAwkJSUxPTpX+Dp6UlAQCAREeHs37/XPFx38WIIq1evZNSo0ZQsWYojRw4RFhZGzZq1/nN9d6MwZGOFC5v+kG/ciLdxJSIicjctW7bh888/pVmzFhZzWUaMGMmnn37IkCGDcXFxpmLFyowePY6xY9/h2LEjBAbWve05HRwc+OyzqUyZ8hmvvTYEV1dX+vZ9zmJl5/fe+4DPP/+E55/vi4eHJ9Wr1+CNN95m8uSJXL58ieLFS/D440/x4Ydj6dfvWf78c43FNVq3bktaWio//TSHuXNnU7CgL+3atTffdp8dXn31dQoWLMjXX39JdHQUJUuWZsSIkeZb67t1e4LY2BjmzPmOiIhwvL29admyDS+9ZFpG4I033uKrr77ggw/eIybmOn5+xXnppaHmCd05yWC8l8HAfGzfvn3UrVuXtWs3Z5oY9l85OTmwfv1fdO/enUKFCnPiRFC2nl9MnJwc8PX1JDo6ntTUu09mlAejdrae3NTWKSnJREZepnDh4jg7Z56PYu+cnBzsvo3zggdp57v9bR06dIC2bZuzd+9e6tSpc8dzac6QjRUpUgSAhATb3VUhIiKSnykM2dhDDz0EQFJS0j3N2BcREZHspTBkY8WKmSa5paenExsbY+NqRERE8h+FIRvz9fU13w559epVG1cjIiKS/ygM2ZiDg4P5lsvISIUhERERa1MYsgPu7u4AREUpDIlI3qQ5kZLdsvNvSmHIDmSEoejoaBtXIiKSvRwdHQFITk6ycSWS12T8TTk6/vclE7Xooh3IeD6PwpCI5DUODo64u3sRF2f6983FxTVXPbE+Pd1AWpp6tXLa/bSz0WgkOTmJuLho3N29LFbqflAKQ3bAw8MUhvSwVhHJi3x8CgGYA1Fu4uDgcE9Pnpf/5kHa2d3dy/y39V8pDNkBDw8PAOLiYm1ciYhI9jMYDBQoUBhvb1/S0lJtXc49c3Q0UKCAB9ev31DvUA56kHZ2dHTKlh6hDApDdiBjzlBCQoKNKxERyTkODg44OOSeR3I4OTng5uZGQkKaHsmRg+yhnTWB2g64ubkBCkMiIiK2oDBkB9zdTcNkiYkKQyIiItamMGQHNEwmIiJiOwpDdiDjbrKkpEQbVyIiIpL/KAzZAQ8PU89QYqIWJRMREbE2hSE7cLNnSGFIRETE2hSG7EDGCtQaJhMREbE+hSE74OnpBUBycrKNKxEREcl/FIbsgJeXqWdIYUhERMT6FIbsgJeXN2AKQ0ajlnwXERGxJoUhO+DlZRomS0lJIT09zcbViIiI5C8KQ3Ygo2coJSWFlJQUG1cjIiKSvygM2YGMu8lSU1NJSck9T3QWERHJCxSG7ICbm2nRxZSUFFJT1TMkIiJiTQpDdiBjBer09HRu3Ii3cTUiIiL5i8KQHcjoGQKIi4uzYSUiIiL5j8KQHXBzczP/HB+vMCQiImJNCkN2wGAw4OzsAqhnSERExNoUhuyEq6spDN24ccPGlYiIiOQvCkN2wsUlIwxpArWIiIg1KQzZCRcXVwASEtQzJCIiYk0KQ3YiY5gsISHBxpWIiIjkLwpDdsLV1XRHmeYMiYiIWJfCkJ1wdc0YJlPPkIiIiDUpDNmJjDCUlJRo40pERETyF4UhO5FxN1liYpKNKxEREclfFIbsRMacoaQkDZOJiIhYk8KQncgYJlPPkIiIiHUpDNkJNzdTGEpOVhgSERGxJoUhO3FzmExhSERExJoUhuxExpPrk5OTbVyJiIhI/qIwZCfc3NwB9QyJiIhYm8KQncjoGUpJUc+QiIiINSkM2Ql3dw9Aw2QiIiLWpjBkJ9zdTcNkyckpNq5EREQkf1EYshMZYUjDZCIiItalMGQnNEwmIiJiGwpDdsLDwxSGUlJSMBqNNq5GREQk/1AYshMZw2Spqamkp6fbuBoREZH8Q2HITri7ewKmMJSammrjakRERPIPhSE74eFxs2coLU1hSERExFoUhuxExgrUpjCUZuNqRERE8g+FITuRsQK1hslERESsS2HITtwahjRMJiIiYj0KQ3bC1dUVgPT0dD2sVURExIoUhuyEq6ub+eeEhHgbViIiIpK/KAzZiYyeIYD4+Bs2rERERCR/URiyE05OTjg4mD6OGzcSbFyNiIhI/qEwZEecnZ0BSExUz5CIiIi1KAzZEWdnFwBu3FAYEhERsRaFITvi4pLRM6RhMhEREWtRGLIjGT1DCQmJNq5EREQk/1AYsiPqGRIREbE+hSE74uRkCkNJSeoZEhERsRaFITuS0TOUlJRs40pERETyD4UhO5IxZ0iP4xAREbEehSE74uKSEYY0TCYiImItCkN2JCMMJSdrmExERMRaFIbsiIuL6flkyckaJhMREbEWhSE7cnOYTGFIRETEWhSG7IiGyURERKxPYciOuLqahslSUlJsXImIiEj+oTBkR1xd3QDNGRIREbEmhSE7op4hERER61MYsiMKQyIiItanMGRHbg6TaQK1iIiItSgM2RE3N1MYUs+QiIiI9SgM2ZGMYbLU1FQbVyIiIpJ/KAzZEXd3d8AUhtLT021cjYiISP6gMGRH3NxuhqG0NPUOiYiIWIPCkB3JCENpaWkaKhMREbEShSE74upqehyHwpCIiIj1KAzZkYyn1puGydJsXI2IiEj+oDBkRzIe1JqWlqYwJCIiYiUKQ3YkIwypZ0hERMR6FIbsSMYwmXqGRERErEdhyI7cOoFat9aLiIhYh8KQHdEEahEREetTGLIjLi7OgIbJRERErElhyI5ozpCIiIj15aswFB4eTuvWrW1dxm3dOkymRRdFRESsI9+Eoe3bt9OvXz+uXLli61JuK2OYDCAlJcmGlYiIiOQf+SYMLV68mKlTp9q6jDvK6BkCSExMtGElIiIi+YeTrQuwlkmTJtm6hLtydb0ZhhISFIZERESsIU+FoeXLl2fq/Wnbti3vvPOObQq6T05OThgMDhiN6SQlKQyJiIhYQ54KQ926daNbt262LuM/cXZ2Ijk5meRkhSERERFryDdzhnILJyfTJOrERE2gFhERsQaFITvj7GwKQxomExERsY77DkPx8fGMHz+eFi1aULduXV5++WUuXLiQ7YXNmDGDvn37WmxLT09n2rRpNGvWjICAAAYMGMD58+fv67yHDx/OzjKznbOzaeQyKUk9QyIiItZw32Fo+PDhrFmzhvfff5+FCxdSuHBhnn32WaKjo7OtqDlz5jBt2rRM22fMmMGvv/7KhAkTWLBgAQaDgUGDBpGcnJxt17Y1Z2fTw1qTkvLOexIREbFn9xWGTpw4webNmxk/fjytWrXC39+fcePG4eXlxfz58zMdHxkZyZEjRyy27du3j7i4uCzPHx4ezvPPP88XX3xB+fLlLfYlJyfz/fffM3ToUFq0aEHVqlWZMmUK4eHhrFmz5n7ehl3LWHhRw2QiIiLWcV9hKCgoCIB69erdPIGDA1WrVmX37t2Zjv/xxx/p168f+/btA2DLli3079+f5cuXZ3n+o0ePUqBAAZYvX05AQIDFvhMnThAfH0+jRo3M23x8fKhWrVqW186tMnqGkpM1TCYiImIN93Vr/UMPPQRAWFgY/v7+5u2hoaFZznF59dVXCQsLY+DAgQwePJgZM2bQv39/evXqleX5W7dufdtnh4WFhQFQvHhxi+1Fixbl8uXL9/M27FrGBOq8NPQnIiJiz+6rZyggIAB/f3/Gjh3L5cuXSU5OZs6cORw/fjzLL28HBwc+/vhjAgMDmTp1Kh06dOD1119/oEITEhIAcHFxsdju6uqapyYbZ6xCrTAkIiJiHfcVhpydnZk+fTopKSm0bNmSwMBAdu/eTffu3fHy8sryNevXr2fXrl1Uq1aNNWvWsGPHjgcq1M3NDcgcEpKSknB3d3+gc9qjjLCnCdQiIiLWcd93k5UvX54FCxawa9cutm/fzvTp07l27RrlypXLdOzq1asZPnw4gwcPZsmSJXTr1o0XXniB7du333ehGcNjERERFtsjIiLw8/O77/PZq5tzhhSGRERErOG+wlBcXBx9+vThyJEjFChQAB8fH2JjY9m2bRvNmjXLdLyLiwtDhgxh2LBhGAwGxo0bR+/evc29PPejatWqeHl5sXPnTvO2mJgYjh07ZjGhO7fL6BlKSVEYEhERsYb7mkDt5eWFwWDgo48+YuzYsRiNRsaPH0+JEiXo0qVLpuNbtWpFq1atzL8bDAZGjhz5QIW6uLjQp08fJk2aRKFChShZsiSfffYZfn5+tGvX7oHOaY80Z0hERMS67vtBrZMnT2bChAn06dMHBwcHWrVqxciRI3Fyyvlnvg4bNozU1FRGjx5NYmIi9evXZ/bs2ZkmVedmGWFIPUMiIiLWcd8JpmjRolmuDp3dJk6cmGmbo6Mjb775Jm+++WaOX99WXFwywlCKjSsRERHJH/SgVjtzs2dIYUhERMQaFIbsjKuraXK5wpCIiIh1KAzZGVfXjLvJFIZERESsQWHIzri5mRaQTE1NJT093cbViIiI5H0KQ3YmY5gsLS2NtLQ0G1cjIiKS9ykM2ZmMBSlTU1MVhkRERKxAYcjOZMwZSktLIz1dYUhERCSnKQzZGQ2TiYiIWJfCkJ3JWE1bw2QiIiLWoTBkZzLCkKlnKNXG1YiIiOR9CkN2JuNxHBomExERsQ6FITuTMYFaw2QiIiLWoTBkZ9QzJCIiYl0KQ3bG2VlzhkRERKxJYcjOaJhMRETEuhSG7IyGyURERKxLYcjOaJ0hERER61IYsjOW6wwpDImIiOQ0hSE74+p66zCZJlCLiIjkNIUhO5PRM2Q0GklOTrZxNSIiInmfwpCdyZhADZCYmGDDSkRERPIHhSE74+7ubv75xg2FIRERkZymMGRnHBwccHJyAiAh4YaNqxEREcn7FIbskKurGwAJCeoZEhERyWkKQ3YoYxXqxMREG1ciIiKS9ykM2aGM2+vVMyQiIpLzFIbskMKQiIiI9SgM2SFXV9MdZUlJGiYTERHJaQpDdsjNzdQzpDlDIiIiOU9hyA5l3E2mMCQiIpLzFIbskJubKQxpmExERCTnKQzZoYxVqJOSkmxciYiISN6nMGSH3NwUhkRERKxFYcgOubtnDJMpDImIiOQ0hSE75O7uAUByssKQiIhITlMYskM3w1CyjSsRERHJ+xSG7JAmUIuIiFiPwpAd8vBQz5CIiIi1KAzZIS8vL8AUhtLS0mxcjYiISN6mMGSHPD1NYSg1NZXU1BQbVyMiIpK3KQzZIQ8PTwBSUlJISVEYEhERyUkKQ3YoY50hhSEREZGcpzBkhzJurU9NTVUYEhERyWEKQ3Yo40Gtpp4h3VEmIiKSkxSG7FDGs8k0TCYiIpLzFIbsUMaiiykpKbqbTEREJIcpDNkhb28fwBSGEhMTbVyNiIhI3qYwZIcyFl0EiIm5bsNKRERE8j6FITvk6uqKo6MjADExsTauRkREJG9TGLJDBoPBPG8oNlY9QyIiIjlJYchO3QxD6hkSERHJSQpDdirjkRxxcQpDIiIiOUlhyE55eprCUGxsnI0rERERydsUhuxURhiKj1fPkIiISE5SGLJTnp6m2+vj4+NtXImIiEjepjBkpzLWGrpx44aNKxEREcnbFIbslLe3NwA3bqhnSEREJCcpDNkpH58CgIbJREREcprCkJ0qWLAgYBomMxqNti1GREQkD1MYslMFC/oCkJSURHJyko2rERERybsUhuxUxpPrk5KS9OR6ERGRHKQwZKe8vEwTqJOTk0lMTLBxNSIiInmXwpCdyribTD1DIiIiOUthyE75+hYCTHeTqWdIREQk5ygM2amHHioKmO4mi4/X88lERERyisKQnSpcuDAGgwGAK1fCbVyNiIhI3qUwZKecnJzw8THdURYerjAkIiKSUxSG7FihQoUBuHLlio0rERERybsUhuxY4cKmMHT1qsKQiIhITlEYsmMZk6ivXr1q40pERETyLoUhO1asmB8AUVGRNq5EREQk71IYsmPFi5cA4Pr166Smpti4GhERkbxJYciOZYSh+Ph44uPjbVyNiIhI3qQwZMeKFjXNGYqLi9PCiyIiIjlEYciOZUygVs+QiIhIzlEYsmO3hqHY2BgbVyMiIpI3KQzZscKFiwBgNBoJC7ts42pERETyJoUhO+bs7Gx+JEdY2CUbVyMiIpI3KQzZuYxHcqhnSEREJGcoDNm5m6tQ65EcIiIiOUFhyM4VLVoMgMjISIxGo42rERERyXsUhuxciRKmhRdjYmJITk6ycTUiIiJ5j8KQnct4Pplp4UWtNSQiIpLdFIbsXNmy5QDTMJlWoRYREcl+CkN2rmrVagBEREQQFxdr42pERETyHoUhO1ehgj+Ojo4kJycTFHTO1uWIiIjkOQpDds7Z2ZlSpUoDcOLEMRtXIyIikvcoDOUCFStWAuDUqZM2rkRERCTvURjKBWrUqAnAmTOnbVyJiIhI3qMwlAs0adIMgKCgIJKTk21cjYiISN6iMJQLNGjwCA4ODsTGxnLixHFblyMiIpKnKAzlAh4eHpQubZpE/c8/m2xcjYiISN6iMJRLVKpUGYDDhw/auBIREZG8RWEol6he3TSJ+tSpk3pgq4iISDZSGMol6tSpD0B4eDgREWE2rkZERCTvUBjKJapUqQJAVFQUly9fsnE1IiIieYfCUC5RunRZHBwcSElJITj4rK3LERERyTMUhnIJZ2dnihcvAcDp06dsXI2IiEjeoTCUi/j7VwT0WA4REZHspDCUizRs2BiAU6dOkZqaauNqRERE8gaFoVykWbPmAJw/f57r16NtXI2IiEjeoDCUiwQG1sXJyYkbN26wf/9eW5cjIiKSJygM5SKurq5Urmy6xX7z5o22LUZERCSPUBjKZerXbwigniEREZFsojCUy7Rs2QYwTaJOSUmxcTUiIiK5n8JQLtOiRUsMBgPR0dHs2bPT1uWIiIjkegpDuYyXlzfVq9cAYO7c2TauRkREJPdTGMqF+vT5HwBLlixmxowvbVyNiIhI7qYwlAs9/XQvSpcuDcD48WNITEy0cUUiIiK5l8JQLuTl5c2HH04EIC0tjdOn9XgOERGRB6UwlEs1adKSsmXLArB9+z82rkZERCT3UhjKpby9valS5WEAtm3bYuNqREREci+FoVysUaNHADh79gxpaXpwq4iIyINQGMrFateuC8Dly5e5cuWKjasRERHJnRSGcrE6derh4uJCTEwMO3dq3pCIiMiDUBjKxTw8PKhTx9Q7tHr1XzauRkREJHdSGMrlHn20IwA7d24nNjbGxtWIiIjkPgpDuVz79p0A07yhJUt+JTo6ysYViYiI5C4KQ7lchQr+uLm5kZqayvnzwezZs8PWJYmIiOQqCkO5nKOjIw8/XA2A8PBwLl8OtXFFIiIiuYvCUB5QvXpNANavX8+BA/s1VCYiInIfFIbygBo1agEQGRnJTz/9xNdff2XjikRERHIPhaE84Omnn2Ho0BHm3zdtWk9kpBZhFBERuRcKQ3mAl5c37703jpkzvwPgypUr7NihRRhFRETuhcJQHlKpUmUAoqKiOH/+HKGhF2xckYiIiP1TGMpDKlTwByA+Pp6EhATWrFlFUlKSjasSERGxbwpDeYiXlzfFivkBkJSUzI0b8Zw7d9rGVYmIiNg3haE8JqN36Pr1WADOnDlpy3JERETsnsJQHtOt2+MA/PTTXC5fvkxIyHn27dtFdHQUSUmJti1ORETEDikM5TEDBgymQ4fOpKSk8PvvvxMaGsqOHVv55Zc5rFy5zNbliYiI2B2FoTzGYDDwySeT8fDwJCIigu+//56oqChWr17Npk0biIgIt3WJIiIidkVhKA8qXrwEixaZeoHS0tKYOXMm27dvZ/78+Rw9etDG1YmIiNgXhaE8ql69Bnz44ScApKSkmLcfOLCPuLhYW5UlIiJidxSG8rCuXR/H2dnZYlto6EW2b9+C0Wi0UVUiIiL2RWEoD/PzK86SJSsYM2Y8JUuWAiAsLIzTp09w8OA+G1cnIiJiH5xsXYDkrIYNG9GwYSNSU1P46KMPiI2NA2D79s14e3sTFnaZlJQUmjVrhaOjo42rFRERsT71DOUTbdo8isFgYPv2bYSHXyE9PZ3Vq//k4MG9HDt2iODgc6SkJGv4TERE8h2FoXyiZs1aDB78MgAzZ07nxImTXL58mf3795Oens7ff//JnDmzWLhwHrGxMTauVkRExHo0TJaPjBnzAWlpqXz33SxWrlyBu7sHERHhxMbG0rhxY37//XdKly6Nu7sH3bp1t3W5IiIiVqEwlI84OzszYcIn7Nq1k0OHDhAba7rFfv369Xh5eXHgwAEOHz5M1apVuXAhmGvXoqhatQYuLi42rlxERCTnaJgsn3FwcGDcuA8zbV+5ciVgWqRxw4YNDBrUj2XLFrNhw2prlygiImJVCkP5UJMmzejevSfOzs707z8QgNTUVPP+PXv2sH//fhYvXszZs6dZunQhhw7t0+RqERHJkxSG8qkvv/ya48fP8d574/D09MrymEuXLrF27VrOnw9i69aNrFjxOxcuBFu3UBERkRymMJRPOTo64uNTAG9vH0aMeMO8vU+ffhbHbd26lb179wJw4UIwK1b8zvXr16xZqoiISI7SBGrh5ZeHcf36dcqVK8+TT/agVKnSnDp1kiVLfgMgNjaeggV9uXYtGqPRyI8/zqZnz74ULVo007lCQy/i4OBA8eIlrP02REREHojCkODk5MR7740z//7aayNJT0+nd+//8dRTXdmxYxszZ37Ht9/OZMOGtRw5coQ5c75n+fLVpKWlUrx4CZydXYiLi6N16yakpqZy5MgZ3N3dbfiuRERE7o3CkGTJwcGBZs1a8PDD1Tl+/Ci1alWx2B8SEsLEieN4+OGHcXV1o2vXJ9m/fx/R0dEATJ8+hTfeeAeAyMhIVq9eSY8ez2R6cKyIiIitac6Q3NFXX31NsWJ+We7bsGED+/fvJzY2hiVLFvDbb/PN+zZv3mhex6hPn6cZPvwVpk//wio1i4iI3A+FIbmjmjUD+OWXxVkOeUVERLBs2TK2b99BenoaZ86cMe87deoUwcFniIyMZO/e3QDMmTOb9PR0Fi/+hd9++5m0tDSrvQ8REZHb0TCZ3FWNGjVZu3YLUVFRfPrpRwQG1uHQoQNs3LgegA0b1lOoUCHOnTtnfk1UVBTr1q2mYMEi5m2RkVd5/fVh7Ny5je7du3PlSjh+fppoLSIitqUwJPekUqXKACxevByA48ePMXBgX86cOf3/2xcBUKRIEWrXrsPatX/z999/4+7uaT5HUlISP//8IwArVqygVat2CkMiImJzGiaTB/Lww9XYtm0vjo6O5m2urq7Mnj2PF154BYCDBw/yzz9bAPD09LR4/eHDh1m16g+io6Puei2j0Uho6AWSkpKy8R2IiIiYKAzJfzJ16nSqVKnKqlXrOHr0DI880oTmzVsSGFiXlJQU0tPTMRgMvPjikEyvPXbsGL/8Moe//lrO6dMnSE1NJSEhIdNxQUFnmDXrK1asWGqFdyQiIvmNhsnkP+nZsxc9e/ay2GYwGBgz5gOeeKIzAN7e3jz99LNMnvwJYOolio+PZ+vWrezatYvKlSvTvn17rl69StGiRWnWrBX+/pXNjwlZunQxX3/9NdWqVaN160cpWNDXum9SRETyNPUMSY5o0qQZI0ea1hl67rlBlCtX3rxvzJjxODk5cePGDa5du8auXbsYP348M2fOZN++fWzdupGffppNVFQkAAsX/gqYepKOHDlw12vn5ANlb9y4QXx8fI6dX0RErE9hSHLMG2+8xY4d+xk58h0MBgOrVq3j448n0b//QNq165Dla44ePUZUVBSbN29m2LAX+fbbryzCzc6d20lJSbntNU1DatM4ffrEHWs7fvwYw4e/QkjIhXt+P6mpqdSvX4v69WuRmpp6z68TERH7pjAkOapCBX/zqtN169Zn4MDBGAwGhg4djoODA5UrV2Hp0pXUqFELgHPnzjJt2jTWrFnD33+vZvv2bVy7ds18vuDgIHOvUXj4ZS5dumhxvY0b15CensaaNSvv2EPUvXs35s+fx5AhL9zze4mICOfKlQiuXr1CRET4fbSCiIjYM4UhsYl69RqwZs0mFi/+g8aNm7Ju3RaLO9My/Pnnn+ZHfAAcPXqU+Pg4fv11Lk8+2YWOHduwZ89Orl69SlJSEsnJyeZjr169ctvrX7kSAcD27f/cc823BqCoqMx3wZ0/H8zKlX/m6DCdiIhkP02gFpupWTPA/LPBYODjjyfx9ddf0blzN3x9C7F7905WrfrT4jVHjhyhePESREdHcfq0aY2jd98dyeHDh6lVK4Bu3bqSmJjImTNnMBh+5aWXhmE0GgkPDyc11YCLS+aVtFNTU3Byuvsz08LDb4ahrIJWx46tuXr1KnPmzKdTpy733A4iImJbCkNiN/r3H0j//gMttv3110pefHEgjz/+JAEBgYwa9Rpr1vxtccz+/fsB2LdvLx06tGft2rXs2bOHRYsWERERRqlSJUlMTMTLy5tevZ7L1HNz5MghateuS1JSIgaDAy4uLiQnJ+Pi4mJxXFjYJfPPV69ezVR/xrbly5fccxgyGo0YDIZ7OlZERHKGwpDYtQ4dOnHiRBCurq6A6Tb77dv/wdXVlTfffJtPP/3IYmjs0KHDBAUFmX//9ddfcHR0JD4+ns6dO1OvXiOLVbEBTp8+gb9/JX777WecnJxwcHBh9OhRPP30s0ycOJlr16IpVKgwQUE3Hzdy7NhhwsNb8Pbbb1CqVGkWLVpg3pdVUMpKfHw8bdo0pUaNWnz33VyLfX/8sZSNGzfw8cefZQplIiKSvRSGxO65ubmZf54z52c2bFhH06YtKFq0KI0bN+XSpVB+/fVn1q79mz///MPitZGRkeafFy1aRJUqVShVqpzFMcePH8XR0ZHExAQuXbrEN9988//Xms2mTRsIDg5izpz5nD1780G0R48e5qOP3ufPP5dlqvfo0SP39L42b97IuXNnOXfuLImJsyze58CB/wOgdu1A+vbtf0/nExGRB6MwJLmKr28hnnyyh/n3evUaAKYJ0WvX3hw+K1++Ap06dWXGjGm4uLiQlJREeno6P/74I+fPn7c454oVK0hPTyc1NZUdO3ZY7MvoDerX71mqV69u3n79+jUiIiKyrDEy8ipBQWcpX97/ju8lYx0lMN1FV62a6fxpaWnm7aGhIXc8x/2Ii4sFwMvLO9vOKSKSF+huMskTunfvSbduT1C/fkMMBgMjRrzJ2LHjOX36AseOnWb+/PkAmYIQQFBQEDNmzOCbb77h0KFDADz++OOZjjt69Kj55/j4eM6dO3vbeubN+56LFy9w4UIQISHnWbVqGWFhl9i8eSNjxrxDXFysxeuPHDlk/jk8PMz8c1paeqZzJyQksG3bVou5T6mpqbz44kC+/HJqlvUkJyfTuHE9mjVrqDWSRET+RWFI8oQCBQry3XdzWbFiDcHBYTzzTG8AfHwK4OtbiB49etCiRSsAnJyc+PbbuRw7do6nnno607nq1q3HgAGD73i9kydPmp+jVrFixUz7//lnK4MG9WPMmLf544/FBAWd5c8/f6d376f5+uuveOutNzh9+pT5+FWrTMN7ly+HsmjRL+btly9fynTuyZM/4fHHOzFhwljztj17drFkyW9MnDg+y+e7nTlzmrCwy4SGXiQ09GKm/Rni4mKJibnOuHHv8fvvi+7YBiIieYWGySTPcXfPfPu8k5MTP/+8gO+//46mTVtQo0ZNwPSg2d27d3Lhws0eo7FjP6RRo0do1KgxO3Zso379hly8GJJlMPH3r8jWrXvw8ysIQLly5QkODjLf4QaQmJhIgQIFiI6OJikpEYCFC3+xOM/588FERIQzdeqnFrfwHz16mLS0NIs1mKZN+xyAL7+cSu/efalQoRLBwaZJ4ykpKRw4sI9HHmlicf4zZ24GrwsXzlO2bLlM7yU1NZV27VpYzI3q2LGLxVwmEZG8SGFI8g0PDw9efHGIxTZXV1eWLVtFREQ4e/fuxtXVjUaNHgHgu+9+5Pvvv2HQoJdISkrk+PGjFCpUmPbtTT1MTZs25/33J+Dg4MCaNZtYvnwpzz7bh8aN61pcY/369Xet7dKlSwwa9D+2b9/+r+2h7Nq1jTFjRlO/fgPef/9Di/0//zyX996bQHDwzTvddu7cnikMnThx3PzzoUP7adashcX+sLDLLFu2xCIIAWzduom2bdvftu60tDQWLJhPixatKFu2TKZ977//LmXKlGXQoJfu8O5FRGxLYUjyvZIlS1GyZCkCAy1DTNGiRXnrrdHm30uUKInRaGTw4JcoUuQhXn31dfMaQQEBgQQEBALw8stDuXTpEqNHv8/06V/www/fWZz31Vdf5YcffiAmJgYXFxfS09OJiorKFIQAoqOjefzxLhiNRg4e3E/16jUs9v/ww/cMHjzU3DMEsHr1SiIjIylTpjSOjgY6dnyMkydvhqGNG9fRr99APD29zPU//3w/du2ynDwOMG/enNuGoRMnjrNu3RrGjRtN9eo12bx5m8X+1atXMWvWDAAGDnwBBweNyouIfTIY9eyAO9q3bx9169Zl7drN1KpVO1vP7eTkgK+vJ9HR8aSmZp4oK9nDlu2cnp7Otm1befJJ0yKMZcqUY+fO/cTEXGfLlk0UL16ChQt/Ye7c7/+/VmdSU2//INoMJUuWJCUlhYiICNq378jGjetJSkrKdFyLFi0oXrw427ZtNw8F1qhRg+7du1OrVh2aNm3J9evXqFSpTKbXAvj6+jJnznymTfucvn37kZCQxNGjh+nVqy9Nm9YnPf1me/br9xxnzpxi2rSZlC5djnfeeZPvvpsFwOHDpylWrBibNm1g4sQJfPHFDCpXrnJ/jSlm+rfDOtTO1pFT7Xzo0AHatm3O3r17qVOnzp1ryLarikgmDg4ONG3anF9/XcLYse8wZcpXODo64utbiG7dngDAz684+/fv49KlUMaOHU+xYn4sXbqYU6dOsmfPrizP6+/vT+HChfn9999ZvXrVba+/adOmTNuOHDlCWloaKSkpPPJIU3buzNwjlCE6Opp+/Z7l2rVr/PPPFhITTXOedu/eaRGEAObO/QGACRPGMWvWD6xfv9a87/LlUIoVK0aPHo8BMHr0KBYuXHrb69qLv/9eRcWKlahQIfMkeRHJOxSGRKygdeu2tG7dNst9pUuXYe3azRbbWrZsTXp6OikpKSxevJCiRYvywgsDiY2NAWDYsNcpX96flStXWvQIFSlS5LYrYBcoUIDr168DcPz4cX766SfKlClv7pXy9fXFx8cn0/ID165dAzAHITDNS7qdEyeOExR0zmLpgDNnTjN9+jTz77cuH2Ato0ePYvv2bSxbtgovL6+7Hr9nzy769OkJwLlzl+7pNSKSO2kQX8ROOTg44OrqSq9efWnbtj2tWrUx72vevBWlS5dhzpz5vPnmWzz5ZA+GDBlOYODNruCAgACL833//U/06tWXxo2b4ujoxPnz53nrrTfNC00+/vgTjB07wXx8YGCg+ef+/Z+jSZNmt621cePG5p+PHz+W6W65ceNGs2zZEotj2rVrwbp1poUyT58+RUBAVcaNew+AK1euMHPmV3zzzQwuXgzh7NnTFucLCjp3X+slGY1GvvlmJocPHzQ/OiUqKpIXXniODRvWZfmazZs3mn+eMWNalseISN6gOUN3oTlDuV9eaeewsMuMGvU6AwYMMq+Z9G+XLoXy3HN96N27D926PcGJEycoU6YsoaEXqV+/ofm4n36aw2uvDTP/XqNGDb79dg4+Pr5Ur25aOXvOnJ/p3783derUYc6c+RQrVpzPPvuISZM+AUzDewYDVKlShebNmxMTE8OcOXPMPUn3aufOAwwY0JejRw////u8Rt++PVmzZrX5GEdHR776ahatWrVh+/ZtPPdcb4YPf4N33hlDSMgFpk2bgrOzE+PGfYTRmM6mTWupUKES/v6VASzmRT35ZHeSkpJZsWK5+fwRETEWNRmNRv73v2fMQ5A1atRi/fqt9/W+HlRqaipOTnfutM8rf9P2Tu1sHZozJCL3zM+vOHPnzr/jMSVKlGT16g3m3xs1amzefqs+ffpz6tQpvv76K5566mnGjfuQokWLAbB+/T94eHhQoYI/W7bsokIFf5ydnQHTSt8ZYWjYsBH873/PcfjwQYoV86NcuXJ4enry6aefkpJimgTu7+/P2bM3h8uef/55vvvO8u66l14aaA5CAIMG9bcIQmC6Tf+ll57HwcHBPFdp6tRJDB78Iv/737Pm11evXhN//wps3ryRYcOGMGTIcF555VVCQ0PN51qyJPNikiEhFyhd2hSWevR4jKCgc0RE3Fzv6ezZ06Snp9/2jriIiAjeeONVevXqS4cOnbI85lbx8fHs37+XJk2ame/oAzh16iSPPtqSRo0eYebM7/D1LXTXc2W3uLhYgoLOUbNmwN0Pljzv7NnTnD17hkcf7WjrUnKUhslE8qkPPviIPXsOM2PGt+YgBFCjRk0qVDD1DlWpUtUchADKlauAn19x3N3d6dSpKy4urtSt24BSpUxBYvz48UyZMoU6depQpUoVixW+nZyc8PPzY9iw4RQrVoxatWoBsG/fXou6/vhj6W1r/vek7WrVKloEqSlTPuPEiWMsXLiQyMhIxo17jwUL5nPp0u1X3Qb4/vtvSU1NJSbmOps2beDChfMkJiaaF7tMSEggJOTCbV+/YMF8/vprBf/73zMMGfICVauWY9++PYCpVypjrleG4cNf4cknu9C9ezeGD3+FGzduALBy5R/cuBHP+vVrmTjx5pDlhQvnzefLSdevX6NJk/q0adOMQ4cO5Pj1xP498khd+vTpya5dO+/rdZGRkbz++qvs3bs7hyrLXgpDIvlYmTJlLXom7sbBwYFly1bx118bMvU2ARgMBp5+ujeff/4l3347l1atbk4aL1OmDDVqBDB69AccPnyaESPetNjXsGFDi3N1797d/PMTTzxpsebTv5UqVQpPT08uXDjPyJFvEB0dbd733ntvsW7dGovja9eubfH79Olf0KxZA1atWmGxPTCwLn5+xQH47rtZxMbGcPDgfr7//luWLl3Eb7/9CsC5czcXq1y48BeioqLo3bsHsbExNGvWkLZtm5t7y1JSUszzp7Zs2cT8+fOYPv0LwHJi+sGDplXM09PTqVevJh06tGb69CksX76InJrd8PHH480rrR89eiRHrpEbBQcH2WTSP5j+tp57rg8HDuyzyfUz7N9/f2F8zJi3mTfvBzp2bHP3g+2AhslE5L6UL1/hjvudnJypVs30uBOj0cjEiZOJj4+nS5eulCt387WdOnWjbt16BAcH0bt3b5o3b8vLLw+iTZs2lCtXlkaNmnH27DmOHz/GqFHvUaGCP3/9tYIDB0wh4dYhs8DAQJKTk1m92jS85unpSbt27Vi1ahXXrl1j9uxvzNetWbMmXbt2pXbt2vj6+hIcfJ61a9dw9uwZhg590eK9lCpVkhs34ggLu8ysWdP59defuX79msUxbm5uFo87yWD6f8ZDCQu7DJhW8/bxKcAzzzyV6diMSd23hraMhTRv/X/Wx44dwWAwEhcXS0pKPLt37+eDD95n4sTJ1KxZ67afSUjIBX7/fZH5Ib/z5v2Kj08BwsPD8PT0Mt8pt3v3zaUc7tabliEuLpannupK2bLl+OabOff0mtwkODiYOnVqUq5ceXbuPHBf/+chOwwePIBDhw6wYcNagoPvLZAtWrSATZs2MHnyNFxcXG57XGxsDNevX6dUqdJZ7r/1TtW7zWP7t4wwn1soDIlIjjEYDAwYMOi2+1auXMe1a9G4uLjg6enFzp0HAFOIMhgM/PbbMmJiYsz/WE+f/g09ez5Br17/49VXX2fTpnX89dcKWrZsTeHCRTh9+jReXl58//1PpKenERsby7p1N+8Wa9OmDR06dCA+Pp5y5crh5OREgQIF8PevwNdff01cXJxFjQUK+FCwYEHz7/8OQgBvvfV6ptdlWLr0d/PP48aN5syZMyQnJ2c6LijoHJ999rHFtqioKLp0eZTChYvccn3T0gjLly/l1VeHmMPgY4915Nw507yoCxfO4+joSMmSpQDTMgbPPdfHYtjrt99+pUGDRnTu3I46deqxdOlKjEajxXIIhw8fyvI9/duiRQvZv38f+/fv45VXXjWvxD59+jTmzp3NwoVLKVeu/D2dyx59++23gCmcRkREUKxYsbu8IntlfG4ZQ6m3ExcXZw61L79s+t9c48ZNefbZPrd9TcuWjQkJucChQyfNPaC3ioqKNP/87yHqu7nf421Nw2QiYjMGgwFf30J4enpl2g7g41PA4v+1VqpUhX37jvHGG2/h7OxM27YdmDTpS7p0eYJHHmnG9u37WbNmC2XKlKNcOX8mTZrGyJFvmV9foYI/vXo9R4MGTejU6TFatDAN43l5edG5c2eLGry8vChVqhSBgYH4+flluhulbNmyODk5ceXKFRISEu76Xo8dO0ZycjLu7u4Wc6n+rX379uYvtV27drBq1Z/mfVFRUURGRjJ06MsWXzZxcbFERUWyc+cOmjVrQIsWjXjttaG88MJz1KxZOdP8n9DQUF58cQCJiYls27aVsLDLREREEB9/M9SdPx9sHo4zGo3s3LkjyzD4yy/zzD//+KNp4c2jR48wYcJYgoODWL165V3bJjv8+edyevZ8gsuXL5GQkEBaWpq59lWrVnDlypV7Pte3386kQYMATp48wfz5N29ayKoH8G6mTp1E3bo1LB4Gnd3Wr1+Dv39JZsz40iJsR0ZG3vY18fHx5nlwWT2KB7BYsywjiN8rhSERERsxGAwWwxglS5bmjTfe4bnnnqdAgYK8/PJwnJ1dqFevIeXK+VOlSjWefroPjRs35/33b/bMPPXUU7z44otUqFCRp556hg8//JiXXnrFvL9Jk6asX/8P48d/bHHtZ599loIFC/LMM8+YtxcpcrNnx8PDgyFDhvDMM8/w6qsjqFevAR063LxLp02bNrz00hDc3d2zfH+HDx/myy+/zHJf1arl6dr1URISEoiJieGnn+by+++Lszz2q6+mcvr0zbWbNm5cT1DQWYtjoqOjuHLFdEfdjBlf0rXro4wa9Tpg+vJ97bWh7Ny5g/37b85l+eOPZaSmpjJy5AhzGHnvvbfp0eOx2/aeJSYmEh4enuW+e5WSksKAAX3YsGEdAQFVKVu2GE2b1icpKYnvv/+Gfv2eZfDg/vd8vnffHUVwcBCPPFKP4OBg8/YzZ05nefy0aZ8zduy7mQKA0Wjko48+ICTkAlOnTnqQt3ZPXn55EEajkffff5eLF29O9E9PT8t0bHp6OlevXuX06ZPmbRmf1b/d2jN07Vp0lsfczq1tkTFfzp5pmExE8ryJEyczceLkLOd7FClSlCJFigIwf/5v/z/cMwwHBwfc3EyhpEYN023m58+HMH/+PCZNmoa3tw8DBgzm2rVrfPbZxzRq9AgdOnSmShXTM9fatGnH4cMH6dOnD9euXePy5cs0b96SpKQEjh49hK9vAbp06WSx4nfduvVp1KhZpuDQqFEjduzYkeXz58By/tS9cnR0xMfHh+joaIYNe4mKFSsB4O7ubg5UISHneffdt8wTvpcs+Y3q1WsyfvwYAH76aS4A9es34ODBA1y7Fk23bh0yPUZm06YNLFq0gP79B5q37d69k0mTJhIdHcWRI4dZvXqDxe38+/fvZcCAvtSsGcDkydOIibnGkiWLKFSoMO3ateeZZ56kbNlyfPjhJxw7dizT+zt79gxHjhziyy+nAvDPP1vMw69gugMwIiKCoUOHEx8fz5gx7/D440/SuHHT27bZiRPH+PnnH/Hx8aFduw64uLhw6VIoEya8D0Dz5i1o0+ZR8/G3PkD51kn99+rfE+UzhsIytme8l1uPuzW83bo8RIavv57O+++/S5cuj5m3XbkSkeX1IyNv9gzdb/23LooaFRV12+HFuLhYfvnlJ/r374O7e4H7ukZ2UhgSkTzvXie9tm3bnrZt2992/7BhrzFs2GsW53399VE8//wLeHv7YDAYOHXqOFFRVxk48GXCwi5z5Uo4RqMRP78SFC9egn37dhESct58d1KFChUoUMD0JdC7d38cHBwYMeJNPvroA1q2bElAQABOTk7mlcIdHZ3o3v0pFi5caP4SfOWVVwgPD6ds2bIcPXqMlStX8G8VKlSgVq1aLF26FDBNJG/UqBFff/01cLPXo3z58hw7doy4uDg6dmyX6TwZQehWVapU5vr1a5w6dcochLp372meGA4wc+aX/PHHUmrVqs2YMR/QubPlub/44nPeeWcMTk5OxMbGMmnSREJDLxIaehFvb2+OHj3CsWOmO9zefvsNc83/vlPwVvv377Xo3Rgz5h1q1KhJhw6dzJPlW7Zsxc6d25k37wfmzfuB9ev/yXSehx+uxvHjxywm4kPmEPr9999ahKFb7w48efK4+TMvVszvtjXfKioqyuL3sLDL7Nq1g3HjRlOwoC/r1m3By8vb4phbA1hWPW7vv/8uAH/+ucy87dYhxKioSKZMmcTAgYMtwlBWQ6S3Mq3yPgMHBweef/5Fi9dGRUXeNgz9+OMc3n//XaKjr/DOO+/f8Ro5SWFIROQ/KlCgoPnnKlWqmX8uW7Y8ZctaTh5u0KAJDRo0ITExgStXIihVqgytW3cEjBQvXgKAV199nUcf7Uh09FWOHDlAeno6/v6ViIy8wqxZP+Dp6YaHhwcLFiygU6dOFC5cmMKFC+PrW5jq1Wubw1Dbtm0pX748QUFB1KlTB29vb3bv3s3Vq1dp0aIFAwaYhgJHjnzDXF/z5i0telocHBzo3Lkz//zzT6YvZ4DSpUtToUIFrl69yqlTpjk1P/zwMw0bPmIRhoKCzhEUdI4tWzZZPOokw4ED+2nTppnFvKUMGUsY/JvBYLDoFalXrz5paWkULVqM1atX8cMP31k8U2/WrOkAtGt3M/AeOnSQ06dvzgV6+eXnAXB2dsbT04v4+DiGDRvBSy9lvhHg371xa9asZsqUz3jppaG4ubmxY8c2874zZ05Tv34t3N3d2bp1Dw899BBRUZEULOhrXswzY75Pxh1gwcHnLM5/+vQp3nrrdRITE4mOjmbRooX07z/QIuzfuszDrcsBRERE4OGR9fDrrT1D7747isWLF7JkyW/07dvPvP1uPUNjxrzNrFkzAKhZs7bFPLpbgxGY7pB0cHAgMLAux48fBbC4UcEWFIZERGzAzc2d0qXLAlCpUmWLfQaDgWrVqgNQr17D/58T8zJpaWl4eHiQmprE1aumnqCSJUtTr15Drly5QrVqNTEajTz6aAdKly5D48aPEBUVSdmy5ShXrgJNm7aiQIGChIRcoFgxP4oWLcYTTzzNli2b+eOP5QQEBPDaa6NYu3YNZ86cpnjx4rRp04aKFSty/Phxcxjq1+859u/fx5kzp+nSpQsODg506tSFS5cuUa9efcqXL8fff/9x2/d++PDBTNsuXAjO8tgqVapy8uQJwLQuVrFiRbl+/Tr169dn2LA3aNjw5jP0OnbsgKOjI8ePHwewCDm3unWF83379lr0ppw4YXpt+/admDjxM4zGJEqUKMe6dWtZtGgBjRo1tgg5lSpVJiIigm7dnmDevB/4+OPx/PzzPD766JNMDzROTEwkMTGR6tX9zQ9VLlbMj1mzvueRR5rw9NOPs3//Xj744GNq1KiZaY2e2bO/sQh3c+bM5n//e46UlJtDUjt23LxmxrIOq1atYMCAPrftkZo/fx4eHh7cuHGDxYsXAqaAFBISYj4mMvJqlo+KuXgxhA8/HGd+HcBXX02xOOb06VM0adKMqVMn8dtvv5p7ISdMmMjZs6bwljG8bCt6Ntld6NlkuZ/a2TrUztaT0dahoRE4O7vd19o36enpxMXF4uHhgZOTaXXxlJQU1q9fS/PmLXF3d+fEieN89910KlS4uS7UlStXWLZsGS1btqRixYqkp6eTlpaGs7Mzbm5u9OjRh59//sFi0u7SpUs5duwYnTp1Mg/Pde7cFQcHR0qUKM7cuT9YfLnfysHBgR49elK/fiPeeONVwDQnp3Xrm8/lc3R0ZNy4caSmpmIwGBg7diwAMTExfP755+bjXnllKL/9toBGjRqxevXq2869atjwEXOAGTHiDd57732Cgk5y7NgJzp8PIikpmRIlSvLnnytYsWI5w4e/wVtvjSYpKQk3Nzdmz57FtGlTzCEEMibW92H+/HlZXvNBPPfc8yxY8As3bsQzZsx4PvjgvSyP8/DwYPnyv+jcud1t3/Od/HsY0MXFhdat2+LvX4m5c79nxIg3Wbfub7ZtMz23r3TpMrddqb169ZoWq8Wb6vPE0dHx/xczPUjp0v42ezaZwtBdKAzlfmpn61A7W4812nrfvl2cOHGUQoUK88gjzfDw8OTEiaNcuhRKWloqKSkp+PtXxsPDg6JF/fDy8mbLlg0cPmxabK9ixSpUrFiZLVs2EBcXy44dO/Dw8KBBgwb4+ZXg/Pkgrl69yrFjx6hSpQq//fYbKSkpvPTSS7i5uZGSkoKzszNlypRnwADTcE27du1o0qQJxYr5mYeAgoKCWLp0KZ07d6Zy5cr4+hYiOjqKLVu2sG7dOkqWLMWbb47k2jVTr9aOHTtYs2ZNlndQnTsXyrx5c1m+/HemTp1OqVIlmDv3u0zHJSUl4edXmu7de5ofV2M0Gjlz5iTp6UZ++mmuecioSJGHWL78Lxo3rgtAv34D8fHxYe3av2nZsjXLli3h0qXQTNcA04Keb701GhcXF+bNm8OlS5dwdXVlxYo1rFjxh3n+z7+1bt2W9evX3vUzfvjh6uZhqooVK932brl7ERAQyC+/LKZ16yYWYfBexcfHk5Rk1INaRUTEftSp04A6dRpYbKtZM5CaNQNv8wpo2LAxrq6uVKxYhUKFCgOmeVMpKak0bNiUXbv+ISkpifPnTcNSxYoVo2rVapQuXYYKFfyJibmOp6cnPj4FuHr1Cunp6Vy4EETfvn05duwY9evXp1ChIjRv3pbffvsJV1dX+vV7nmbNWnDx4gXatu1I5coPs3OnaRJ0hQoV8PLyMgchMN2ZV7t2bYxGI4cPH2blypX/X4sfa9euokOHDtSrV5eNG1dbBCY3N3cSE03zYFxdXfHx8TIHNqPRyNGjh9i82bTAZ9OmTZkz53uSkhJp3rwFFStW4rXX3iQkJITXXnsTT09P3ntvHACPPfYE06Z9zrZt/2AwwFdffU1UVBSpqWk0adKM6OhItm/fTM+eT9OhQ1fKl6+IwWBg8OCXOHv2DPPm/ZDpcxg37iMOHtxvXmeoTp26TJ/+DTNmfEnbtu0pVao0K1Ys46mnetKkST3ANOF9x45tbNy4/l7/RMwCAgJZs2YTAD16PMOXX5qGyTw8PLlxI97i2Hr1GvDss31Ytux3Nm82PVS6ZMlSeHh4kJRkeaw1qWfoLtQzlPupna1D7Ww9ubGtU1JS2LJlPUlJSaSkpFC2bDlq1gw0Tx5OT08nIeEG7u4e5m0XLgRx8OA+ihcvyZUrEVy9GkHnzk9QqFBhwsPD8Pb2xsPDE6PRaL7tPGPIcPPmdRw5Yjk3qUGDJuza9Q8FCxbi+vVojEYjKSkpnDlzjtKlS+Pl5ZFl7e3adaREidKsWbMCX9/CHDt22Dxxu337Lhw+fCDT40tiY+M4c+Ysgwa9RHJyAnXrNuLatSiWLl2IweBAixZtePjhGqSlpbJs2SIuXw6lUqWqnDljWv/n6af7UqTIQyxdusDcc1SpUlVat25PZOQVvLy8cXV1o3HjugQFWU60PnDgGHPnzmbKlMkArFmzybwy+K0SExMpU8a0rMSyZauoUMGfkSNfo1+/5wgNDWXmzC+ZMmU6Xbua7pD79dclODs707PnE3Tt+hjnzwezb99eJk6cbF5p/vz5YFq1akKNGjWZOfM7vvhiMi+88DJHjhxm/fq1fPTRZ3h6ejJx4ng+//wzAFq1asP69Wuz/e9Zw2S3ER4ezrPPPsv69feefBWGcj+1s3Wona0nv7b1resE3c21a9EsXDjPvN5NzZq1adrUNN/IYDAQFxfLyZPH2blza5avd3V1pWzZ8iQnJ9Kx42MYDI7mfZs2rePo0cyTwEuUKEmjRs3YuHEtUVGWd1D5+BQgNTXV3FNiMBgoWbIMFy9mvTJ17dr1qF//EWbPnm6et+Pk5ISXl7fFAoiJiUls2bKVBg0asmHDOj744GP27dvx/+/Bg5Ily9C+fUeCgs5StGgxvLy8SUtLJTo6ioIFC/HZZx8TFRXJJ598bg6hqakphIRcoGzZ8jg4OPDHH8uIiAhnwIBBGAwGoqIi8fEpQHJyMrt27aBZsxakpqaQmpqKp6cXsbExuLi44urqetvPZ+PG9Tz99OMArFy5ho4d29o0DOWbYbLt27czbty4+1qSXURE7Mf9TBQvWNCXXr2ew8XFhZSUFNzc3C1e7+XlTUBAHS5cCOLyZVPPS9GixfDw8OLChSDatOlIxYoVswydzZq1omzZ8qxcudS87bHHulOyZBkAevTozd69O9m3b5c5yMTEmB5n4enpyUMP+REcfPa2QQjg+PHDpKamkJ6ejouLK8nJSaSmpmZaCdrNzZV27Ux3nT3++GPmIARQqlQp2rfvxLFjh9m4cQ2FChWhUSPTcOXVq6bvwgIFPClatDApKcm4uroBsHPnNg4e3Evt2vX+/zb/VPr0+Z+5/TKGQBMSbuDj40VSUiLLly8mNvY6zzzTD29vH9LSUjl+/AhlypTL9LgdgBYtWjFt2kxq1apNrVo1b9sO1pJvwtDixYuZOnUqPXr0sHUpIiJiBRkLErq4ZN1D4eTkxOOPP82lSyEkJCRQqlQZ3NzcSU1NMd9plxUHBwfzUgVnz56iUaOmFC9e0rzf0dGRBg0aU6FCJUJDQ/DzK86JE8dwdHSgTp0GuLi48s8/mzh27BBVqlSjVq06XL8eTUjIBerXb8Ty5Yv+f2VuU+9TtWo1OHRovzlYNWrUlB07TD1aPj4FzEHr386ePcXKlcsICQkGICrqqkWAy5CUlMiFC8F4enqRnJzMwYN7AThwYI/5mGXLFtG6dXuKFHkIMA1rrlixlKioq+ZaMl5TtKgfu3dvJybmOn5+JXjiiZ6ZgqzBYOCZZ3qTlpbGqVMnqFWrGraUb8LQpEk591wYERHJnTKGq251pyB0q1q1AqlV6/YTyosUecgcHooVs3wqfIsWbWjatCWOjo7mY/39TetNdenyFBs2rCY2NoZKlapSr14jPDy82Lt3B+3adaJUqbKEh4fh5ORImzYd+euvPwgOtny2nKOjI2lpaZm2A5Qv70/t2vXZsOFv8+TyNWvu/EDdq1cjWLhwHhUqVMLNzZ0rV8IzDQUCHD58wOL3sLBLhIaGUKCALw4Ohky9RNu3b+bQof3Ex18nMLDhHWvISXkqDC1fvpypU6dabGvbti3vvPOObQoSERG5jYwg9G/e3t5069bdYlvt2nUJCKhj7mHp2LGbed+jj3bm3LnTrF27CjD1Fj39dF9CQ0OIj4/DaDTy0ENFOXv2NBUrVjYHsx49enP+fBB///0nYFp1+98PVTUYDLi6uprXgzp37rTFvjp1GrB3706L1/j6FsbJydG8svXy5YvM+4oVK46HhwfOzi5cuxZNRIRpiYTy5S1Xare2PBWGunXrRrdu3e5+oIiISC5zuzlTTk5OVK78MC4uruzfv5sWLdri4uJC+fL+Fsf5+ZWw+N3Z2ZkKFSpSpkw5HBwcaNnyUWJjr7Njx1bq1WtEoUKFSUtLw8vLm8uXQ/n9d9PjVcqUKUeZMuXx96+Ep6cXpUqVISIijCpVqpOWloq3tw8AiYkJ/Prrjxa314eHZ16DqGLFypQvX57oaNvdWp+nwpCIiEh+Va5cBcqVq3D3A2/h4OBAly5Pmn/38PDgsccyz60tXrwkXbs+hbu7O0WKFLXYV7JkaUqWLJ3pNW5u7nTs2I1Dh/ZTo0YAe/fu4urVCAICTHd2ubt7ULCgL8WLF8/0WmtTGBIREZG7yniW3v0oVqw47dqZwk7nzo8DmXu4HB0d/nNt/9V9V5CSksKUKVNo2bIlgYGB9OrVi3379mV7YTNmzKBv374W29LT05k2bRrNmjUjICCAAQMGcP787W9NzMrhw4fvfpCIiIhkK4PBcF/LI1jTfYehmTNnsnjxYiZMmMDSpUupUKECgwYNIjw8PNuKmjNnDtOmTcu0fcaMGfz6669MmDCBBQsWYDAYGDRoEMnJydl2bREREclf7jsMrVu3ji5dutC0aVPKli3LW2+9RVxcHAcOHMh0bGRkJEeOHLHYtm/fPuLi4rI8d3h4OM8//zxffPFFppnlycnJfP/99wwdOpQWLVpQtWpVpkyZQnh4OGvWrLnftyEiIiICPEAYKliwIBs2bODixYukpaWxYMECXFxcePjhhzMd++OPP9KvXz/zMNqWLVvo378/y5cvz/LcR48epUCBAixfvpyAgACLfSdOnCA+Pp5GjRqZt/n4+FCtWjV27959v29DREREBHiACdTvvvsuI0aMoE2bNjg6OuLg4MAXX3xBmTJlMh376quvEhYWxsCBAxk8eDAzZsygf//+9OrVK8tzt27dmtatW2e5LyzMtBbBv2edFy1alMuXM9+qJyIiInIv7rtn6OzZs/j4+DB9+nQWLFjAk08+yahRozhx4kTmkzs48PHHHxMYGMjUqVPp0KEDr7/++gMVmpCQAPD/z0m5ydXVlaSkpAc6p4iIiMh9haHQ0FDefPNNXn/9ddq2bUvNmjX54IMPqFy5Ml9++WWWr1m/fj27du2iWrVqrFmzhh07dmR53N24uZkeIPfvydJJSUm4u7s/0DlFRERE7isMHTp0iJSUFGrWtHzCbEBAAMHBwZmOX716NcOHD2fw4MEsWbKEbt268cILL7B9+/b7LjRjeCwiIsJie0REBH5+fvd9PhERERG4zzCUEUhOnjxpsf3UqVOULZt5MSYXFxeGDBnCsGHDMBgMjBs3jt69e5t7ee5H1apV8fLyYufOm89AiYmJ4dixY9SrV+++zyciIiIC9zmBulatWtSrV49Ro0YxduxY/Pz8WLp0Kdu3b2f+/PmZjm/VqhWtWrUy/24wGBg5cuQDFeri4kKfPn2YNGkShQoVomTJknz22Wf4+fnRrl27BzqniIiIyH2FIQcHB2bMmMHUqVN5++23uX79OpUrV2bOnDnUrl07h0q8adiwYaSmpjJ69GgSExOpX78+s2fPzjSpWkRERORe3fet9QUKFGDs2LGMHTs2J+oxmzhxYqZtjo6OvPnmm7z55ps5em0RERHJP2z/dDQRERERG1IYEhERkXztvofJ8puMxR5Pnz55lyPvn6OjAz4+7sTEJJCWlp7t5xcTtbN1qJ2tR21tHWpn68ipds743s74Hr8ThaG7yFg/6aWXBtm2EBEREblvwcHBNGnS5I7HGIxGo9FK9eRKV69eZfXq1ZQrV04rXYuIiOQSCQkJBAcH0759e4oUKXLHYxWGREREJF/TBGoRERHJ1xSGREREJF9TGBIREZF8TWFIRERE8jWFIREREcnXFIZEREQkX1MYEhERkXxNYUhERETyNYUhERERydcUhkRERCRfUxgSERGRfE1hyAbS09OZNm0azZo1IyAggAEDBnD+/Hlbl5WrzZgxg759+1psO378OH369KF27dq0bNmS2bNnW+zX53Bvrl27xpgxY2jevDl16tTh2WefZc+ePeb9aufsExkZyZtvvkmjRo0IDAxk8ODBnDlzxrxfbZ39goKCCAwMZMmSJeZtaufsERoaSpUqVTL999tvvwF21s5Gsbovv/zS+Mgjjxg3btxoPH78uHHAgAHGdu3aGZOSkmxdWq70ww8/GKtUqWLs06ePeVtUVJSxYcOGxnfffdd45swZ46JFi4w1a9Y0Llq0yHyMPod789xzzxm7detm3L17t/Hs2bPG8ePHG2vVqmU8c+aM2jmb9ejRw9izZ0/joUOHjGfOnDEOHTrU2KRJE+ONGzfU1jkgOTnZ+OSTTxorV65sXLx4sdFo1L8d2WndunXGmjVrGsPDw40RERHm/xISEuyunRWGrCwpKckYGBhonD9/vnnb9evXjbVq1TL++eefNqws9wkLCzMOHDjQWLt2bWOHDh0swtDXX39tbNasmTElJcW8bfLkycb27dsbjUZ9DvcqODjYWLlyZePevXvN29LT043t2rUzTp06Ve2cjaKioowjRowwnjp1yrzt+PHjxsqVKxsPHjyots4BkydPNvbt29ciDKmds8/MmTON3bp1y3KfvbWzhsms7MSJE8THx9OoUSPzNh8fH6pVq8bu3bttWFnuc/ToUQoUKMDy5csJCAiw2Ldnzx7q16+Pk5OTeVujRo0ICgoiMjJSn8M98vX15ZtvvqFGjRrmbQaDAaPRyPXr19XO2cjX15fPP/+cSpUqAXD16lVmz56Nn58fFStWVFtns927d7NgwQI++eQTi+1q5+xz8uRJKlasmOU+e2tnhSErCwsLA6B48eIW24sWLcrly5dtUVKu1bp1ayZPnkzp0qUz7QsLC8PPz89iW9GiRQG4dOmSPod75OPjQ4sWLXBxcTFvW7VqFRcuXKBp06Zq5xzy3nvv0aRJE/766y8+/PBDPDw81NbZKCYmhpEjRzJ69OhM7aV2zj6nTp0iMjKSXr160bhxY5599lm2bNkC2F87KwxZWUJCAoDFlwuAq6srSUlJtigpT0pMTMyyjQGSkpL0OTygvXv38s4779CmTRtat26tds4h/fr1Y/HixXTr1o1XXnmFo0ePqq2z0fvvv0/t2rXp2rVrpn1q5+yRnJxMcHAwcXFxDB8+nG+++YaaNWsyaNAgtm/fbnft7HT3QyQ7ubm5AaY/lIyfwfThu7u726qsPMfNzY3k5GSLbRn/A/Lw8NDn8ADWrl3LG2+8QUBAAJ9//jmgds4pGUML48eP58CBA/z0009q62yydOlS9uzZwx9//JHlfrVz9nBxcWH37t04OTmZA02NGjU4e/Yss2fPtrt2Vs+QlWV0+UVERFhsj4iIyNRlKA/Oz88vyzYGKFasmD6H+/TTTz8xdOhQmjdvzrfffmv+x0ntnH0iIyP5888/SUtLM29zcHDA39/f3F5q6/9u8eLFREZG0rJlSwIDAwkMDARg7NixdO7cWe2cjTw8PDL17FSuXJnw8HC7a2eFISurWrUqXl5e7Ny507wtJiaGY8eOUa9ePRtWlrfUr1+fvXv3WnyxbN++nfLly1O4cGF9Dvdh/vz5jB8/nt69ezN16lSLf9zUztknIiKC119/nV27dpm3paSkcOzYMfz9/dXW2WTSpEmsXLmSpUuXmv8DGDZsGN98843aOZucOHGCwMBAizXJAI4cOULFihXtr52z/f40uavPP//c2KBBA+PatWvNayc8+uijWqPiPxg1apTFrfVXr1411q9f3zhq1Cjj6dOnjYsXLzbWrFnTuGTJEvMx+hzu7ty5c8bq1asbX3nlFYt1QiIiIowxMTFq52yUnp5uHDBggLF9+/bG3bt3G0+ePGkcMWKEsX79+sbQ0FC1dQ669dZ6tXP2SEtLM/bo0cPYpUsX4+7du41nzpwxfvTRR8YaNWoYT5w4YXftrDBkA6mpqcZPP/3U2KhRI2Pt2rWNgwYNMoaEhNi6rFzt32HIaDQaDx48aHz66aeNNWrUMLZq1co4b948i/36HO5u5syZxsqVK2f536hRo4xGo9o5O8XExBjHjh1rbNKkibFWrVrGAQMGWKw7pLbOGbeGIaNR7ZxdIiMjjW+//baxSZMmxpo1axp79uxp3L17t3m/PbWzwWg0GrO/v0lEREQkd9CcIREREcnXFIZEREQkX1MYEhERkXxNYUhERETyNYUhERERydcUhkRERCRfUxgSERGRfE1hSERERPI1hSERERHJ1xSGREREJF9TGBIREZF87f8AhIFCmJ0OwNcAAAAASUVORK5CYII=","text/plain":["<Figure size 640x480 with 1 Axes>"]},"metadata":{},"output_type":"display_data"}],"source":["# Plot the learning curve\n","plt.semilogy(history.history['loss'], color='gray', label='training loss')\n","plt.semilogy(history.history['val_loss'], color='black', label='validation loss')\n","plt.legend()\n","plt.savefig('figures/learning_curve')\n"]},{"cell_type":"markdown","id":"61b23dfe-23f2-4ed0-9767-f9f92b884a0f","metadata":{"id":"61b23dfe-23f2-4ed0-9767-f9f92b884a0f"},"source":["We expect both validation and training loss to decrease. If the validation stabilise or increase while the training loss continues to decrease, it means that the neural network overfits on the training set. In that case, the training can be stopped earlier by reducing the number of epochs (it is called \"Early stopping\")."]},{"cell_type":"code","execution_count":null,"id":"2817e9eb-9d9b-4dfa-80bc-def5f7de70c2","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"2817e9eb-9d9b-4dfa-80bc-def5f7de70c2","outputId":"88925f5d-87f0-49e9-c121-2bfdeb9e989d"},"outputs":[{"name":"stderr","output_type":"stream","text":["WARNING:absl:Found untraced functions such as _jit_compiled_convolution_op, _jit_compiled_convolution_op, _jit_compiled_convolution_op, _jit_compiled_convolution_op, _update_step_xla while saving (showing 5 of 5). These functions will not be directly callable after loading.\n"]},{"name":"stdout","output_type":"stream","text":["INFO:tensorflow:Assets written to: Output/damodel_nn/assets\n"]},{"name":"stderr","output_type":"stream","text":["INFO:tensorflow:Assets written to: Output/damodel_nn/assets\n"]}],"source":["# Save the model and the standardisation coefficients\n","if learning=='analysis':\n","    name = 'damodel'\n","else:\n","    name = 'true_model'\n","model.save(outputdir / f'{name}_nn')\n","dump(scaler_x, open(outputdir / f'{name}_scalerx', 'wb'))\n","dump(scaler_y, open(outputdir / f'{name}_scalery', 'wb'))"]},{"cell_type":"markdown","id":"973230a5-686a-4baf-9a55-1966c3a8e517","metadata":{"id":"973230a5-686a-4baf-9a55-1966c3a8e517"},"source":["## <span style=\"color:blue\"> IV. Validate the model (offline).</span>\n","In this section, we apply the neural network to the validation dataset and see how well the neural net has been trained."]},{"cell_type":"code","execution_count":null,"id":"15a7cd0e-3d94-41e7-9c5c-4fa1c6c30454","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"15a7cd0e-3d94-41e7-9c5c-4fa1c6c30454","outputId":"c77f0073-2ad5-4f41-8750-a31eb228fa78"},"outputs":[{"name":"stdout","output_type":"stream","text":["32/32 [==============================] - 0s 736us/step\n"]}],"source":["ypredict = scaler_y.inverse_transform(model.predict(scaler_x.transform(Xval)).squeeze())"]},{"cell_type":"code","execution_count":null,"id":"effce0c6-37d2-4ae4-b127-24eb04e9e826","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":413},"id":"effce0c6-37d2-4ae4-b127-24eb04e9e826","outputId":"6852400c-3d16-4423-942e-d5f651f6feb0"},"outputs":[{"data":{"image/png":"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","text/plain":["<Figure size 640x480 with 1 Axes>"]},"metadata":{},"output_type":"display_data"}],"source":["ax = scatter_plot (yval[:,20], ypredict[:,20])\n","ax.set_xlabel('target error')\n","ax.set_ylabel('predicted error');\n","ax.axis('equal');\n","\n"]},{"cell_type":"markdown","id":"09266f98-55bc-4654-8016-3cfb3c983247","metadata":{"id":"09266f98-55bc-4654-8016-3cfb3c983247"},"source":["There is a significant correlation between the target error and the predicted error (around 0.18), which is what we hoped for.\n","The neural network is supposed to learn the model error: if the analysis is used, the model error is estimated and therefore noisy because of the noise in the observation and the uncertainty of the analysis. So ideally, we don't aim at a correlation of 1, since it would mean that the neural network has fitted also the noise of the data. On the contrary, if the neural is trained using the *true* model error (`learning=truth`), you aim at the best possible correlation."]},{"cell_type":"markdown","id":"lVGH6ZUK1_Hr","metadata":{"id":"lVGH6ZUK1_Hr"},"source":["## <span style=\"color:blue\"> V. On our way to part III  </span>\n","\n","If you are under Colab or kaggle, click here to continue with part III: \n","\n","[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg) ](https://colab.research.google.com/github/ecmwf-projects/mooc-machine-learning-weather-climate/blob/main/tier_3/data_assimilation/mooc_ecmwf_bocquet_brajard_part3.ipynb)\n","\n","[![Kaggle](https://kaggle.com/static/images/open-in-kaggle.svg)](https://kaggle.com/kernels/welcome?src=https://github.com/ecmwf-projects/mooc-machine-learning-weather-climate/blob/main/tier_3/data_assimilation/mooc_ecmwf_bocquet_brajard_part3.ipynb)"]},{"cell_type":"code","execution_count":null,"id":"8168ba9c","metadata":{"id":"8168ba9c"},"outputs":[],"source":[]}],"metadata":{"colab":{"provenance":[{"file_id":"https://github.com/ecmwf-projects/mooc-machine-learning-weather-climate/blob/main/tier_3/data_assimilation/mooc_ecmwf_bocquet_brajard_part2.ipynb","timestamp":1681336718165}]},"gpuClass":"standard","kernelspec":{"display_name":"Python 3 (ipykernel)","language":"python","name":"python3"},"language_info":{"codemirror_mode":{"name":"ipython","version":3},"file_extension":".py","mimetype":"text/x-python","name":"python","nbconvert_exporter":"python","pygments_lexer":"ipython3","version":"3.10.0"},"widgets":{"application/vnd.jupyter.widget-state+json":{"013c2ca16a9f43bbbde8c9f746837eda":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"09ba5482356046c1905faafcf7f77c9d":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"0c091879d3ba48edb9b762aff8c5cd92":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"0ca41fb9aee44015b0e1112daa393634":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"0dfef0c43c87493d803b6849ecf7fa11":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"10bd83a61fce44e9b951d9f72cee9c1e":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"125c84f09fcb4666b79236f179d19827":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"12762e6f19e44be9b4824d54b97acd15":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"19c75562bce7421e8da187e97c4b08c3":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_98ccbf5dfc0e48d29b21891e60ab6c4c","placeholder":"​","style":"IPY_MODEL_6fa25cae71b9497da2a408cbb2e44ccf","value":"true model integration: 100%"}},"21bd22f8052b49b98acbb98a922d7d52":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"2233d06a6bc949c1b21917875d699b85":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"2265c62e2c654e25a51d7bba6d6744f9":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"23ad55be58a045ea82d6aa18519d6739":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_21bd22f8052b49b98acbb98a922d7d52","placeholder":"​","style":"IPY_MODEL_09ba5482356046c1905faafcf7f77c9d","value":" 500/500 [00:00&lt;00:00, 807.95it/s]"}},"2403d9c27b92416595809f8851b7e786":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"24eb5bf591a3440b92ef49cba0b848ff":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"FloatProgressModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_013c2ca16a9f43bbbde8c9f746837eda","max":500,"min":0,"orientation":"horizontal","style":"IPY_MODEL_670eb83f4a3047f38456fd21ab792418","value":500}},"2b2a5f4feec34418a815cbaddac8c9e6":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HBoxModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_19c75562bce7421e8da187e97c4b08c3","IPY_MODEL_e48505496f2142ecac7cc2f43105dfe2","IPY_MODEL_8153b322cbc9499a9444429bb39e797c"],"layout":"IPY_MODEL_0dfef0c43c87493d803b6849ecf7fa11"}},"320eba1b9795450db713503c124ef4c8":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"35b724c3fb1f4517ae9e33f151b578e9":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"38b06265149143a09648875f5a435782":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_2233d06a6bc949c1b21917875d699b85","placeholder":"​","style":"IPY_MODEL_82a0a5bcffbd431cb12359259e78b76e","value":" 500/500 [03:53&lt;00:00,  2.28it/s, mse=0.796, val_mse=0.813]"}},"3acf1cac85e9490483a4ef865289b917":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ProgressStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"3f80dd8cc17842b0b542dffa7d0b1f8b":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"46df280ab0e1494aada1d25fd0c6cd93":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"4a2e89f1f6bf4eedbca0ad25d5a04de0":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_2403d9c27b92416595809f8851b7e786","placeholder":"​","style":"IPY_MODEL_69b622f0565f42e8a16844d3d6fef28b","value":"physical model integration: 100%"}},"5c276d02f53b4cb88de8b336dc660cfd":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"FloatProgressModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_0ca41fb9aee44015b0e1112daa393634","max":500,"min":0,"orientation":"horizontal","style":"IPY_MODEL_da7f410b46fa4574adf21b5b7c1be74d","value":500}},"5e313cac6dd94aac9ec6e55789cc5136":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HBoxModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_9fb61e3b0e6949aea6f4a6b7154f4ce0","IPY_MODEL_95a89ac1b6924365b4bccff08a482f49","IPY_MODEL_f8cab4150ab040aeb033f64745faad74"],"layout":"IPY_MODEL_46df280ab0e1494aada1d25fd0c6cd93"}},"608345d3c72442538d04f01665b690e9":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ProgressStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"63ab33992e6747789c6aa30d7fd3ca7b":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"FloatProgressModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_125c84f09fcb4666b79236f179d19827","max":500,"min":0,"orientation":"horizontal","style":"IPY_MODEL_ec55d4f2783d4264a14534273656a4a5","value":500}},"670eb83f4a3047f38456fd21ab792418":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ProgressStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"675634702dbb4e31ba310b6cfb154ab6":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"6887d263554c4ce2a3e629693d29825b":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ProgressStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"69b622f0565f42e8a16844d3d6fef28b":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"6fa25cae71b9497da2a408cbb2e44ccf":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"747d3c6ccb794e67bfc262ce031f9fc5":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HBoxModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_973fca2dd29b4e0881bace37d1eca805","IPY_MODEL_24eb5bf591a3440b92ef49cba0b848ff","IPY_MODEL_38b06265149143a09648875f5a435782"],"layout":"IPY_MODEL_cb7abe9177294838930daa210455aef3"}},"8153b322cbc9499a9444429bb39e797c":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_f966d3ecfee44e5cacdedcd292c4c6b0","placeholder":"​","style":"IPY_MODEL_d6c6b760b1cc472dbd2df015c2c3efa9","value":" 12000/12000 [00:12&lt;00:00, 961.20it/s]"}},"82a0a5bcffbd431cb12359259e78b76e":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"835d070e8f344cfd92b79773ba7023b5":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HBoxModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_ffb331998a1d41cd8132263b449c68b9","IPY_MODEL_5c276d02f53b4cb88de8b336dc660cfd","IPY_MODEL_23ad55be58a045ea82d6aa18519d6739"],"layout":"IPY_MODEL_10bd83a61fce44e9b951d9f72cee9c1e"}},"85a3c4eb85ed4672a8ee46a037372300":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HBoxModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_4a2e89f1f6bf4eedbca0ad25d5a04de0","IPY_MODEL_63ab33992e6747789c6aa30d7fd3ca7b","IPY_MODEL_ddb4dc44dafc4f7ab81ee0be4bd36d29"],"layout":"IPY_MODEL_12762e6f19e44be9b4824d54b97acd15"}},"89b15c11862e4bfda72fcff7ea1e92d1":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"9315aadffe264b1a89bddc22ed186249":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"95a89ac1b6924365b4bccff08a482f49":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"FloatProgressModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_db781734b9984ade9318d80c0a70eb5d","max":5000,"min":0,"orientation":"horizontal","style":"IPY_MODEL_608345d3c72442538d04f01665b690e9","value":5000}},"973fca2dd29b4e0881bace37d1eca805":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_9ded00ca570047ed9d9b6f7f120982ff","placeholder":"​","style":"IPY_MODEL_0c091879d3ba48edb9b762aff8c5cd92","value":"NN training: 100%"}},"979bb47d6beb4f3ba39930094adabb8e":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"FloatProgressModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_cf8df982c4134fe39681a3d9906fadc8","max":5000,"min":0,"orientation":"horizontal","style":"IPY_MODEL_3acf1cac85e9490483a4ef865289b917","value":5000}},"98ccbf5dfc0e48d29b21891e60ab6c4c":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"9de6ad16dac748058be999dde9ffd367":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_ab611e839b6e45c985e821d7af6d17fb","placeholder":"​","style":"IPY_MODEL_89b15c11862e4bfda72fcff7ea1e92d1","value":"generating observations: 100%"}},"9ded00ca570047ed9d9b6f7f120982ff":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"9fb61e3b0e6949aea6f4a6b7154f4ce0":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_a4df1248688a4522b24d39639de0c3dd","placeholder":"​","style":"IPY_MODEL_d25593e56677436da0eb4d72a3c18f4e","value":"running EnKF: 100%"}},"a4df1248688a4522b24d39639de0c3dd":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"ab611e839b6e45c985e821d7af6d17fb":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"ada68873de254113a3534c1154e0347b":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"b0e7cc4350bf4ed3ac85b424eabab9f8":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"c0e73a04cfd24537921109d7d648dbf5":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"cb7abe9177294838930daa210455aef3":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"cf8df982c4134fe39681a3d9906fadc8":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"d25593e56677436da0eb4d72a3c18f4e":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"d6c6b760b1cc472dbd2df015c2c3efa9":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"d6dc2ba5f0a14b248c56e3fe88e94c3c":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"da7f410b46fa4574adf21b5b7c1be74d":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ProgressStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"db781734b9984ade9318d80c0a70eb5d":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"ddb4dc44dafc4f7ab81ee0be4bd36d29":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_b0e7cc4350bf4ed3ac85b424eabab9f8","placeholder":"​","style":"IPY_MODEL_3f80dd8cc17842b0b542dffa7d0b1f8b","value":" 500/500 [00:00&lt;00:00, 2675.46it/s]"}},"e48505496f2142ecac7cc2f43105dfe2":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"FloatProgressModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_2265c62e2c654e25a51d7bba6d6744f9","max":12000,"min":0,"orientation":"horizontal","style":"IPY_MODEL_6887d263554c4ce2a3e629693d29825b","value":12000}},"ea90e17d28754b61970391b99450da70":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HBoxModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_9de6ad16dac748058be999dde9ffd367","IPY_MODEL_979bb47d6beb4f3ba39930094adabb8e","IPY_MODEL_f0c6e5f86cd4441e8c299e4bd01515fc"],"layout":"IPY_MODEL_9315aadffe264b1a89bddc22ed186249"}},"ec55d4f2783d4264a14534273656a4a5":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ProgressStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"f0c6e5f86cd4441e8c299e4bd01515fc":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_ada68873de254113a3534c1154e0347b","placeholder":"​","style":"IPY_MODEL_d6dc2ba5f0a14b248c56e3fe88e94c3c","value":" 5000/5000 [00:00&lt;00:00, 17815.67it/s]"}},"f8cab4150ab040aeb033f64745faad74":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_675634702dbb4e31ba310b6cfb154ab6","placeholder":"​","style":"IPY_MODEL_c0e73a04cfd24537921109d7d648dbf5","value":" 5000/5000 [01:04&lt;00:00, 110.97it/s, rmse=0.200 spreadx=0.205 rmsei=2.714 infl=1.000 beta=1.157 nu=5100.000]"}},"f966d3ecfee44e5cacdedcd292c4c6b0":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"ffb331998a1d41cd8132263b449c68b9":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_320eba1b9795450db713503c124ef4c8","placeholder":"​","style":"IPY_MODEL_35b724c3fb1f4517ae9e33f151b578e9","value":"true model integration: 100%"}}}}},"nbformat":4,"nbformat_minor":5}