Upload folder using huggingface_hub

.gitattributes

@@ -53,3 +53,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.jpg filter=lfs diff=lfs merge=lfs -text
 *.jpeg filter=lfs diff=lfs merge=lfs -text
 *.webp filter=lfs diff=lfs merge=lfs -text
+data/tcwv05/tcwv_total.nc filter=lfs diff=lfs merge=lfs -text

data/tcwv05/downsampling.ipynb

@@ -0,0 +1,93 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import numpy as np\n",
+    "import os"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "s = np.load('/home/X-neural-lam/data/tcwv05/samples/train/tcwv.npy')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 13,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "split = 'test'\n",
+    "root_path = '/home/X-neural-lam/data/tcwv05'\n",
+    "data_path = os.path.join(root_path, 'samples',split)\n",
+    "sin = np.load(os.path.join(data_path, 'sin_feature.npy'))\n",
+    "cos = np.load(os.path.join(data_path, 'cos_feature.npy'))\n",
+    "land_sea_mask = np.load(os.path.join(data_path, 'land_sea_mask.npy'))\n",
+    "tcwv = np.load(os.path.join(data_path, 'tcwv.npy'))\n",
+    "\n",
+    "# sin = sin[:365*5,::4,::4]\n",
+    "# cos = cos[:365*5,::4,::4]\n",
+    "# tcwv = tcwv[:365*5,::4,::4]\n",
+    "\n",
+    "# land_sea_mask = land_sea_mask[::4,::4]\n",
+    "\n",
+    "sin = sin[:365]\n",
+    "cos = sin[:365]\n",
+    "tcwv = tcwv[:365,::4,::4]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 14,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "np.save(os.path.join(data_path, 'sin_feature.npy'), sin)\n",
+    "np.save(os.path.join(data_path, 'cos_feature.npy'), cos)\n",
+    "np.save(os.path.join(data_path, 'land_sea_mask.npy'), land_sea_mask)\n",
+    "np.save(os.path.join(data_path, 'tcwv.npy'), tcwv)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 15,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# statics\n",
+    "pos = np.load(os.path.join(root_path, 'static', 'pos_feature.npy'))\n",
+    "pos = pos[:,::4,::4]\n",
+    "np.save(os.path.join(root_path, 'static', 'pos_feature.npy'), pos)"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": ".venv",
+   "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": "3.9.18"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}

data/tcwv05/repair_data.ipynb

@@ -0,0 +1,46 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import numpy as np\n",
+    "import matplotlib.pyplot as plt\n",
+    "\n",
+    "data = np.load('/root/Desktop/machine_learning/neural-lam/data/tcwv05/samples/test/tcwv.npy')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "tcwv = data[:365,::4,::4]"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": ".venv",
+   "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": "3.9.18"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}

data/tcwv05/samples/test/cos_feature.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:148f38c1e61703ba383bcb1b500ebccb5c2f6b3acd7d54f3a7cf910bef3c6a8e
+size 5886848

data/tcwv05/samples/test/land_sea_mask.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c53aeaf3fb52dac7688d9806cc3ebf28eb26e585b1bb9320b5b89275d9930bd7
+size 8192

data/tcwv05/samples/test/sin_feature.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:148f38c1e61703ba383bcb1b500ebccb5c2f6b3acd7d54f3a7cf910bef3c6a8e
+size 5886848

data/tcwv05/samples/test/tcwv.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5dcadf28cc605d22f888d19e5a724ab137c7942505e97921ba19ca5ce2aac654
+size 2943488

data/tcwv05/samples/train/cos_feature.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:18837195adf9a94e09c24483d5bb3fa6530864441ee6454d62f72e4cf13cb0f8
+size 29433728

data/tcwv05/samples/train/land_sea_mask.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c53aeaf3fb52dac7688d9806cc3ebf28eb26e585b1bb9320b5b89275d9930bd7
+size 8192

data/tcwv05/samples/train/sin_feature.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:61aec1781001d16947093eb0885876790902b5853c10aee5d4d8a8f74f9ab719
+size 29433728

data/tcwv05/samples/train/tcwv.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4e9a9ae9d3dbc931d0ca821782982d12124c76c92e924fca5072b41f9bfaef8a
+size 14716928

data/tcwv05/samples/val/cos_feature.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d6db66738848c90625d769685371ddda012ade6883d3f5d49eef113b55c30327
+size 5886848

data/tcwv05/samples/val/land_sea_mask.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c53aeaf3fb52dac7688d9806cc3ebf28eb26e585b1bb9320b5b89275d9930bd7
+size 8192

data/tcwv05/samples/val/sin_feature.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d6db66738848c90625d769685371ddda012ade6883d3f5d49eef113b55c30327
+size 5886848

data/tcwv05/samples/val/tcwv.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6829c3cd5768c226488f4c38db6cfa0c54523fae5a90f91b56e458f5ccddd211
+size 2943488

data/tcwv05/static/diff_mean.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:90938f64832d7627ed14282f65cbea2e05a53eedc1c8d29ee8b39320ccef8563
+size 753

data/tcwv05/static/diff_std.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:374fae2ba472fa31ab5b827a56596fe854450a74c1e13855ec45ede41d25ae9e
+size 750

data/tcwv05/static/flux_stats.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cfc2bd06c03f47d4bd9758be1f15af2c20121a3086901132d037d2b60da8fdbb
+size 756

data/tcwv05/static/grid_features.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:25db888e9ca87094e44448360feb847a40c6dd11f840203444ae0acb92a5b462
+size 16893

data/tcwv05/static/parameter_mean.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:31ce5452c0560d1d7c9e80969463e19dcbe25eb526376f4c87c7d11d1c99ffa3
+size 768

data/tcwv05/static/parameter_std.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c1b2769dab47c20a37844344ae0e62546408267f0c773c2eff67defd5f1a6e50
+size 765

data/tcwv05/static/parameter_weights.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4396be9607d0a994dba58eb282e281a00da2c554c420229e52b8900d8ac701b1
+size 132

data/tcwv05/static/pos_feature.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:08d49a91f7038afb62c3886292a208ce1e0560fd9300ab8f3bb3f944b7fd2825
+size 16256

data/tcwv05/task.yaml

@@ -0,0 +1,16 @@
+task:
+  n_features: 1
+  n_init_state: 2
+  pred_length: 1
+  n_nodes: 2016
+  resolution: 2
+  dt: 24 # hours
+
+train:
+  n_t: 1825
+
+test:
+  n_t: 365
+
+val:
+  n_t: 365

data/tcwv05/tcwv_total.nc

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:926b713ac3715ce1126d76dcc31060386f8a3a07c867d2bef5e74a317b404676
+size 3744291671

data/tcwv05/view.data.ipynb

@@ -0,0 +1,164 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import numpy as np\n",
+    "import xarray as xr\n",
+    "import matplotlib.pyplot as plt\n",
+    "\n",
+    "dataset = xr.open_dataset('/home/X-neural-lam/data/tcwv05/tcwv_total.nc')\n",
+    "data = dataset['tcwv']\n",
+    "data = np.array(data)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "(1825, 21, 96)\n",
+      "(365, 21, 96)\n",
+      "(365, 21, 96)\n"
+     ]
+    }
+   ],
+   "source": [
+    "data_ds = data[-7*365:,::4,::4]\n",
+    "train = data_ds[:5*365,...]\n",
+    "val = data_ds[5*365:6*365,...]\n",
+    "test = data_ds[6*365:,...]\n",
+    "# print shape\n",
+    "print(train.shape)\n",
+    "print(val.shape)\n",
+    "print(test.shape)\n",
+    "# append a dimension\n",
+    "train = train[:,...,np.newaxis]\n",
+    "val = val[:,...,np.newaxis]\n",
+    "test = test[:,...,np.newaxis]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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",
+      "text/plain": [
+       "<Figure size 640x480 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "plt.imshow(train[0,...])\n",
+    "np.save('/home/X-neural-lam/data/tcwv05/samples/train/tcwv.npy', train)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 9,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAh8AAACbCAYAAADC4/k2AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8WgzjOAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAxbUlEQVR4nO2de5TdZX3un33fe257bpmZTJJJQoiG+y0QIhypJYrIqiKurroWtkHX0aqhgnRpxQoutTSsev6gdlFd7bFQjyItvWClPSgNGkrlGuQOAUxIJpe5ZTJ779kz+/47f+CZMHmerRNJ9gzk+ay110q+8/5+7/t7b/PObz/v84aCIAhgjDHGGNMgwvNdAGOMMcYcX3jxYYwxxpiG4sWHMcYYYxqKFx/GGGOMaShefBhjjDGmoXjxYYwxxpiG4sWHMcYYYxqKFx/GGGOMaShefBhjjDGmoXjxYYwxxpiGcswWH7feeitWrFiBZDKJdevW4dFHHz1WWRljjDHmTUToWJzt8g//8A/4gz/4A3zrW9/CunXrcMstt+Cuu+7C9u3b0dPT8yuvrdVq2LdvH1pbWxEKhY520YwxxhhzDAiCALlcDv39/QiHf827jeAYcN555wWbNm2a+X+1Wg36+/uDzZs3/9prBwcHAwD++OOPP/7448+b8DM4OPhrf9dHcZQplUrYtm0brr/++plYOBzGhg0b8NBDD1H6YrGIYrE48//gly9izvjglxCJJWfisXyNrg1VOf9yi15tTS2a2zdMkXJAsUSGYxAhGatDtMDPkxqapli1OUaxYjvHItN8PwBIHuB7hncNU6yyeglfXOMHikyXKRaaKlIMAKq/eJXTnnES552OU2y6i58xNcZ5x58flHkfvHgVxcrN/CatmuBYSLRjLDf3xg0iHFN9NVrge8bynDAi+kq4JG4IIJrnOkKFrw+SXMjwdIWvLZY4Fue2AQCIN5VBRNR5S4Ji4yenKJZZzfUTn9DjOBDhSGFusZAYOlUuDuJqHgCQ3sl1FB+bolh4LEOxWi7H5YnzeFB1CwBob+N7ppsoduCMVopNd/E9m0b4GStJnXeVmxE1UfSqiJW6dP89nCBWZ9zp6Y5o7svzPQN+nun9zRRr2SEGMoCWvaLsoopqMQ627eDyhMpifOe4/7x2UzGW86JTl3hODqqi3OrthMijHuH07P5XqZXw06Hb0NrK/e1wjvriY2xsDNVqFb29vbPivb29ePHFFyn95s2b8ZWvfIXikVgSkfihxUe0JBYfot6CmJ6cIok5Lj7Eb56IGADqF9QRLT6q/DzRqMgnyhN9NSYWH+IXDABEI3zPcFjMBtEkx9TiI8L1GNJjFKEQlzMUETNWlGNR8YzRKGcUVc8CzOo7/59aXMwQavEhqjISPwaLjyrfMxoTiw/RtuFancWHaB8EYsKK8NAPR8TCRSweEHlji4+Q6GuqvcJJ0f/qjGNV56Lry5hqb4huWq8PRKNcpmiE20eNu1qIYyERQ7jO4kOMp5qIqfqNiL6vnjFQ4waQdaSKrtKFU41ZfESaeEGtFh/hlKofPbGpMaoWH1Wx+FD9IiTGcjhcr37EWA6LyhC/oAI1CalfonJAaMJh0bjAnCQT877b5frrr0cmk5n5DA7qv2SNMcYY89bgqL/56O7uRiQSwfDw7Ff7w8PD6Ovro/SJRAKJBK+eWndNz3oTEIiVf/YEfjcaKeqVcmJCvAEQf3nGpub2FUukyKvDqUW6OlU+Uy28qm7eyX95FnrF+19BTfwVAwCR8UmK5S7gryQSE5x39AC/zguJ1/C1Nl3G6rvOpthUL/9p1LyXXxGqryQqKa6z2tkrZd4K9bYqmhd/XYtvGiriEdVf20CdtxzipUJNdJew+NpPfaUQrvOmKxB/hdeaxZ+j4q1WpZXThQv8F6HK47Uf8D1r4q/HzAmcz8SauY07VWcAEM9yLMLfOKIqXvCpr1gi4pvEsPhWCgCmesSbyCme04p9SylWSXFdqjksUtR/CVfjfH1+MZen1MbzQ+tu8fZVjLtIUc8tRfG3q/wqvJnvmdrH/UKNz3rtXW4V/beTG2hatENTM89rXSsOUuxgJ399BQDZNWqciLdfea63zIn8dYT6KjA63S7zVn1DvtAQ04OSLlSSXO5ih25vNQcWumeXp1YoAF+SlxNH/c1HPB7HOeecgy1bthwqUK2GLVu2YP369Uc7O2OMMca8yTjqbz4A4LrrrsPGjRuxdu1anHfeebjllluQz+fx0Y9+9FhkZ4wxxpg3Ecdk8fF7v/d7GB0dxY033oihoSGceeaZuPfee0mEaowxxpjjj2Oy+ACAq6++GldfffWxur0xxhhj3qQcs8XHGyVcrs7aTphZ3UJpprtZGJN+VQvxEhMsRormhYJMbBEqt3A1qS2EShgIAMVWIQIUuxXHz+zgdKKFCmJ/fjyrhba59/VTLCpEtVWxhTEi/ETU3vVCWj94fokQLomk092sZAqXuIwt+1lZle3XXTg1zv2g2MGZl7lbQezEg9r5FmUtLwCg1MmxGts5yLoMVfl5Sq2cLjalt7sqUWRUeMAU2+Ym94pNi75SZ+vl5BK+53SPEHqLMnY9yfdUnhHqWQAgnhMC8G4halRCvFGORUT/Uz4xAFBsF1tW+1nZquag3BJu74rQOYYC3d7KayNxkMueHBPbuoV4UYnj40KUDQCVlNiyKsZJyx6OlZQNxJHs+lTi8QOiLqtcxtwUp5sSW3Kr01pRHmll9Xg1zmlLQsyJsBCMlsTcO6nHZ7VZbLVVVhDiuYOYqMwoN1gsJdTxAGpqi/Lh23ynhHq2DvO+1dYYY4wxxxdefBhjjDGmoXjxYYwxxpiG4sWHMcYYYxqKFx/GGGOMaSgLdrfLwTWt8jCk19M6yErd1H6ttg0J6+dyG8vpy82sWla7WJTiX1qzA1BnBxc6xcmKizhd206OKQX4+HlaoRzOcObKLjlSFDtBhLo/UIffCatuAAgLi+pKGxc+OcJ1rpT0k/1zSwcAsSmOt7/MvumvfkgdR6kOU5v7wXLRcaGmX6oOd+N75s/jMqonDMmTDYFkiq8vFnmnRHmYdxglR0X9NokdEbqrISoO4mznsyTRtovLWG4T1vliR1m4Mvd2aBvkgqpjGsrN3PcLYmdUpUn3NWXRr06CrfQJG3Zlly0esWlUb/uITXK8KvIupEX9RjldYpx3feT79U4bNT+onTrpnaKM4jiIfL+ox446nvZiN0f8oNjdpHZ9xEV5MvyMYbGrBQCqOZFW2KuHRNHV74JA7ICpdOnnTrbz77d4jNOqnSnxKKdLimvrMV3i557MH/Y7Wm0VrIPffBhjjDGmoXjxYYwxxpiG4sWHMcYYYxqKFx/GGGOMaSgLVnAan6wh+jo7WCXwVMLLclp4MgOIFFmcWmplgZISnymLcyUUUyIqQNuhV5rF9X2s0KwNsn9ysYOvXdSXkXkXu7jwxRLHymVhIS+Eb7WsEM1N17ECFmLFqLANLgur5arQGofGuR5b9wg/ZwATJ3LbTvVx3+hbNkyxiFD87d3RTbFE97TMuwgufPdibp9Tu/dTrDXKgrJlyXGK7VTqZABLEwdFWi77T6urKVYUIujwXn6WUns9z2sh3FzE92zbxVdWhL2/GndBeO5/L8UzLKbLLufxlDrAfah5mGOFDm23rQTXkTqi3MOJKbt4Me7qHd0QLvP10bwSdfPcoq6ttPD4To3pMdY0IizJ41zQ3DKut8KFfN5Aa5KFyJmMULACUoXNVwMQ1uOxJpFSZFOr1an0BNdHIOzZI3GOlbPc/3qX8pidEuJOAChM8xymhMM9rXz2Q1U8jxKuh+uI2eMRfu6Tu2fPn+V8CWKPhMRvPowxxhjTULz4MMYYY0xD8eLDGGOMMQ3Fiw9jjDHGNJQFKzhte3QQ0fAhcc3wZSspTUg4UdbRyqDUKtZZIeF+J5IpN1KhAUSg9WgotQsHu3YWI8WTHMtcIMRsERaK9TSzwKgeFSE8GplsodhUgcVNqSUssswd7nL3S4KKEJdOcZerJoXYMMHPWBTixfySOuvnAbbb7O5gkduipjzFdh3soFjnsgmKLU+zUAwA8l0sKutvZsFpSqgSa6IDPjaxgmJ7J9My7/+791SKNb3C7dixj/vV5FLh2svVg5JwIwWAvkf5ecZOZ+HcyFqun6rQiScmOJY8oMWuSpA+uZRvmsjy9ROruE/WhN6vc7t2g4zmWYhXFQJaJUytCYFmSLm41jGOLHZw2cNlIWaXjsxcF+Eix5SgEQDyfZx3vp/Trr54B8VKVe5Dag46bWCfzHt0upliYzG+vpTjPhBWjqIV4QAr5i8AQFH0/7L4XdLFdXnBqS9T7NVsJ8XO7N0rs1ZzxkuZHooNjvIcVs5zp06mWYi8vEv8cgOQjAgBd3n23F8uz/19ht98GGOMMaahePFhjDHGmIbixYcxxhhjGooXH8YYY4xpKAtWcDr67uWIxA+JWVJCaFaLCZFPHWFWII7nHj6P09U62P0u/QQL5JQwVR2hDADlRSwSChWEu+q0OKo5xs+dEsem7xjrknn3pllkGQ+zQO69y16g2KtTfM/JCtdFvlm7yhYrc+teSnzWFOM662vOUmx4StijAuhIsOC0P8XXZ8XztPWwy2hM1Fk8rBs8HRfHXou05Ro/909eeRvFTlw8QrF9+1ikBuh+Nd3PZZ9aWc+l9LD77eQ+ufhh6SWJyTpHrx9O4iAL/oodPD5LbeLiOq6TTaP8jEpQqeYM5aZbbuMyjp6u+3NsUrgDi6RKxJoc5XySE0Js3aafOyLEpUiKumwVgtyEOIJeuUazjhMAUDiZxef/fMG3KPa/9l9CseczvRRTAvcXCyymBIBlXRMUO2UxOwZnulIUm5jmBs/k2OK0v4fzALT7aG6S82lqYjHn6DRXZq7Ac9BIXM9r0TA3UFeSVeGTrXzP8Sr3IbV54dUxPbdUhAt2JDp73NWmeO6rh998GGOMMaahePFhjDHGmIbixYcxxhhjGooXH8YYY4xpKAtWcNq8v4Jo7HUiPXG+e6TEYpncUhbaAEBuOQuuLv0f2yj241fWUCyWZyFUqa2OslUQ38cCpVKvECtWhBhOiHwSwh11MqddRveU2OmuIhz6WoVI8sz0HooNFvh+7W36aPnuGItdMxUWdk2UObZjksWuSrT57t4X55x3USj+WiNc9uEyu4fuEMfSKzdSALiofTvFnskvpVhblOv8ylMflfc8nCVN7JgKAAnhQlgVKuxdkywqG86xyK3Ww9fuHNBiuPQLnLbcwuM2cYCvbdnDY7ncxPeLsIYPADCxmvt0EBZC5v1cnvZXWKyaG+BrUyNapBsXR9gHYS57bokQA4v6nerndNW4tm4OibaNFDimtNGlVr5neRmLic9etUvmvbyJnTA/8fxHKHYww26kwT6er6rNon6FGykA7Am1U6y9hUXmp3exQ2qlhev3p5nVFFvRpp0+nxvto1hNuG1nD/Bzp+IspC8UeV7a/kq/zDs6IRyiE1xH4ZL4/RTjdAUhuA8Lt1YAEF0aOLxfCcF7PfzmwxhjjDENxYsPY4wxxjQULz6MMcYY01C8+DDGGGNMQ/HiwxhjjDENJRQEYhvJPJLNZpFOp/HOC29ANHpIET3Vy6rcqR5eO9WzAq6czbsfEnFheb2Nd3MkDvL9hDM2wixkBgAUeKMEUsJWuZQWtshpsctnWtjK1xEZxyY5VmEnYBT6uS7Wn/YyxboT4oZ1uKiNd33kaqxy/8ZL76LY59/+Y4o9kH07xc5s2S3zHimzN3dCNFAEXL+xEO9+2F3k3SE/+PH5Mu+r3nc/xYrCb3tJnDvW7iLv8kmKcmdUIwJYIjprU5h3MDSHedtIWXQitUvn5UlteT06xer+4VHeORQe5d1jauNQy24O1ursz1PW5aJpZTplKS66gCwjANTE6QI1sbNAjVG1wyOIioLX2fURyYmdMeKeLb08bgOxU6ZcntuxD69dL4Jit0NI7OKLZbgy1S6LUofeYRREROatPIdFkzx2KiVhh18WjVus0+Bi10d4mtMGog+oaxXJ/XpCj4qNhapfqT4tf0eoNqxTRjGNUNpqsYAXbv0iMpkM2trU+Qivu9+v/KkxxhhjzFHGiw9jjDHGNBQvPowxxhjTULz4MMYYY0xDWbD26qNnpBBJHBInFts5TeVEVt8s7xXezQCGs2wJPbmTxXBxsRzLLxHiMeVem9dKnZqwRs4L99zIHAQ9ABBn7SwSB7UgTZUzkRUiN2GN/dLjbDX/vLC8Vm0DAD89+0SKXbnqcYp9dNXDFKuKdfHF6ecpVgi0GK47ypW0v9xOsZYIW5zfP3oyxXYeYMFppUmL4ZRA9GCJLeTHytwnW0V5OqJ5ip2UZNtoAKiKDrMqNkqx54rcAfeU+Bn7Emzjvq71FzLvshDV7lrCausnsyxi7RFC5pp4lsmKUHcC2J3jsvc3c9nHi9wOxSqX+7193Nfu3HGOzHtiTKjcS9x/lfASylFciBdrTUIBCyC5kvv50vYJiilx6ct7WTicSHHfVaJNAChPsHg8Ocx1qcS7SsRfFU0bzeu/j6tCbx3bIRSVNd6ooDYlVITVfEhY0r/2A5GNEjILi/NwUYhveXgjNaLn8+YhLlQtxvfM94njBkS5lTC13sYJ9YyHi7BDpbnvX/GbD2OMMcY0FC8+jDHGGNNQvPgwxhhjTEM54sXHAw88gN/5nd9Bf38/QqEQ7r777lk/D4IAN954IxYvXoxUKoUNGzbg5ZfZrMoYY4wxxydHLDjN5/M444wz8LGPfQxXXHEF/fwv/uIv8I1vfAN///d/j5UrV+KGG27AJZdcgueffx7JJAuU6jF5WhHh1OsUMsLhryvNSp1sQedReond1mqdLN4JcsL1r4MVU219LPTK7daObtEpXuNFhThVOtUJ0dPUYq6LAuvtfpk351NpFmIksQyNTnM+8SzHqnEttC280E6x53oXU2xN8zDFvrNzHcWuOuEhiimnTgDYWVxEsZpQXD04uopivSlu266WKYqtOHdQ5j0s3FV35tm5NNbC/UoJbZXgtC86IfPOC4Fdq1CQnZPksivBqKrHV2p9Mu8Tk0MU29D6LMW6Y1y/j2VXUqw5ym0bCWmB8emdeym2Z6qd805yXU6UWL344+GTZD4S5XhZEGN+Uox58Tjxt2UpdukKFsACQFFYvuYqPAe+NMHt2NPN+YxNsBozGNJzalK4lCoHTiWqVeJFNbdIET7mLp6sqqKH+eKIcFeth3K3Vn1grs65ETGF1XPTLbVx5vEM/5JIjQm32FYhgq5ynYfrCG0r4iEPf8ZQnfZSHPHi49JLL8Wll14qfxYEAW655RZ86Utfwgc+8AEAwHe+8x309vbi7rvvxoc//OEjzc4YY4wxbzGOquZj586dGBoawoYNG2Zi6XQa69atw0MP8V+tAFAsFpHNZmd9jDHGGPPW5aguPoaGXnvt2tvbOyve29s787PD2bx5M9Lp9Mxn2bJlR7NIxhhjjFlgzPtul+uvvx6ZTGbmMziov0c3xhhjzFuDo+pw2tf3mhBteHgYixcfEhYODw/jzDPPlNckEgkkEiyS6+yaRKTpkFCuWGGhTXuK1U2xsHYCTJ3Fort9B9jhtNhLIYTE0cqpON+v5xTtOrlvggWIFfE8sRiXPRZm1VJJHAkdiWi3zakJFtNNqRO7E5x3rSLWpiKm6gfQTo0/28ECz9RqrsubT/pnin1913sp9s7uV2TevTH++k65f164iN06HxhhZ9ZTO/dTbLqqxY+rkiMU+9Ft76DYmf9zD5dRKM22T7HAs57QdrjMffrlEl8/KZR4yu1ViXTr8VKBxcQT1WaKvVpg11NFWqgXBxLjMm1ZqLUTQjk3LWw0C8LhdLrCbRuL6rmlpYPFyJMBO6mWO3mMrl7KfeWDi3/OeSulIoBnptgt9qL27RSLi7p4dpzbqyaOlo/n6oxvIS6MclVAaGJR5SkftU7ua8lR7ZgZE2L4cFmJJ4W4lLu5FPvX6/pKLFurI7qn8ogyKmFqpI5TaLjC8VqM20c9T6Q4N/fRUL1k4hEjhz1PIMpXj6P65mPlypXo6+vDli1bZmLZbBaPPPII1q9ffzSzMsYYY8yblCN+8zE5OYlXXjn01+bOnTvx5JNPorOzEwMDA7j22mvxZ3/2Z1i9evXMVtv+/n5cfvnlR7PcxhhjjHmTcsSLj8cffxzvete7Zv5/3XXXAQA2btyI22+/HZ///OeRz+fxiU98AhMTE7jwwgtx7733HpHHhzHGGGPeuhzx4uO3fuu3EAT1v9cJhUL46le/iq9+9atvqGDGGGOMeWtyVAWnR5OzFw0i3nJIHKbEfWHwImi0KM5LBtASZ4Heyt4DFOtazg6IzcJmb/80i0jjdcSuH1z1NMU6hWvleIXFeZNCmaWEeGVpuwe8OsXOmpMVvmdbjFVYFSF+PFhgIV2+rI85H8vy89RqrFrKlfmt2PMFFtJ9fOl/UWxbfoXMe0QcV58ps/h2SXKCYu/sYRHraInvtyLF/QcA2sLcPjd/5tsUe3DybRRT/WJ5Yoxie0ra0nZpnAWZKu2YOFf8zm++m2J9D/L9hr4ms8blK7ifP5I5gWIXtb9IsZYI99P+OJ+7PlHl/gcAGfE8SqSZiLEisk2Mp3Q7xzIVcY47tMvoLzrZUTQslIXvXfQcxQ6KeaAe57bspNglTbsppvpApZ3njH27uR2UIBIAhCEuiqx3Ri3B83SlWYhDxXHztYgWcoaEGlQJRGtiaqqK8ig9b2pU550a5QqJFjlWSc5NhFqNCVFsHXGomuarzTxPK2GqKmM5xdfGJnWDKyfWw8sZKevfgYp532prjDHGmOMLLz6MMcYY01C8+DDGGGNMQ/HiwxhjjDENZcEKTnviOSTih0SmSnSnxGc7oiz0AoCaEE8OFVhEqESsSuyqxKUnt7EL5mt5s6CoIGzysuIo7K4YCxAfOsDHj7fFhW0fgHPSLD57JreEYupY8ZE8i/iyeS6jcmsFtGNrSxOrlsaL3I5KfJupchmVmyignR9VO46W+BlPbOJ7djRxO6yIc58EgKYQP+NEjZ9xZWKUYsqpc1+pg2KxOudePz3FZyM1CSvKnjg7wL7/D7dSLPyHPB72KVUhgGey/RQb/8pyij140WkU2/iB+yn2f/acT7EzO9gVFgA6hJB0RIiEVbrFoi4WRTnWnhL2nQByNe6XK5IsRlZtGw9xOyph6ylNe2XeVTG3JEOcz+LYBMX+/eApFAunuDyVZv1rQglR4xNCCCpE5rWocOpMCtGm7mqIjfN8Xurm+SZU5rzDJSXwFEfQ18k7CAmRZp6fp9zM9+zczm7O5Ra+X+KgHt+BEuDWhKC3RTj+HuC8I6Jta1EtlFXurKCdr/PkcGqMMcYY8+vw4sMYY4wxDcWLD2OMMcY0FC8+jDHGGNNQvPgwxhhjTENZsLtdXpjsQyw45I17sMK7BU5rYuV7Tzwn77drmm2DlzZNUGy8xPlMVdijNy9ii+MZmfdjWVb8D0+xPXtXkndUTAhL8KrYufPUXt7BAgDNUd7psDfPMu7dzyym2Fnnss34WIx3h+zer62+C9PcvcopVmFPih0049PcDplJrot6xwyd0MO7DRY38Q6GSIhv8MO9p1Ns3zO9FLv+srtl3iPgti2K3U1NYeFXLNg53U0x1f8AYHUL79RRNuMjJS5jXlj5T4p8dmS4PABwSifv9lr19W0UC8Z4PKhjBD634kcUy9b0AZWPTbKNu5oL1G62JTG2cVdjbLyqj25IhnkXQTrCY7ktwjvSkiG+djTGu3TWxPVOuh/leOfQd7Ncl+MVLnuX2L0zVOadVYHYhQIA8YM8lkUzQnXz6DTvqCh3CmtusSsGAEo9XKbYAZ5vYjmx+0bZwnfy/ap6iMlnFCdwoHUPP4/a2dK0j/tFsVNkAqAqLNvVzhixiQpTi/meEWG5Hstpi/RanMsezR+WUUXv0lH4zYcxxhhjGooXH8YYY4xpKF58GGOMMaahePFhjDHGmIYSCoJ6kr35IZvNIp1O45Q7P4dI0yGBTHOCFT2xCAtjzu4clPfdW2inWFxYVO/Nc7o9BzgWEkLF9QOvyrx3ZlnsqlA27IOD4toqp4uNa+1w+mWOTfXy9W+/lBO+OsFCUiUOXbaIBXsAEBZ1tO8/2f67muJ0saywpF/E6YRzOACg3CFEcm0s7gtKvP7u7GVh6pmL9lGsOzEp81Y27vf87GyKrT2bBb3Pj/RR7OzF3KfbolqsOlZiW/rBXDvFklHu+9mCsM6vCuvnmBaVNce5MU5KD1MsJdR523Ms6H1+D4ugP3vWf8q8L0hxXf739IkU64pwm6ljGjqjnG4gOi7zTgp1X0KIfAeFV3izUGNGhCi2ntD2n8fXUuykZhanqnsOl1l0/MAw19ngbi0wDudYcBoR1uVhYV0ehMVYFlboQR3BabmV4xEhYq208jwQNGlBJd1vQs+p8YM8JoRDP4RmGbmVXB5V7kU/188dLQiB6KR4HnF5uZXbKznGY7HYoZW2ifFfL5CvVArY+shNyGQyaGvj/vV6/ObDGGOMMQ3Fiw9jjDHGNBQvPowxxhjTULz4MMYYY0xDWbCC0z984EOItxwS7u2bZrFWQQj7dh1khz4AWNzGiqAlTexIuiPHAs/9B1k4U97Dwr5aWgvxWrvY7TAQ4tJomMVEOeHqmd6qxWeKiTVCzJnjNWdZCLOieSEeW8NCvEVtWng5uHMRxdqfZhGXMK9F0zCXOz7JZQyX6wjSmvkZq3F+HuVWWG7ldNnV3LZvX7NX5t0SY2FWpcZirybhPruiiZ1Zd0+z8PeZERZjAsDyDhb/KmH1TiEmbhWi7l17eTw0p9mREQA2LN9OsYxw6O2M8XhQvDv9HMVGKuz+CWj30EURHvODZX6eQsDzSF+U54Y+cT8A2Fvluak9zO6hLxb755S3clwNQ7uMbi9yP5is8vwwJew6lSPzK9M9FHs5y+MYAPaMt1OsKsTwlSKP+UhMCEHVnFhH3KzSRiJ8z8I+nqdjvdMUK+WFyFIIYAEgIoS20EmJarNoRyHMR03fMLVH1KUQ3adG+J4qGzV/JrJakBsRYtfDqVQKePCnX7Hg1BhjjDELDy8+jDHGGNNQvPgwxhhjTEPx4sMYY4wxDUVbuC0AHti7apbD6dl9eyiNcgRd26cdTp8eY7FXZ4JFYZNFViB2tHK6yMksmhs6wMIzAMgNsUgu0spum9VJbo6ux4XASOj9pvq0QKllN8dL7ZwucYDXoSGhL4pv5WcZb9EiwM5xFjNFpzmW3sWisugUi54qTSz0CkL6uZUTIEJira2uz3EZu7Zx3nt38tHwAFAWJ69XmvieS8/Qx6QfzmSF+2ReOM0CwGiCBXYTk6zobUmxKDYzzfdMNLOarVwWgjsA97zIx7snknx9YUqI+0Qz3BM7lWK97TmZ97nduyj2gY4nKFYVf2+VxBnrj+RXUWxlYkTm3SXcUGPC4XS8ym3TGubBvL3AItKTklrc/PYE96H/nnwbxYaLPEbLAbfjiSl+xqqYZwGgWOF6my6zgDbezuN7ssB9Ws3nytkaAOLC3bo7xe3wdGkJxUoZoTIXQlkIF1YAqLaIuUWlVdVW4WBIuAgHcS3uLKeFs+sY3zPDXUA6yIZEeVIjur2bRtWcOvu/lfLclxR+82GMMcaYhuLFhzHGGGMaihcfxhhjjGkoXnwYY4wxpqEsWMFpbqgF4dQhAdzWXadQmkhBuOn1aIFSqpUFdjsy7HaYL7AYrqOFBaepGAtGY3Htxoc2IRISbnzhgyz4qyb4GRMZvrakta5SZCRO8UZMmJTGJufmMlo7UEf0WRTPXeTrK0nhuNrEsUgdN1NFuCSc+wrcPrWYEHupUSGyjpS08HJyqeiXLXyDXa+wm+T+LnYFrA6yYDR5ghZeHphgtWv8KRY6jvdwLNLP/fz0JfsotifXLvOeKrHYsLOJ3SR3DXPeUeEaWW7ivjJUx/nx7pEzKHZ/G6vuvnzyPRTri05QbKzMAs1cjd1aAaA94HprDnFfW50YpthTUwMUW9u8g2L5mhBJAijUuM5XJVk0Oi3coLvFoJ+q8fwXU8pzAC1xnkjaEiygHZxop5gSl+azPP8Fwm0YAPIZTtu3mh1o+zo5NhSIyVLZf9YR2irX1WKO2ycU4XsGEb5noPIu6fcCLa+K68U01CTcqZMH5+YQXYvq51a/i2L52fes5zit8JsPY4wxxjQULz6MMcYY01C8+DDGGGNMQ/HiwxhjjDENxYsPY4wxxjSUBbvbJVQLIfQ6ZXuQZDvdlpdYwZ16gmMAUGxnFXcxxjsLwsIae9+JrKxu62Z79XaxKwYAAqGaHh5mxXXHyeMUK+R5R05kmu+36OdakZ5dwVLoYrvYCTLB14bF5p0wNwMiJZ23UnZXE3OzcZeK62COFsYAQlWhNA8LtfYkP2S4LFThRU4Xm9Q7EKJF7oNTGX7u3Al8bbnMO1tSK3lnS7Gg+3klw/28chLvQIineLeWskZWuxJOSB+Qeb94gHfv7H2CrcJbhX1zfinXeZDizhZP6B1lkRRfn5nguvz6L95DsRtO/HeKnSCs1CeqfD8A2FfuoNiZ4vqT47zbJQwud3OId+ydEOe5AQD+94ELKfaxrv+mWFLc84n8CoqNFXkC7E6IrXAAWmK822Uoz3NqscT9qiQs9oMij5F8We8w6ujhMfH0Pj5CoyTGSSB2kixbxn06JyzgAWBSHG0QjnNfrU3N7ddrqMDpwqU6O05EdSTEMRZqThWO/6ikuC6iBb1jJTrFNz18B0y1znEXCr/5MMYYY0xD8eLDGGOMMQ3liBYfmzdvxrnnnovW1lb09PTg8ssvx/bt22elKRQK2LRpE7q6utDS0oIPfehDGB7m143GGGOMOT45osXH1q1bsWnTJjz88MO47777UC6X8Z73vAf5/CH9w2c/+1n88Ic/xF133YWtW7di3759uOKKK456wY0xxhjz5uSIBKf33nvvrP/ffvvt6OnpwbZt2/DOd74TmUwG3/72t3HHHXfgt3/7twEAt912G0466SQ8/PDDOP/88+ecVxAJELzOnjY+KsRwQnNX14JbCOeiU5w2KjSjrbtF3pF2io2zmzMAoLyMxV4t7Ww7XanxWnB6BV/bsZ3LU2jX68jUCD9j07ASnCqFkgjV+NpSi7YZr6nepe4phFBK2Kps2CPCRh0AKk1cJmVDrMSl8RHuBJU0K70ml7BoDgDK7B6OpKjf5BOcrhrnChqNixvWIZIXdTTGwrlShxasHs7Pp5dTLDyppw113EHfo/zchbQQHVc5ll3DscK0rvN4ggW0PYvYWntxM8eqolOuiI1R7L/KeoC/r+VZig1EWbj5cIE7dUQoA3+WX02xpog4EwHAp7sfoNioUCWuSeyn2JSwbP9ZaRXFYmqAAihV5yZQrlZ44IUmRP+LKjtynXdBWPlXRD5hYXEeamLR8liOx1hJCMcBIBjheqsJwXNIHG2h5r8wd10pGAX03FJpEkJ6cfpCdEgddyFidTYQFDq5fg8XpwZ1jj9QvCHNRyaTAQB0dnYCALZt24ZyuYwNGzbMpFmzZg0GBgbw0EMPvZGsjDHGGPMW4Tfealur1XDttdfiggsuwKmnngoAGBoaQjweR3t7+6y0vb29GBoakvcpFosoFg+t6rNZ/svEGGOMMW8dfuM3H5s2bcKzzz6LO++88w0VYPPmzUin0zOfZcuWvaH7GWOMMWZh8xu9+bj66qtxzz334IEHHsDSpUtn4n19fSiVSpiYmJj19mN4eBh9fX3yXtdffz2uu+66mf9nMhkMDAygVphtjFQr8PdN1aI4urysvyOsCmOZQPsVESF17HCV866xl9Nr8WnWbVSFQU9IHK1cm+bnVmZQ1TqmNOCsERJmXRWhfZir5qNS1pqPQH11KO/JsZpoRnXCdXAERzir8oQr3AkqVW6bSkV8l13n2Gtxejmqcyyn0h/UpnWfVoQKQk8htBi16TpfKh9GEIi8hSlSvXxUv1L1psZybZrbJhQVX5ADqFY5Xg24HcthHhBTOX7GZFhoVUR5AGBSdKxsnGP5AsemREcvTPOzhCNzzztfFSZ5YpBNF/ie5TzXT7Gs61ylrUwLHcgUT4xqrgzUEfSq/wGoxsU9hZmZmjTUPFsV80CtqPMOhAlXTZjFzVXzAdH3Q+L3C6DHiXgchIVESM7xalpS6QBUS2KeP2xeq5Zfa5dAGUJS3kdArVYLNm3aFPT39wcvvfQS/XxiYiKIxWLBP/3TP83EXnzxxQBA8NBDD80pj8HBweCXVeKPP/74448//rzJPoODg7/2d30oCOayRHmNT3/607jjjjvwgx/8AG9/+9tn4ul0GqnUawrrT33qU/iP//gP3H777Whra8Mf/dEfAQB+9rOfzSmPWq2Gffv2obW1FblcDsuWLcPg4CDa2ti218wf2WzWbbOAcfssXNw2Cxe3zRsjCALkcjn09/cjHP7Vqo4jWnyE6vi233bbbbjqqqsAvGYy9sd//Mf4/ve/j2KxiEsuuQR//dd/Xfdrl19FNptFOp1GJpNxR1hguG0WNm6fhYvbZuHitmkcR6T5mMs6JZlM4tZbb8Wtt976GxfKGGOMMW9dfLaLMcYYYxrKgl58JBIJfPnLX0YioY82NvOH22Zh4/ZZuLhtFi5um8ZxRJoPY4wxxpg3yoJ+82GMMcaYtx5efBhjjDGmoXjxYYwxxpiG4sWHMcYYYxrKgl183HrrrVixYgWSySTWrVuHRx99dL6LdFyyefNmnHvuuWhtbUVPTw8uv/xybN++fVaaQqGATZs2oaurCy0tLfjQhz6E4eHheSrx8cvNN9+MUCiEa6+9dibmtpk/9u7di4985CPo6upCKpXCaaedhscff3zm50EQ4MYbb8TixYuRSqWwYcMGvPzyy/NY4uODarWKG264AStXrkQqlcKqVavwta99bZaPldumAczpwJUGc+eddwbxeDz4u7/7u+C5554LPv7xjwft7e3B8PDwfBftuOOSSy4JbrvttuDZZ58NnnzyyeB973tfMDAwEExOTs6k+eQnPxksW7Ys2LJlS/D4448H559/fvCOd7xjHkt9/PHoo48GK1asCE4//fTgmmuumYm7beaH8fHxYPny5cFVV10VPPLII8GOHTuCH/3oR8Err7wyk+bmm28O0ul0cPfddwdPPfVU8P73vz9YuXJlMD09PY8lf+tz0003BV1dXcE999wT7Ny5M7jrrruClpaW4C//8i9n0rhtjj0LcvFx3nnnBZs2bZr5f7VaDfr7+4PNmzfPY6lMEATByMhIACDYunVrEASHDhO86667ZtK88MILATD3wwTNGyOXywWrV68O7rvvvuCiiy6aWXy4beaPP/mTPwkuvPDCuj+v1WpBX19f8PWvf30mNjExESQSieD73/9+I4p43HLZZZcFH/vYx2bFrrjiiuDKK68MgsBt0ygW3NcupVIJ27Ztw4YNG2Zi4XAYGzZswEMPPTSPJTMAkMlkAACdnZ0AgG3btqFcLs9qrzVr1mBgYMDt1SA2bdqEyy67bFYbAG6b+eTf/u3fsHbtWvzu7/4uenp6cNZZZ+Fv//ZvZ36+c+dODA0NzWqbdDqNdevWuW2OMe94xzuwZcsWvPTSSwCAp556Cg8++CAuvfRSAG6bRnFEZ7s0grGxMVSrVfT29s6K9/b24sUXX5ynUhngtROHr732WlxwwQU49dRTAQBDQ0OIx+Nob2+flba3txdDQ0PzUMrjizvvvBNPPPEEHnvsMfqZ22b+2LFjB775zW/iuuuuwxe/+EU89thj+MxnPoN4PI6NGzfO1L+a59w2x5YvfOELyGazWLNmDSKRCKrVKm666SZceeWVAOC2aRALbvFhFi6bNm3Cs88+iwcffHC+i2IADA4O4pprrsF9992HZDI538Uxr6NWq2Ht2rX48z//cwDAWWedhWeffRbf+ta3sHHjxnku3fHNP/7jP+J73/se7rjjDpxyyil48sknce2116K/v99t00AW3Ncu3d3diEQipMgfHh5GX1/fPJXKXH311bjnnnvwk5/8BEuXLp2J9/X1oVQqYWJiYlZ6t9exZ9u2bRgZGcHZZ5+NaDSKaDSKrVu34hvf+Aai0Sh6e3vdNvPE4sWLcfLJJ8+KnXTSSdi9ezcAzNS/57nG87nPfQ5f+MIX8OEPfxinnXYafv/3fx+f/exnsXnzZgBum0ax4BYf8Xgc55xzDrZs2TITq9Vq2LJlC9avXz+PJTs+CYIAV199Nf71X/8V999/P1auXDnr5+eccw5isdis9tq+fTt2797t9jrGXHzxxXjmmWfw5JNPznzWrl2LK6+8cubfbpv54YILLqAt6S+99BKWL18OAFi5ciX6+vpmtU02m8UjjzzitjnGTE1NIRye/asvEomgVqsBcNs0jPlWvCruvPPOIJFIBLfffnvw/PPPB5/4xCeC9vb2YGhoaL6LdtzxqU99Kkin08FPf/rTYP/+/TOfqampmTSf/OQng4GBgeD+++8PHn/88WD9+vXB+vXr57HUxy+v3+0SBG6b+eLRRx8NotFocNNNNwUvv/xy8L3vfS9oamoKvvvd786kufnmm4P29vbgBz/4QfD0008HH/jAB7ydswFs3LgxWLJkycxW23/5l38Juru7g89//vMzadw2x54FufgIgiD4q7/6q2BgYCCIx+PBeeedFzz88MPzXaTjEgDyc9ttt82kmZ6eDj796U8HHR0dQVNTU/DBD34w2L9///wV+jjm8MWH22b++OEPfxiceuqpQSKRCNasWRP8zd/8zayf12q14IYbbgh6e3uDRCIRXHzxxcH27dvnqbTHD9lsNrjmmmuCgYGBIJlMBieccELwp3/6p0GxWJxJ47Y59oSC4HW2bsYYY4wxx5gFp/kwxhhjzFsbLz6MMcYY01C8+DDGGGNMQ/HiwxhjjDENxYsPY4wxxjQULz6MMcYY01C8+DDGGGNMQ/HiwxhjjDENxYsPY4wxxjQULz6MMcYY01C8+DDGGGNMQ/HiwxhjjDEN5f8BEz9hA3GsAssAAAAASUVORK5CYII=",
+      "text/plain": [
+       "<Figure size 640x480 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "plt.imshow(test[0,...])\n",
+    "np.save('/home/X-neural-lam/data/tcwv05/samples/test/tcwv.npy', test)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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",
+      "text/plain": [
+       "<Figure size 640x480 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "plt.imshow(val[0,...])\n",
+    "np.save('/home/X-neural-lam/data/tcwv05/samples/val/tcwv.npy', val)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 13,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<matplotlib.image.AxesImage at 0x7f586c3b2880>"
+      ]
+     },
+     "execution_count": 13,
+     "metadata": {},
+     "output_type": "execute_result"
+    },
+    {
+     "data": {
+      "image/png": "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",
+      "text/plain": [
+       "<Figure size 640x480 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "s = np.load('/home/X-neural-lam/data/tcwv05/samples/train/tcwv.npy')\n",
+    "plt.imshow(s[0,...])"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": ".venv",
+   "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": "3.9.18"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}