Skip to content

Instantly share code, notes, and snippets.

@kayhan-batmanghelich

kayhan-batmanghelich/unet_ntk_share.ipynb Secret

Created January 22, 2020 19:02
Show Gist options
  • Save kayhan-batmanghelich/f444e6cec65139070f1b3e5ade230de5 to your computer and use it in GitHub Desktop.
Save kayhan-batmanghelich/f444e6cec65139070f1b3e5ade230de5 to your computer and use it in GitHub Desktop.
Download ZIP
UNet_NTK_share.ipynb
Display the source blob
Display the rendered blob
Raw
unet_ntk_share.ipynb
{
"nbformat": 4,
"nbformat_minor": 0,
"metadata": {
"colab": {
"name": "UNet_NTK_share.ipynb",
"provenance": [],
"collapsed_sections": [],
"authorship_tag": "ABX9TyOZlnLkHTK/Ewnk+H5ryY/x",
"include_colab_link": true
},
"kernelspec": {
"name": "python3",
"display_name": "Python 3"
},
"accelerator": "TPU"
},
"cells": [
{
"cell_type": "markdown",
"metadata": {
"id": "view-in-github",
"colab_type": "text"
},
"source": [
"<a href=\"https://colab.research.google.com/gist/kayhan-batmanghelich/f444e6cec65139070f1b3e5ade230de5/unet_ntk_share.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
]
},
{
"cell_type": "code",
"metadata": {
"id": "LSkUhNhI1Qov",
"colab_type": "code",
"outputId": "b637e28b-3c1c-41bf-a9e3-b4eb0ac940b5",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 34
}
},
"source": [
"# Grab newest JAX version.\n",
"#!pip install --upgrade -q jax==0.1.54 jaxlib==0.1.37\n",
"\n",
"# Make sure the Colab Runtime is set to Accelerator: TPU.\n",
"import requests\n",
"import os\n",
"if 'TPU_DRIVER_MODE' not in globals():\n",
" url = 'http://' + os.environ['COLAB_TPU_ADDR'].split(':')[0] + ':8475/requestversion/tpu_driver_nightly'\n",
" resp = requests.post(url)\n",
" TPU_DRIVER_MODE = 1\n",
"\n",
"# The following is required to use TPU Driver as JAX's backend.\n",
"from jax.config import config\n",
"config.FLAGS.jax_xla_backend = \"tpu_driver\"\n",
"config.FLAGS.jax_backend_target = \"grpc://\" + os.environ['COLAB_TPU_ADDR']\n",
"print(config.FLAGS.jax_backend_target)"
],
"execution_count": 1,
"outputs": [
{
"output_type": "stream",
"text": [
"grpc://10.79.68.74:8470\n"
],
"name": "stdout"
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "0f9QvjF51shW",
"colab_type": "code",
"outputId": "54848092-cf01-4f7b-e030-d7145799a97b",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 102
}
},
"source": [
"!pip3 install torch torchvision"
],
"execution_count": 2,
"outputs": [
{
"output_type": "stream",
"text": [
"Requirement already satisfied: torch in /usr/local/lib/python3.6/dist-packages (1.3.1)\n",
"Requirement already satisfied: torchvision in /usr/local/lib/python3.6/dist-packages (0.4.2)\n",
"Requirement already satisfied: numpy in /usr/local/lib/python3.6/dist-packages (from torch) (1.17.5)\n",
"Requirement already satisfied: pillow>=4.1.1 in /usr/local/lib/python3.6/dist-packages (from torchvision) (6.2.2)\n",
"Requirement already satisfied: six in /usr/local/lib/python3.6/dist-packages (from torchvision) (1.12.0)\n"
],
"name": "stdout"
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "EQxvptgM17wL",
"colab_type": "code",
"outputId": "04a04727-676d-4280-c017-f904e34e5618",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 119
}
},
"source": [
"!git clone https://github.com/usuyama/pytorch-unet.git"
],
"execution_count": 3,
"outputs": [
{
"output_type": "stream",
"text": [
"Cloning into 'pytorch-unet'...\n",
"remote: Enumerating objects: 7, done.\u001b[K\n",
"remote: Counting objects: 100% (7/7), done.\u001b[K\n",
"remote: Compressing objects: 100% (7/7), done.\u001b[K\n",
"remote: Total 55 (delta 2), reused 1 (delta 0), pack-reused 48\u001b[K\n",
"Unpacking objects: 100% (55/55), done.\n"
],
"name": "stdout"
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "yCyW-zFE_kB4",
"colab_type": "code",
"outputId": "80e5e319-9bdc-470f-def5-0ebaf18204c4",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 136
}
},
"source": [
"!git clone --single-branch --branch unet https://github.com/kayhan-batmanghelich/neural-tangents.git"
],
"execution_count": 4,
"outputs": [
{
"output_type": "stream",
"text": [
"Cloning into 'neural-tangents'...\n",
"remote: Enumerating objects: 88, done.\u001b[K\n",
"remote: Counting objects: 1% (1/88)\u001b[K\rremote: Counting objects: 2% (2/88)\u001b[K\rremote: Counting objects: 3% (3/88)\u001b[K\rremote: Counting objects: 4% (4/88)\u001b[K\rremote: Counting objects: 5% (5/88)\u001b[K\rremote: Counting objects: 6% (6/88)\u001b[K\rremote: Counting objects: 7% (7/88)\u001b[K\rremote: Counting objects: 9% (8/88)\u001b[K\rremote: Counting objects: 10% (9/88)\u001b[K\rremote: Counting objects: 11% (10/88)\u001b[K\rremote: Counting objects: 12% (11/88)\u001b[K\rremote: Counting objects: 13% (12/88)\u001b[K\rremote: Counting objects: 14% (13/88)\u001b[K\rremote: Counting objects: 15% (14/88)\u001b[K\rremote: Counting objects: 17% (15/88)\u001b[K\rremote: Counting objects: 18% (16/88)\u001b[K\rremote: Counting objects: 19% (17/88)\u001b[K\rremote: Counting objects: 20% (18/88)\u001b[K\rremote: Counting objects: 21% (19/88)\u001b[K\rremote: Counting objects: 22% (20/88)\u001b[K\rremote: Counting objects: 23% (21/88)\u001b[K\rremote: Counting objects: 25% (22/88)\u001b[K\rremote: Counting objects: 26% (23/88)\u001b[K\rremote: Counting objects: 27% (24/88)\u001b[K\rremote: Counting objects: 28% (25/88)\u001b[K\rremote: Counting objects: 29% (26/88)\u001b[K\rremote: Counting objects: 30% (27/88)\u001b[K\rremote: Counting objects: 31% (28/88)\u001b[K\rremote: Counting objects: 32% (29/88)\u001b[K\rremote: Counting objects: 34% (30/88)\u001b[K\rremote: Counting objects: 35% (31/88)\u001b[K\rremote: Counting objects: 36% (32/88)\u001b[K\rremote: Counting objects: 37% (33/88)\u001b[K\rremote: Counting objects: 38% (34/88)\u001b[K\rremote: Counting objects: 39% (35/88)\u001b[K\rremote: Counting objects: 40% (36/88)\u001b[K\rremote: Counting objects: 42% (37/88)\u001b[K\rremote: Counting objects: 43% (38/88)\u001b[K\rremote: Counting objects: 44% (39/88)\u001b[K\rremote: Counting objects: 45% (40/88)\u001b[K\rremote: Counting objects: 46% (41/88)\u001b[K\rremote: Counting objects: 47% (42/88)\u001b[K\rremote: Counting objects: 48% (43/88)\u001b[K\rremote: Counting objects: 50% (44/88)\u001b[K\rremote: Counting objects: 51% (45/88)\u001b[K\rremote: Counting objects: 52% (46/88)\u001b[K\rremote: Counting objects: 53% (47/88)\u001b[K\rremote: Counting objects: 54% (48/88)\u001b[K\rremote: Counting objects: 55% (49/88)\u001b[K\rremote: Counting objects: 56% (50/88)\u001b[K\rremote: Counting objects: 57% (51/88)\u001b[K\rremote: Counting objects: 59% (52/88)\u001b[K\rremote: Counting objects: 60% (53/88)\u001b[K\rremote: Counting objects: 61% (54/88)\u001b[K\rremote: Counting objects: 62% (55/88)\u001b[K\rremote: Counting objects: 63% (56/88)\u001b[K\rremote: Counting objects: 64% (57/88)\u001b[K\rremote: Counting objects: 65% (58/88)\u001b[K\rremote: Counting objects: 67% (59/88)\u001b[K\rremote: Counting objects: 68% (60/88)\u001b[K\rremote: Counting objects: 69% (61/88)\u001b[K\rremote: Counting objects: 70% (62/88)\u001b[K\rremote: Counting objects: 71% (63/88)\u001b[K\rremote: Counting objects: 72% (64/88)\u001b[K\rremote: Counting objects: 73% (65/88)\u001b[K\rremote: Counting objects: 75% (66/88)\u001b[K\rremote: Counting objects: 76% (67/88)\u001b[K\rremote: Counting objects: 77% (68/88)\u001b[K\rremote: Counting objects: 78% (69/88)\u001b[K\rremote: Counting objects: 79% (70/88)\u001b[K\rremote: Counting objects: 80% (71/88)\u001b[K\rremote: Counting objects: 81% (72/88)\u001b[K\rremote: Counting objects: 82% (73/88)\u001b[K\rremote: Counting objects: 84% (74/88)\u001b[K\rremote: Counting objects: 85% (75/88)\u001b[K\rremote: Counting objects: 86% (76/88)\u001b[K\rremote: Counting objects: 87% (77/88)\u001b[K\rremote: Counting objects: 88% (78/88)\u001b[K\rremote: Counting objects: 89% (79/88)\u001b[K\rremote: Counting objects: 90% (80/88)\u001b[K\rremote: Counting objects: 92% (81/88)\u001b[K\rremote: Counting objects: 93% (82/88)\u001b[K\rremote: Counting objects: 94% (83/88)\u001b[K\rremote: Counting objects: 95% (84/88)\u001b[K\rremote: Counting objects: 96% (85/88)\u001b[K\rremote: Counting objects: 97% (86/88)\u001b[K\rremote: Counting objects: 98% (87/88)\u001b[K\rremote: Counting objects: 100% (88/88)\u001b[K\rremote: Counting objects: 100% (88/88), done.\u001b[K\n",
"remote: Compressing objects: 100% (62/62), done.\u001b[K\n",
"remote: Total 843 (delta 42), reused 62 (delta 26), pack-reused 755\u001b[K\n",
"Receiving objects: 100% (843/843), 7.99 MiB | 23.92 MiB/s, done.\n",
"Resolving deltas: 100% (561/561), done.\n"
],
"name": "stdout"
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "tM1DVgy7DSqi",
"colab_type": "code",
"outputId": "33a2061f-7d06-4ca5-f4b6-9f64f8648026",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 701
}
},
"source": [
"#os.chdir('neural-tangents')\n",
"!pip install -e neural-tangents\n",
"#os.chdir('..')"
],
"execution_count": 5,
"outputs": [
{
"output_type": "stream",
"text": [
"Obtaining file:///content/neural-tangents\n",
"Collecting jaxlib>=0.1.37\n",
"\u001b[?25l Downloading https://files.pythonhosted.org/packages/24/bf/e181454464b866f30f09b5d74d1dd08e8b15e032716d8bcc531c659776ab/jaxlib-0.1.37-cp36-none-manylinux2010_x86_64.whl (25.4MB)\n",
"\u001b[K |████████████████████████████████| 25.4MB 1.9MB/s \n",
"\u001b[?25hCollecting jax>=0.1.57\n",
"\u001b[?25l Downloading https://files.pythonhosted.org/packages/ae/f2/ea981ed2659f70a1d8286ce41b5e74f1d9df844c1c6be6696144ed3f2932/jax-0.1.57.tar.gz (255kB)\n",
"\u001b[K |████████████████████████████████| 256kB 43.9MB/s \n",
"\u001b[?25hCollecting frozendict\n",
" Downloading https://files.pythonhosted.org/packages/4e/55/a12ded2c426a4d2bee73f88304c9c08ebbdbadb82569ebdd6a0c007cfd08/frozendict-1.2.tar.gz\n",
"Requirement already satisfied: numpy>=1.12 in /usr/local/lib/python3.6/dist-packages (from jaxlib>=0.1.37->neural-tangents==0.1.5) (1.17.5)\n",
"Requirement already satisfied: absl-py in /usr/local/lib/python3.6/dist-packages (from jaxlib>=0.1.37->neural-tangents==0.1.5) (0.9.0)\n",
"Requirement already satisfied: six in /usr/local/lib/python3.6/dist-packages (from jaxlib>=0.1.37->neural-tangents==0.1.5) (1.12.0)\n",
"Requirement already satisfied: scipy in /usr/local/lib/python3.6/dist-packages (from jaxlib>=0.1.37->neural-tangents==0.1.5) (1.4.1)\n",
"Requirement already satisfied: protobuf>=3.6.0 in /usr/local/lib/python3.6/dist-packages (from jaxlib>=0.1.37->neural-tangents==0.1.5) (3.10.0)\n",
"Requirement already satisfied: opt_einsum in /usr/local/lib/python3.6/dist-packages (from jax>=0.1.57->neural-tangents==0.1.5) (3.1.0)\n",
"Requirement already satisfied: fastcache in /usr/local/lib/python3.6/dist-packages (from jax>=0.1.57->neural-tangents==0.1.5) (1.1.0)\n",
"Requirement already satisfied: setuptools in /usr/local/lib/python3.6/dist-packages (from protobuf>=3.6.0->jaxlib>=0.1.37->neural-tangents==0.1.5) (42.0.2)\n",
"Building wheels for collected packages: jax, frozendict\n",
" Building wheel for jax (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
" Created wheel for jax: filename=jax-0.1.57-cp36-none-any.whl size=297710 sha256=ed6f5e14c84eb44514571e8d4c4b1c63d7675c9fe714492506a4e9c7d35b4fea\n",
" Stored in directory: /root/.cache/pip/wheels/8a/b4/75/859bcdaf181569124306615bd9b68c747725c60bfa68826378\n",
" Building wheel for frozendict (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
" Created wheel for frozendict: filename=frozendict-1.2-cp36-none-any.whl size=3149 sha256=efbadb5f600bd00e8ceeb354c9e3ba5ef644cb71a5d8fa64156396c489f47404\n",
" Stored in directory: /root/.cache/pip/wheels/6c/6c/e9/534386165bd12cf1885582c75eb6d0ffcb321b65c23fe0f834\n",
"Successfully built jax frozendict\n",
"Installing collected packages: jaxlib, jax, frozendict, neural-tangents\n",
" Found existing installation: jaxlib 0.1.36\n",
" Uninstalling jaxlib-0.1.36:\n",
" Successfully uninstalled jaxlib-0.1.36\n",
" Found existing installation: jax 0.1.52\n",
" Uninstalling jax-0.1.52:\n",
" Successfully uninstalled jax-0.1.52\n",
" Running setup.py develop for neural-tangents\n",
"Successfully installed frozendict-1.2 jax-0.1.57 jaxlib-0.1.37 neural-tangents\n"
],
"name": "stdout"
},
{
"output_type": "display_data",
"data": {
"application/vnd.colab-display-data+json": {
"pip_warning": {
"packages": [
"jax",
"jaxlib"
]
}
}
},
"metadata": {
"tags": []
}
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "F-ejLz7FOYzh",
"colab_type": "code",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 241
},
"outputId": "0733a812-2db8-4811-f0e1-66d23fc4529b"
},
"source": [
"!pip install neural-tangents"
],
"execution_count": 17,
"outputs": [
{
"output_type": "stream",
"text": [
"Requirement already satisfied: neural-tangents in ./neural-tangents (0.1.5)\n",
"Requirement already satisfied: jaxlib>=0.1.37 in /usr/local/lib/python3.6/dist-packages (from neural-tangents) (0.1.37)\n",
"Requirement already satisfied: jax>=0.1.57 in /usr/local/lib/python3.6/dist-packages (from neural-tangents) (0.1.57)\n",
"Requirement already satisfied: frozendict in /usr/local/lib/python3.6/dist-packages (from neural-tangents) (1.2)\n",
"Requirement already satisfied: absl-py in /usr/local/lib/python3.6/dist-packages (from jaxlib>=0.1.37->neural-tangents) (0.9.0)\n",
"Requirement already satisfied: numpy>=1.12 in /usr/local/lib/python3.6/dist-packages (from jaxlib>=0.1.37->neural-tangents) (1.17.5)\n",
"Requirement already satisfied: six in /usr/local/lib/python3.6/dist-packages (from jaxlib>=0.1.37->neural-tangents) (1.12.0)\n",
"Requirement already satisfied: protobuf>=3.6.0 in /usr/local/lib/python3.6/dist-packages (from jaxlib>=0.1.37->neural-tangents) (3.10.0)\n",
"Requirement already satisfied: scipy in /usr/local/lib/python3.6/dist-packages (from jaxlib>=0.1.37->neural-tangents) (1.4.1)\n",
"Requirement already satisfied: opt-einsum in /usr/local/lib/python3.6/dist-packages (from jax>=0.1.57->neural-tangents) (3.1.0)\n",
"Requirement already satisfied: fastcache in /usr/local/lib/python3.6/dist-packages (from jax>=0.1.57->neural-tangents) (1.1.0)\n",
"Requirement already satisfied: setuptools in /usr/local/lib/python3.6/dist-packages (from protobuf>=3.6.0->jaxlib>=0.1.37->neural-tangents) (42.0.2)\n"
],
"name": "stdout"
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "39v1WzEgPdUd",
"colab_type": "code",
"colab": {}
},
"source": [
"from neural_tangents import stax"
],
"execution_count": 0,
"outputs": []
},
{
"cell_type": "markdown",
"metadata": {
"id": "u_W7G-mvW_Om",
"colab_type": "text"
},
"source": [
"# Imports"
]
},
{
"cell_type": "code",
"metadata": {
"id": "PnmQM9FD2Ip-",
"colab_type": "code",
"colab": {}
},
"source": [
"import itertools\n",
"import time\n",
"\n",
"import matplotlib.pylab as plt\n",
"import sys, os\n",
"import numpy as onp\n",
"import torch\n",
"\n",
"sys.path.append('pytorch-unet')\n",
"sys.path.append('BayesianUNet/src')\n",
"\n",
"import simulation\n",
"import helper\n",
"\n",
"from jax.experimental import optimizers"
],
"execution_count": 0,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"id": "Zw7DuMzJYOnI",
"colab_type": "code",
"colab": {}
},
"source": [
"import numpy as onp\n",
"from torch.utils.data import Dataset, DataLoader\n",
"from torchvision import transforms, datasets, models"
],
"execution_count": 0,
"outputs": []
},
{
"cell_type": "markdown",
"metadata": {
"id": "08nEnV8MXG-a",
"colab_type": "text"
},
"source": [
"# Simulate Images"
]
},
{
"cell_type": "code",
"metadata": {
"id": "mYl5G7XAXLXc",
"colab_type": "code",
"outputId": "daa27650-8cc7-4c91-dcb5-73895bf3406c",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 772
}
},
"source": [
"# Generate some random images\n",
"input_images, target_masks = simulation.generate_random_data(192, 192, count=3)\n",
"\n",
"for x in [input_images, target_masks]:\n",
" print(x.shape)\n",
" print(x.min(), x.max())\n",
"\n",
"# Change channel-order and make 3 channels for matplot\n",
"input_images_rgb = [x.astype(onp.uint8) for x in input_images]\n",
"\n",
"# Map each channel (i.e. class) to each color\n",
"target_masks_rgb = [helper.masks_to_colorimg(x) for x in target_masks]\n",
"\n",
"# Left: Input image, Right: Target mask (Ground-truth)\n",
"helper.plot_side_by_side([input_images_rgb, target_masks_rgb])"
],
"execution_count": 5,
"outputs": [
{
"output_type": "stream",
"text": [
"(3, 192, 192, 3)\n",
"0 255\n",
"(3, 6, 192, 192)\n",
"0.0 1.0\n"
],
"name": "stdout"
},
{
"output_type": "display_data",
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAdsAAAKvCAYAAAAiIWV+AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjIsIGh0\ndHA6Ly9tYXRwbG90bGliLm9yZy8li6FKAAAgAElEQVR4nO3df4wc9X3/8dcrdhOplK+AsLUsg2tA\nDhFE7QVWblACXwghMYhvHBqJ2qoSJ0E90Bek9ttK/ZLwVUH9CilKQ5GifAM5hGVTJYY0LoEvX7eJ\nvygKSQWFc+I6hmCwiRG+r7EvuAIUIlLj9/ePmyPDsXe3uzOfmx/3fEir3f3szM57zjf38nt2ZtYR\nIQAAkM47qi4AAIC2I2wBAEiMsAUAIDHCFgCAxAhbAAASI2wBAEgsWdjaXmt7r+19tm9MtRwAAOrO\nKc6ztb1E0jOSLpN0UNITkjZExFOlLwwAgJpL1dmukbQvIp6LiF9LulfSukTLAgCg1pYmet8Vkl7I\nPT8o6Q9nm9g2l7HCYvaLiOhUXURZTj311Fi1alXVZQCV2LlzZ8/tOVXYzsv2qKTRqpYP1MjzVRdQ\nVH57XrlypcbHxyuuCKiG7Z7bc6rdyBOSTs89Py0be1NEjEVENyK6iWoAsEDy23On05omHShNqrB9\nQtJq22fYfqek9ZIeTLQsAABqLclu5Ig4ZvsGSd+VtETSpoh4MsWyAACou2Sf2UbEdknbU70/AABN\nwRWkAABIjLAFACAxwhYAgMQIWwAAEiNsAQBIjLAFACAxwhYAgMQIWwAAEqvsiwiA1Ib5rmbbCSoB\nUNTjH7xk4HnW/Mv3E1QyHDpbAAASI2wBAEiMsAUAIDHCFgCAxAhbAAASI2wBAEiMsM0Z5lQRAPXk\nr19ZdQnAm4YOW9un2/6+7adsP2n7z7LxW2xP2N6V3a4or9z0CFygPQhc1EWRzvaYpL+MiHMkfUDS\n9bbPyV67PSJGstv2wlUuMAIXaA8CF3UwdNhGxKGI+HH2+FVJP5O0oqzCqkbgAu1B4KJqpXxma3uV\npPdL+tds6Abbu21vsn1yGcuoAoELtAeBiyoVDlvbvyNpm6Q/j4hXJN0h6SxJI5IOSbptlvlGbY/b\nHi9aQ0pVB25EzHoD6iK/PU9OTlZdzqyqDtxXd71j1hvazUX+aNv+LUkPSfpuRPxdj9dXSXooIt43\nz/vUIjnm+lks9AXqB/l34eL5jbczIrpVF1GWbrcb4+PV/x96rmCNax9awEo0UJieOHI8YSVIzXbP\n7bnI0ciWdLekn+WD1vby3GRXSdoz7DLqZCE7yUGXRZcLDGYhO9xBu1a63HYq8q/6QUmfkvThGaf5\nfMn2T23vlnSJpP9WRqF1QKgB7VH1LmUsLkN/n21E/EhSr/2XjTvVZxARkXS37bCBnrouoI389SuT\n7lIetkt9ddc72J3cMuyvGAIdLtAedLhYCITtkAhcoD0IXKRG2BZA4ALtQeAiJcK2IAIXaA8CF6kQ\ntiUgcIH2IHCRAmFbEgIXaA8CF2UjbEtURuAOe/oOp/0A5SojcIc9fYfTftqHsC0ZHS7QHnS4KAth\nm0DRwB20S6WrBdIpGriDdql0te1E2CZSRuDOF6L9TAOguDICd74Q7WcaNNfQl2vE/Mq4hCJhCtRD\nGZd2JEwXL8I2h2AD2mOhv0YPmAu7kQEASIywBQAgMcIWAIDECFsAABIrfICU7QOSXpX0hqRjEdG1\nfYqk+yStknRA0tUR8e9FlwUAQBOV1dleEhEjEdHNnt8o6eGIWC3p4ew5AACLUqrdyOskbckeb5H0\niUTLAQCg9soI25D0Pds7bY9mY8si4lD2+EVJy0pYDgAAjVTGRS0+FBETtn9X0g7bT+dfjIiw/bZr\nF2bBPDpzHEDz5LfnlStXVlwNUD+FO9uImMjuj0i6X9IaSYdtL5ek7P5Ij/nGIqKb+5wXQEPlt+dO\np1N1OUDtFApb2yfYPnH6saSPStoj6UFJG7PJNkp6oMhyAABosqK7kZdJuj+7pvBSSd+MiH+2/YSk\nb9m+RtLzkq4uuBwAABqrUNhGxHOS/qDH+EuSLi3y3gAAtAVXkAIAIDHCFrUUEYp420HsABpoyw8v\n0JYfXlB1GZUibAEASIywBQAgMcIWAIDECFsAABIjbAEASIywBQAgsTK+iAAYSj+n9sw1TXblMgA1\n0M+pPXNNs/HCR8ssp3bobAEASIzOFpWZqzOd7mjpXoFmmKszne5o2969zoXOFgCAxAhbAAASI2wB\nAEiMsAUAIDHCFgCAxIY+Gtn22ZLuyw2dKemvJZ0k6U8lTWbjX4iI7UNXCABAww0dthGxV9KIJNle\nImlC0v2SPivp9oj4cikVYlHilB+gPRbzKT/TytqNfKmk/RHxfEnvBwBAa5QVtuslbc09v8H2btub\nbJ9c0jIAAGikwmFr+52SPi7pH7KhOySdpaldzIck3TbLfKO2x22PF60BQLXy2/Pk5OT8MwCLTBmd\n7eWSfhwRhyUpIg5HxBsRcVzSXZLW9JopIsYiohsR3RJqAFCh/Pbc6XSqLgeonTLCdoNyu5BtL8+9\ndpWkPSUsAwCAxir0RQS2T5B0maRrc8Nfsj0iKSQdmPEaAACLTqGwjYhfSnr3jLFPFaoIAICW4QpS\nAAAkRtgCAJAYYQsAQGKELQAAiRG2AAAkRtgCAJAYYQsAQGKELQAAiRG2AAAkRtgCAJAYYQsAQGKE\nLQAAiRG2AAAkRtgCAJAYYQsAQGKELQAAifUVtrY32T5ie09u7BTbO2w/m92fnI3b9lds77O92/Z5\nqYoHAKAJ+u1sN0taO2PsRkkPR8RqSQ9nzyXpckmrs9uopDuKlwkAQHP1FbYR8YikozOG10nakj3e\nIukTufF7Yspjkk6yvbyMYgEAaKIin9kui4hD2eMXJS3LHq+Q9EJuuoPZGAAAi1IpB0hFREiKQeax\nPWp73PZ4GTUAqE5+e56cnKy6HKB2ioTt4endw9n9kWx8QtLpuelOy8beIiLGIqIbEd0CNQCogfz2\n3Ol0qi4HqJ0iYfugpI3Z442SHsiNfzo7KvkDkl7O7W4GAGDRWdrPRLa3SrpY0qm2D0q6WdIXJX3L\n9jWSnpd0dTb5dklXSNon6TVJny25ZgAAGqWvsI2IDbO8dGmPaUPS9UWKAgCgTbiCFAAAiRG2AAAk\nRtgCAJAYYQsAQGKELQAAiRG2AAAk1tepPwDQVBecdWtp7/Xo/ptKey8sLnS2s5g6Xbj4PQAAhO0s\nbJdyDwAAYTsLOlsAQFkI21nQ2QIAykLYzoLOFgBQFsJ2FnS2AICyELazoLMFAJSFsJ0FnS0AoCyE\n7SzobAEAZZk3bG1vsn3E9p7c2N/aftr2btv32z4pG19l+1e2d2W3O1MWnxKdLQCgLP10tpslrZ0x\ntkPS+yLi9yU9I+nzudf2R8RIdruunDIXHp0tAKAs84ZtRDwi6eiMse9FxLHs6WOSTktQW6XobAEA\nZSnjiwg+J+m+3PMzbP9E0iuS/kdE/LCEZSy4iJDtwvdt108Hvxh+DkAbjG6bmHeasU+uWIBK2qdQ\n2Nq+SdIxSd/Ihg5JWhkRL9k+X9J3bJ8bEa/0mHdU0miR5adEZzu3QXaTT0/b9p/JYpbfnleuXFlx\nNRhUPyE7c1pCdzBDH41s+zOSrpT0J5H9NY2I1yPipezxTkn7Jb2n1/wRMRYR3YjoDltDSnxmO7th\n163NP5PFLr89dzqdqsvBAAYJ2jLmW6yG6mxtr5X0V5L+c0S8lhvvSDoaEW/YPlPSaknPlVLpAqOz\n7a1XYM61rjOnXyy714Em6BWYc3WsM6cf3TZBh9unfk792SrpUUln2z5o+xpJX5V0oqQdM07xuUjS\nbtu7JH1b0nURcbTnG9ccne3bDRq0s73exp8N0DSDBu1sr9Ph9mfezjYiNvQYvnuWabdJ2la0qDqg\ns51fv+s4fcAYgPrqt0Md++QKAnYIrsMfQdvVF4E5zfw9GeY/E2W8R0vtrOuxC8PodrsxPj5edRmY\nw8ywHGZXcBnv0Ua2e27PXK4RAxs2JAlXoH6GDUnCdTCELQAAiRG2AAAkRtgCAJAYYQsAQGKELQAA\niRG2AAAkRthiYFwbGWgPro28MAhb9GXmObKDBicXtADqY+Y5soMGJxe0GBxhi6H1G7h0tED99Ru4\ndLTDIWzRt2G+VGCYLy8AkN4wXyowzJcXYEqhL4/H4tPrSwUG6VwJWqA+en2pwCCdK0HbPzpbDIxr\nIwPtwbWRFwadLYYyHZz9dLWELFBv08HZT1dLyA6HsEUhBCnQHgRpOvPuRra9yfYR23tyY7fYnrC9\nK7tdkXvt87b32d5r+2OpCgcAoCn6+cx2s6S1PcZvj4iR7LZdkmyfI2m9pHOzeb5me0lZxQIA0ETz\nhm1EPCLpaJ/vt07SvRHxekT8XNI+SWsK1AcAQOMVORr5Btu7s93MJ2djKyS9kJvmYDYGAMCiNWzY\n3iHpLEkjkg5Jum3QN7A9anvc9viQNQCoifz2PDk5WXU5QO0MFbYRcTgi3oiI45Lu0m92FU9IOj03\n6WnZWK/3GIuIbkR0h6kBQH3kt+dOp1N1OUDtDBW2tpfnnl4lafpI5Qclrbf9LttnSFot6fFiJQIA\n0Gzznmdre6ukiyWdavugpJslXWx7RFJIOiDpWkmKiCdtf0vSU5KOSbo+It5IUzoAAM0wb9hGxIYe\nw3fPMf2tkm4tUhQAAG3CtZEBAEiMsAUAIDHCFgCAxAhbAAASI2wBAEiMsAUAIDHCFgCAxAhbAAAS\nI2wBAEiMsAUAIDHCFgCAxAhbAAASI2wBAEiMsAUAIDHCFgCAxAhbAAASmzdsbW+yfcT2ntzYfbZ3\nZbcDtndl46ts/yr32p0piwcAoAmW9jHNZklflXTP9EBE/PH0Y9u3SXo5N/3+iBgpq0AAAJpu3rCN\niEdsr+r1mm1LulrSh8stCwCA9ij6me2Fkg5HxLO5sTNs/8T2D2xfWPD9AQBovH52I89lg6StueeH\nJK2MiJdsny/pO7bPjYhXZs5oe1TSaMHlA6iB/Pa8cuXKiqsB6mfoztb2Ukl/JOm+6bGIeD0iXsoe\n75S0X9J7es0fEWMR0Y2I7rA1AKiH/Pbc6XSqLgeonSK7kT8i6emIODg9YLtje0n2+ExJqyU9V6xE\nAACarZ9Tf7ZKelTS2bYP2r4me2m93roLWZIukrQ7OxXo25Kui4ijZRYMAEDT9HM08oZZxj/TY2yb\npG3FywIAoD24ghQAAIkRtgAAJEbYAgCQGGELAEBihC0AAIkRtgAAJEbYAgCQGGELAEBihC0AAIkR\ntgAAJEbYAgCQmCOi6hpke1LSLyX9oupaSnCqWI86acJ6/F5EtOZ76Wy/Kmlv1XWUoAm/O/1gPRZW\nz+25FmErSbbH2/DdtqxHvbRlPZqkLT9z1qNemr4e7EYGACAxwhYAgMTqFLZjVRdQEtajXtqyHk3S\nlp8561EvjV6P2nxmCwBAW9WpswUAoJUIWwAAEiNsAQBIjLAFACAxwhYAgMQIWwAAEiNsAQBIjLAF\nACAxwhYAgMQIWwAAEiNsAQBIjLAFACAxwhYAgMQIWwAAEiNsAQBIjLAFACAxwhYAgMQIWwAAEiNs\nAQBIjLAFACAxwhYAgMQIWwAAEiNsAQBIjLAFACAxwhYAgMQIWwAAEiNsAQBIjLAFACAxwhYAgMSS\nha3ttbb32t5n+8ZUywEAoO4cEeW/qb1E0jOSLpN0UNITkjZExFOlLwwAgJpL1dmukbQvIp6LiF9L\nulfSukTLAgCg1lKF7QpJL+SeH8zGAABYdJZWtWDbo5JGs6fnV1UHUAO/iIhO1UUUkd+eTzjhhPPf\n+973VlwRUI2dO3f23J5The2EpNNzz0/Lxt4UEWOSxiTJdvkfHAPN8XzVBRSV35673W6Mj49XXBFQ\nDds9t+dUu5GfkLTa9hm23ylpvaQHEy0LAIBaS9LZRsQx2zdI+q6kJZI2RcSTKZYFAEDdJfvMNiK2\nS9qe6v0BAGgKriAFAEBihC0AAIkRtgAAJEbYAgCQGGELAEBihC0AAIkRtgAAJEbYAgCQGGELAEBi\nhC0AAIkRtgAAJEbYAgCQGGELAEBihC0AAIkRtqiliFBEVF0GgBJs+eEF2vLDC6ouo1KELQAAiQ0d\ntrZPt/1920/ZftL2n2Xjt9iesL0ru11RXrkAADTP0gLzHpP0lxHxY9snStppe0f22u0R8eXi5QEA\n0HxDh21EHJJ0KHv8qu2fSVpRVmEAALRFKZ/Z2l4l6f2S/jUbusH2btubbJ9cxjIAAGiqwmFr+3ck\nbZP05xHxiqQ7JJ0laURTne9ts8w3anvc9njRGgBUK789T05OVl0OUDuFwtb2b2kqaL8REf8oSRFx\nOCLeiIjjku6StKbXvBExFhHdiOgWqQFA9fLbc6fTqbocoHaKHI1sSXdL+llE/F1ufHlusqsk7Rm+\nPAAAmq/I0cgflPQpST+1vSsb+4KkDbZHJIWkA5KuLVQhWmeQi1X0M+3U//sAVGGQi1X0M+3GCx8t\nUk5tFTka+UeSev2V2z58OQAAtE+RzhYYSj+d6HRHS9cK1Fs/neh0R9vWrrUfXK4RAIDECFsAABIj\nbAEASIywBQAgMcIWAIDECFsAABIjbAEASIywBQAgMS5qgVriYhZAeyzmi1lMo7MFACAxwhYAgMQI\nWwAAEiNsAQBIjLAFACAxwhYAgMQKn/pj+4CkVyW9IelYRHRtnyLpPkmrJB2QdHVE/HvRZQEA0ERl\ndbaXRMRIRHSz5zdKejgiVkt6OHsOAMCilGo38jpJW7LHWyR9ItFyAACovTLCNiR9z/ZO26PZ2LKI\nOJQ9flHSshKWAwBAI5VxucYPRcSE7d+VtMP20/kXIyJsx8yZsmAenTkOoHny2/PKlSsrrgaon8Kd\nbURMZPdHJN0vaY2kw7aXS1J2f6THfGMR0c19zgugofLbc6fTqbocoHYKha3tE2yfOP1Y0kcl7ZH0\noKSN2WQbJT1QZDkAADRZ0d3IyyTdn31Dy1JJ34yIf7b9hKRv2b5G0vOSri64HAAAGqtQ2EbEc5L+\noMf4S5IuLfLeAAC0BVeQAgAgMcIWAIDECFsAABIr4zxbAAAkSRecdevQ8z66/6YSK6kXOlvUQkQM\ndQ8ATUDYohay08cGvgeAJiBsUQt0tgDajLBFLdDZAmgzwha1QGcLoM0IW9QCnS2ANiNsUQt0tgDa\njLBFLdDZAmgzwha1QGcLoM0IW9QCnS2ANiNsUQt0tgDajLBFLdDZAmizob+IwPbZku7LDZ0p6a8l\nnSTpTyVNZuNfiIjtQ1eIRSEiZHvgewBogqHDNiL2ShqRJNtLJE1Iul/SZyXdHhFfLqVCLAp0tgDa\nrKyv2LtU0v6IeJ4/ggCweLX5a/KKKOsz2/WStuae32B7t+1Ntk8uaRkAADRS4bC1/U5JH5f0D9nQ\nHZLO0tQu5kOSbptlvlHb47bHi9YAoFr57XlycnL+GYBFpozdyJdL+nFEHJak6XtJsn2XpId6zRQR\nY5LGsuk4j0P9nc7CbnrUUX577na7bM+SRrdNzDvN2CdXLEAlqIMydiNvUG4Xsu3ludeukrSnhGW0\nXr/njXJ+KVB//QTtINOh+Qp1trZPkHSZpGtzw1+yPSIpJB2Y8RpmGCY8p+ehywXqZZjwnJ6HLrfd\nCoVtRPxS0rtnjH2qUEWL3FwBSlcLNMtcAUpXu7hwBakKzQzP+TrVma8TvkB9zAzP+TrVma8Tvu1G\n2FZk0KCdbToCF6jeoEE723QEbnsRtjUw6GevfFYL1Negn73yWe3iQNgCAJAYYQsAQGKELQAAiRG2\nAAAkRtgCAJAYYQsAQGKEbQ0Meq4s59YC9TXoubKcW7s4ELYVGfbiFMNeDANAOsNenGLYi2GgeQjb\nCg0auAQtUF+DBi5Bu7iU8X22KBG7iIH2YBcxphG2ic0Vnrbf7E4HCVk6WqAar+6afWfgiSPH3+xO\nBwlZOtrFgd3ICQ2yW3jYLyIAsDDmCtqZrw/7RQRoLzrbRAY54Gk6QAlSoJ7mC9r8dCeOHJdEkOKt\n+voNsr3J9hHbe3Jjp9jeYfvZ7P7kbNy2v2J7n+3dts9LVXxdcSoP0B79Bu2w02Nx6Pe3YrOktTPG\nbpT0cESslvRw9lySLpe0OruNSrqjeJkAADRXX2EbEY9IOjpjeJ2kLdnjLZI+kRu/J6Y8Jukk28vL\nKBYAgCYq8pntsog4lD1+UdKy7PEKSS/kpjuYjR1SBTjKF2iPxz94Sd/TrvmX7yesBBhMKR8uxFSi\nDfTBo+1R2+O2x8uoAUB18tvz5ORk1eUAtVMkbA9P7x7O7o9k4xOSTs9Nd1o29hYRMRYR3YjoFqgB\nQA3kt+dOp1N1OUDtFAnbByVtzB5vlPRAbvzT2VHJH5D0cm53MwAAi06/p/5slfSopLNtH7R9jaQv\nSrrM9rOSPpI9l6Ttkp6TtE/SXZL+a+lV19ygn/3yWTFQX9PnzaaaHotDXwdIRcSGWV66tMe0Ien6\nIkW1ge2+Ds4iaIH6O3HkeF/nzxK0mA1nXyc0X5AStEBzzBekBC3mwuUaEyNQgfYgUDEsOlsAABIj\nbAEASIywBQAgsdZ/ZstnpkB7cAlGNBWdLQAAiRG2AAAkRtgCAJAYYQsAQGKELQAAiRG2AAAkRtgC\nAJAYYQsAQGKELQAAiRG2AAAkNm/Y2t5k+4jtPbmxv7X9tO3dtu+3fVI2vsr2r2zvym53piweAIAm\n6Kez3Sxp7YyxHZLeFxG/L+kZSZ/PvbY/Ikay23XllAkAQHPNG7YR8YikozPGvhcRx7Knj0k6LUFt\nAAC0Qhmf2X5O0j/lnp9h+ye2f2D7whLeHwCARiv0FXu2b5J0TNI3sqFDklZGxEu2z5f0HdvnRsQr\nPeYdlTRaZPkA6iG/Pa9cubLiaoD6Gbqztf0ZSVdK+pOICEmKiNcj4qXs8U5J+yW9p9f8ETEWEd2I\n6A5bA4B6yG/PnU6n6nKA2hkqbG2vlfRXkj4eEa/lxju2l2SPz5S0WtJzZRQKAEBTzbsb2fZWSRdL\nOtX2QUk3a+ro43dJ2mFbkh7Ljjy+SNLf2P4PScclXRcRR3u+MQAAi8S8YRsRG3oM3z3LtNskbSta\nFAAAbcIVpAAASIywBQAgMcIWA8kOPAfQAv76lVWXsGgQthgYgQu0B4G7MAhbDIXABdqDwE2PsMXQ\nCFygPQjctAhbFELgAu1B4KZD2KIwAhdoDwI3DcIWpSBwgfYgcMtH2KI0BC7QHgRuuQhblIrABdqD\nwC0PYYvSEbhAexC45SBskQSBC7QHgVscYYtkCFygPQjcYghbJEXgAu1B4A6PsEVyBC7QHgTucOYN\nW9ubbB+xvSc3dovtCdu7stsVudc+b3uf7b22P5aqcDQLgQu0B4E7uH46282S1vYYvz0iRrLbdkmy\nfY6k9ZLOzeb5mu0lZRWLZiNwgfYgcAczb9hGxCOSjvb5fusk3RsRr0fEzyXtk7SmQH1oGQIXaA8C\nt39FPrO9wfbubDfzydnYCkkv5KY5mI0BbyJwgfYgcPszbNjeIeksSSOSDkm6bdA3sD1qe9z2+JA1\noMEI3HbJb8+Tk5NVl4MFRuDOb6iwjYjDEfFGRByXdJd+s6t4QtLpuUlPy8Z6vcdYRHQjojtMDWg+\nArc98ttzp9OpuhxUgMCd21Bha3t57ulVkqaPVH5Q0nrb77J9hqTVkh4vViLajMAF2oPAnd3S+Saw\nvVXSxZJOtX1Q0s2SLrY9IikkHZB0rSRFxJO2vyXpKUnHJF0fEW+kKR1VsF11CQBKEtc+VHUJi8a8\nYRsRG3oM3z3H9LdKurVIUQAAtAlXkAIAIDHCFgCAxAhbAAASI2wBAEiMsAUAIDHCFgCAxAhbAAAS\nI2wBAEiMsAUAIDHCFgCAxAhbAAASI2wBAEiMsAUAIDHCFgCAxAhbAAASI2wBAEhs3rC1vcn2Edt7\ncmP32d6V3Q7Y3pWNr7L9q9xrd6YsHgCAJljaxzSbJX1V0j3TAxHxx9OPbd8m6eXc9PsjYqSsAgEA\naLp5wzYiHrG9qtdrti3pakkfLrcsAADao+hnthdKOhwRz+bGzrD9E9s/sH1hwfcHAKDx+tmNPJcN\nkrbmnh+StDIiXrJ9vqTv2D43Il6ZOaPtUUmjBZcPoAby2/PKlSsrrgaon6E7W9tLJf2RpPumxyLi\n9Yh4KXu8U9J+Se/pNX9EjEVENyK6w9YAoB7y23On06m6HKB2iuxG/oikpyPi4PSA7Y7tJdnjMyWt\nlvRcsRIBAGi2fk792SrpUUln2z5o+5rspfV66y5kSbpI0u7sVKBvS7ouIo6WWTAAAE3Tz9HIG2YZ\n/0yPsW2SthUvCwCA9uAKUgAAJEbYAgCQGGELAEBihC0AAIkRtgAAJEbYAgCQGGELAEBihC0AAIkR\ntgAAJEbYAgCQGGELAEBihC0AAIk5IqquQbYnJf1S0i+qrqUEp4r1qJMmrMfvRURrvgTW9quS9lZd\nRwma8LvTD9ZjYfXcnmsRtpJke7wNXyTPetRLW9ajSdryM2c96qXp68FuZAAAEiNsAQBIrE5hO1Z1\nASVhPeqlLevRJG35mbMe9dLo9ajNZ7YAALRVnTpbAABaibAFACAxwhYAgMQIWwAAEiNsAQBIjLAF\nACAxwhYAgMQIWwAAEiNsAQBIjLAFACAxwhYAgMQIWwAAEiNsAQBIjLAFACAxwhYAgMQIWwAAEiNs\nAQBIjLAFACAxwhYAgMQIWwAAEiNsAQBIjLAFACAxwhYAgMQIWwAAEiNsAQBIjLAFACAxwhYAgMQI\nWwAAEiNsAQBILFnY2l5re6/tfbZvTLUcAADqzhFR/pvaSyQ9I+kySQclPSFpQ0Q8VfrCAACouVSd\n7RpJ+yLiuYj4taR7Ja1LtCwAAGptaaL3XSHphdzzg5L+MD+B7VFJo9nT8xPVATTBLyKiU3URReS3\n5xNOOOH89773vRVXBFRj586dPbfnVGE7r4gYkzQmSbbL35cNNMfzVRdQVH577na7MT4+XnFFQDVs\n99yeU+1GnpB0eu75adkYAACLTqqwfULSattn2H6npPWSHky0LAAAai3JbuSIOGb7BknflbRE0qaI\neDLFsgAAqLtkn9lGxHZJ2455ClEAABDTSURBVFO9PwAATcEVpAAASIywBQAgMcIWAIDECFsAABIj\nbAEASIywBQAgMcIWAIDECFsAABIjbAEASIywBQAgMcIWAIDECFsAABIjbAEASIywBQAgMcIWAIDE\nCFsAABIbOmxtn277+7afsv2k7T/Lxm+xPWF7V3a7orxyAQBonqUF5j0m6S8j4se2T5S00/aO7LXb\nI+LLxcsDAKD5hg7biDgk6VD2+FXbP5O0oqzCAABoi1I+s7W9StL7Jf1rNnSD7d22N9k+eZZ5Rm2P\n2x4vowYA1clvz5OTk1WXA9RO4bC1/TuStkn684h4RdIdks6SNKKpzve2XvNFxFhEdCOiW7QGANXK\nb8+dTqfqcoDaKRS2tn9LU0H7jYj4R0mKiMMR8UZEHJd0l6Q1xcsEAKC5ihyNbEl3S/pZRPxdbnx5\nbrKrJO0ZvjwAAJqvyNHIH5T0KUk/tb0rG/uCpA22RySFpAOSri1UIQAADVfkaOQfSXKPl7YPXw4A\nAO3DFaRQmYiougQAJfHXr6y6hFojbFEpAhdoDwJ3doQtKkfgAu1B4PZG2KIWCFygPQjctyNsURsE\nLtAeBO5bEbaoFQIXaA8C9zcIW9QOgQu0B4E7hbBFLRG4QHsQuIQtaozABdpjsQdukcs1Asn1E7hT\nl+kGUHfn//2dOv+3/8uc04x9sp1fi05ni8aLCLpgoMbO++0rdd5vT3W2O1/733NOO7ptQqPbJhai\nrAVF2KI1CFygfqZDNm++wJXUusBlNzIaZeYu45kBGxHsVgZqolfQ/vi1h968j2sfestrMwN2dNtE\na3Yr09miUWaGa69gpcMFqjdX0E6bedBUr2BtS4dL2KJx+glcAPUyM2in9RO4bUDYopHmC1y6W6A6\nM7va2YJ22nyB24butnDY2j5g+6e2d9kez8ZOsb3D9rPZ/cnFSwXeig4XqL/5gnZa2zvcsjrbSyJi\nJCK62fMbJT0cEaslPZw9B0pHBwu0R5svfJFqN/I6SVuyx1skfSLRcgACF2iRtgZuGWEbkr5ne6ft\n0WxsWUQcyh6/KGnZzJlsj9oen971DBRB4FYrvz1PTk5WXQ4aro2BW0bYfigizpN0uaTrbV+UfzGm\n/gq+7S9hRIxFRDe36xkohMCtTn577nQ6VZeDFmhb4Ba+qEVETGT3R2zfL2mNpMO2l0fEIdvLJR0p\nuhy0T/6ApnxQFjnQicAFqpG/QEX+6OGdn7pu6Pdsw1HI0wp1trZPsH3i9GNJH5W0R9KDkjZmk22U\n9ECR5QAA0GRFdyMvk/Qj2/8m6XFJ/yci/lnSFyVdZvtZSR/JngN9GbY7pasF6mfY7rRNXa1UMGwj\n4rmI+IPsdm5E3JqNvxQRl0bE6oj4SEQcLadctFXRi1Jwzi1QH0UvSjFz+jacc8sVpFBb/QYuHS1Q\nf/0Gbts62mmELWpjmC8V6PU6XS1QvWG+VKDX623oaiW+Yg81Y7vn1+YNMj+Aehj75IqeX5s3yPxt\nQWeL2hk2MAlaoH6GDcw2Ba1EZ4uamg7OfrpaQhaot+ng7KerbVvITiNsUWsEKdAebQ3SfrAbGQCA\nxAhbAAASI2wBAEiMsAUAIDHCFgCAxAhbAAASI2wBAEiMsAUAIDHCFgCAxAhbAAASG/pyjbbPlnRf\nbuhMSX8t6SRJfyppMhv/QkRsH7pCAAAabuiwjYi9kkYkyfYSSROS7pf0WUm3R8SXS6kQAICGK2s3\n8qWS9kfE8yW9HwAArVFW2K6XtDX3/Abbu21vsn1yrxlsj9oetz1eUg0AKpLfnicnJ+efAVhk3M/3\nhc75BvY7Jf0/SedGxGHbyyT9QlJI+p+SlkfE5+Z5j2JFAM22MyK6VRdRlm63G+Pj/B8ai5Ptnttz\nGZ3t5ZJ+HBGHJSkiDkfEGxFxXNJdktaUsAwAABqrjLDdoNwuZNvLc69dJWlPCcsAAKCxhj4aWZJs\nnyDpMknX5oa/ZHtEU7uRD8x4DQCARadQ2EbELyW9e8bYpwpVBABAy3AFKQAAEiNsUUhEqOgR7QDq\nYcsPL9CWH15QdRmtRNgCAJAYYQsAQGKELQAAiRG2AAAkRtgCAJAYYQsAQGKELQAAiRG2AAAkVuhy\njWi3QS5W0c+0touUA6CAQS5W0c+0Gy98tEg5iw6dLQAAidHZYlb9dKLTHS1dK1Bv/XSi0x0tXWv5\n6GwBAEiMsAUAIDHCFgCAxPoKW9ubbB+xvSc3dortHbafze5PzsZt+yu299nebfu8VMUDANAE/R4g\ntVnSVyXdkxu7UdLDEfFF2zdmz/+7pMslrc5ufyjpjuy+VgY5rYWDf4B6e/yDl/Q97Zp/+X7CSoDe\n+upsI+IRSUdnDK+TtCV7vEXSJ3Lj98SUxySdZHt5GcUCANBERT6zXRYRh7LHL0palj1eIemF3HQH\ns7G3sD1qe9z2eIEaANRAfnuenJysuhygdko5QCqm9sn2v192ap6xiOhGRLeMGgBUJ789dzqdqssB\naqfIRS0O214eEYey3cRHsvEJSafnpjstG0ML8Xk20B5czCKdIp3tg5I2Zo83SnogN/7p7KjkD0h6\nObe7GQCARaevztb2VkkXSzrV9kFJN0v6oqRv2b5G0vOSrs4m3y7pCkn7JL0m6bMl1wwAQKP0FbYR\nsWGWly7tMW1Iur5IUcBMESHbfd8DqK8Lzrq172kf3X9TwkoWDleQQiNMB2i/9wBQJ4QtGmH6IiT9\n3gNAnRC2aAQ6WwBNtmi/z5Y/ys3CZ7aYC5dgRN3R2aIR6GwBNBlhi0bgM1sATUbYohHobAE0GWGL\nRqCzBdBkhC0agc4WQJMRtmgEOlsATbZoT/1Bs9DZAu3RlkswDoLOFgCAxAhbAAASI2wBAEiMz2zn\n0M/BNnxGCDTDq7vm7y1OHDm+AJVgMaKznUW/R7Vy9CtQf/0E7SDTAYOa9zfL9ibbR2zvyY39re2n\nbe+2fb/tk7LxVbZ/ZXtXdrszZfEpRMTAATrMPADSe3XXOwYO0GHmAebTz2/UZklrZ4ztkPS+iPh9\nSc9I+nzutf0RMZLdriunTAAAmmvesI2IRyQdnTH2vYg4lj19TNJpCWpbcEW7U7pboD6Kdqd0tyhT\nGb9Nn5P0T7nnZ9j+ie0f2L5wtplsj9oetz1eQg2FlRWUBC4Wo/z2PDk5WXU5pQUlgYuyFPpNsn2T\npGOSvpENHZK0MiLeL+kvJH3T9n/qNW9EjEVENyK6RWoAUL389tzpdKouB6idocPW9mckXSnpTyJr\n5yLi9Yh4KXu8U9J+Se8poU4AABprqLC1vVbSX0n6eES8lhvv2F6SPT5T0mpJz5VRKAAATTXvRS1s\nb5V0saRTbR+UdLOmjj5+l6Qd2UUdHsuOPL5I0t/Y/g9JxyVdFxFHe74xAACLxLxhGxEbegzfPcu0\n2yRtK1oUAABtwqF2AAAkRtgCAJAYYQsAQGKEbaasb+/hW4CA6pX17T18CxDKQtjmFA1Kghaoj6JB\nSdCiTIQtAACJ8eXxM0x3p4Nc45iOFqin6e50kGsc09EiBTrbWfQboAQtUH/9BihBi1TobOdAkALt\nQZCiSnS2AAAkRtgCAJAYYQsAQGKELQAAiRG2AAAkRtgCAJAYYQsAQGLzhq3tTbaP2N6TG7vF9oTt\nXdntitxrn7e9z/Ze2x9LVTgAAE3RT2e7WdLaHuO3R8RIdtsuSbbPkbRe0rnZPF+zvaSsYgEAaKJ5\nwzYiHpF0tM/3Wyfp3oh4PSJ+LmmfpDUF6gMAoPGKfGZ7g+3d2W7mk7OxFZJeyE1zMBt7G9ujtsdt\njxeoAUAN5LfnycnJqssBamfYsL1D0lmSRiQdknTboG8QEWMR0Y2I7pA1AKiJ/Pbc6XSqLgeonaHC\nNiIOR8QbEXFc0l36za7iCUmn5yY9LRsDAGDRGipsbS/PPb1K0vSRyg9KWm/7XbbPkLRa0uPFSgQA\noNnm/Yo921slXSzpVNsHJd0s6WLbI5JC0gFJ10pSRDxp+1uSnpJ0TNL1EfFGmtIBAGiGecM2Ijb0\nGL57julvlXRrkaIAAGgTriAFAEBihC0AAIkRtgAAJEbYAgCQGGELAEBihC0AAIkRtgAAJEbYAgCQ\nGGELAEBihC0AAIkRtgAAJEbYAgCQGGELAEBihC0AAIkRtgAAJEbYAgCQ2Lxha3uT7SO29+TG7rO9\nK7sdsL0rG19l+1e51+5MWTwAAE2wtI9pNkv6qqR7pgci4o+nH9u+TdLLuen3R8RIWQUCANB084Zt\nRDxie1Wv12xb0tWSPlxuWQAAtEfRz2wvlHQ4Ip7NjZ1h+ye2f2D7wtlmtD1qe9z2eMEaAFQsvz1P\nTk5WXQ5QO0XDdoOkrbnnhyStjIj3S/oLSd+0/Z96zRgRYxHRjYhuwRoAVCy/PXc6narLAWpn6LC1\nvVTSH0m6b3osIl6PiJeyxzsl7Zf0nqJFAgDQZEU6249IejoiDk4P2O7YXpI9PlPSaknPFSsRAIBm\n6+fUn62SHpV0tu2Dtq/JXlqvt+5ClqSLJO3OTgX6tqTrIuJomQUDANA0/RyNvGGW8c/0GNsmaVvx\nsgAAaA+uIAUAQGKELQAAiRG2AAAkRtgCAJAYYQsAQGKELQAAiRG2AAAkRtgCAJAYYQsAQGKELQAA\niRG2AAAk5oiougbZnpT0S0m/qLqWEpwq1qNOmrAevxcRrfkSWNuvStpbdR0laMLvTj9Yj4XVc3uu\nRdhKku3xNnyRPOtRL21ZjyZpy8+c9aiXpq8Hu5EBAEiMsAUAILE6he1Y1QWUhPWol7asR5O05WfO\netRLo9ejNp/ZAgDQVnXqbAEAaKXKw9b2Wtt7be+zfWPV9QzC9gHbP7W9y/Z4NnaK7R22n83uT666\nzplsb7J9xPae3FjPuj3lK9m/z27b51VX+VvNsh632J7I/k122b4i99rns/XYa/tj1VTdbmzPC4/t\nuRnbc6Vha3uJpP8l6XJJ50jaYPucKmsawiURMZI7JP1GSQ9HxGpJD2fP62azpLUzxmar+3JJq7Pb\nqKQ7FqjGfmzW29dDkm7P/k1GImK7JGW/V+slnZvN87Xs9w8lYXuuzGaxPdd+e666s10jaV9EPBcR\nv5Z0r6R1FddU1DpJW7LHWyR9osJaeoqIRyQdnTE8W93rJN0TUx6TdJLt5QtT6dxmWY/ZrJN0b0S8\nHhE/l7RPU79/KA/bcwXYnpuxPVcdtiskvZB7fjAba4qQ9D3bO22PZmPLIuJQ9vhFScuqKW1gs9Xd\nxH+jG7JdZJtyu/2auB5N0/SfMdtzPbVie646bJvuQxFxnqZ2zVxv+6L8izF1qHfjDvduat2ZOySd\nJWlE0iFJt1VbDhqE7bl+WrM9Vx22E5JOzz0/LRtrhIiYyO6PSLpfU7sxDk/vlsnuj1RX4UBmq7tR\n/0YRcTgi3oiI45Lu0m92LTVqPRqq0T9jtuf6adP2XHXYPiFpte0zbL9TUx94P1hxTX2xfYLtE6cf\nS/qopD2aqn9jNtlGSQ9UU+HAZqv7QUmfzo5i/ICkl3O7p2pnxudPV2nq30SaWo/1tt9l+wxNHSDy\n+ELX13Jsz/XB9lw3EVHpTdIVkp6RtF/STVXXM0DdZ0r6t+z25HTtkt6tqaP/npX0fyWdUnWtPWrf\nqqldMv+hqc86rpmtbknW1BGm+yX9VFK36vrnWY+/z+rcrakNcnlu+puy9dgr6fKq62/jje25ktrZ\nnhuwPXMFKQAAEqt6NzIAAK1H2AIAkBhhCwBAYoQtAACJEbYAACRG2AIAkBhhCwBAYoQtAACJ/X9O\nJChJb/7R0gAAAABJRU5ErkJggg==\n",
"text/plain": [
"<Figure size 576x864 with 6 Axes>"
]
},
"metadata": {
"tags": []
}
}
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "_Ig85VI7Y9B0",
"colab_type": "text"
},
"source": [
"# Prepapre Dataset and DataLoader"
]
},
{
"cell_type": "code",
"metadata": {
"id": "dIveqpflYG0l",
"colab_type": "code",
"colab": {}
},
"source": [
"def numpy_collate(batch):\n",
" if isinstance(batch[0], onp.ndarray):\n",
" return onp.stack(batch)\n",
" elif isinstance(batch[0], (tuple,list)):\n",
" transposed = zip(*batch)\n",
" return [numpy_collate(samples) for samples in transposed]\n",
" else:\n",
" #print(\"--- print(len(batch)):\", len(batch))\n",
" return onp.array(batch)\n",
"\n",
"class NumpyLoader(DataLoader):\n",
" def __init__(self, dataset, batch_size=1,\n",
" shuffle=False, sampler=None,\n",
" batch_sampler=None, num_workers=0,\n",
" pin_memory=False, drop_last=False,\n",
" timeout=0, worker_init_fn=None):\n",
" super(self.__class__, self).__init__(dataset,\n",
" batch_size=batch_size,\n",
" shuffle=shuffle,\n",
" sampler=sampler,\n",
" batch_sampler=batch_sampler,\n",
" num_workers=num_workers,\n",
" collate_fn=numpy_collate,\n",
" pin_memory=pin_memory,\n",
" drop_last=drop_last,\n",
" timeout=timeout,\n",
" worker_init_fn=worker_init_fn)\n",
" \n",
"\n",
"class FlattenAndCast(object):\n",
" def __call__(self, pic):\n",
" return onp.ravel(onp.array(pic, dtype=np.float32))\n",
"\n",
"class SimDataset(Dataset):\n",
" def __init__(self, count, transform=None):\n",
" self.input_images, self.target_masks = simulation.generate_random_data(64, 64, count=count)\n",
" self.transform = transform\n",
"\n",
" def __len__(self):\n",
" return len(self.input_images)\n",
"\n",
" def __getitem__(self, idx):\n",
" image = self.input_images[idx]\n",
" mask = self.target_masks[idx]\n",
" if self.transform:\n",
" image = self.transform(image)\n",
"\n",
"\n",
" return [image, mask]\n",
"\n",
"\n",
"class SimDataset2(Dataset):\n",
" def __init__(self, count, transform=None):\n",
" self.input_images, self.target_masks = simulation.generate_random_data(64, 64, count=count)\n",
" self.target_masks = self.target_masks.transpose((0,2,3,1))\n",
" self.transform = transform\n",
"\n",
" def __len__(self):\n",
" return len(self.input_images)\n",
"\n",
" def __getitem__(self, idx):\n",
" image = self.input_images[idx]\n",
" mask = self.target_masks[idx]\n",
" if self.transform:\n",
" image = self.transform(image)\n",
"\n",
"\n",
" return [image, mask]\n",
"\n",
"# use the same transformations for train/val in this example\n",
"trans = transforms.Compose([\n",
" transforms.ToTensor(),\n",
" transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225]), # imagenet,\n",
"])\n",
"\n",
"\n",
"nrmFcn = lambda x: (x + 0.0)/255.0\n",
"castFcn = lambda x: x.astype(onp.float32)\n",
"trans2 = transforms.Compose([ transforms.Lambda(nrmFcn), transforms.Lambda(castFcn) ])\n",
"\n",
"train_set = SimDataset2(2000, transform = trans2)\n",
"val_set = SimDataset2(200, transform = trans2)\n",
"\n",
"image_datasets = {\n",
" 'train': train_set, 'val': val_set\n",
"}\n",
"\n",
"batch_size = 25\n",
"\n",
"dataloaders = {\n",
" 'train': NumpyLoader(train_set, batch_size=batch_size, shuffle=True, num_workers=0),\n",
" 'val': NumpyLoader(val_set, batch_size=batch_size, shuffle=True, num_workers=0)\n",
"}"
],
"execution_count": 0,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"id": "bHR_xdjUZKka",
"colab_type": "code",
"colab": {}
},
"source": [
"inputs, masks = next(iter(dataloaders['train']))\n"
],
"execution_count": 0,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"id": "LvoR0n0r3Y6X",
"colab_type": "code",
"colab": {}
},
"source": [
"import numpy as onp\n",
"from torch.utils.data import Dataset, DataLoader\n",
"from torchvision import transforms, datasets, models\n",
"\n",
"\n",
"def numpy_collate(batch):\n",
" if isinstance(batch[0], onp.ndarray):\n",
" return onp.stack(batch)\n",
" elif isinstance(batch[0], (tuple,list)):\n",
" transposed = zip(*batch)\n",
" return [numpy_collate(samples) for samples in transposed]\n",
" else:\n",
" return onp.array(batch)\n",
"\n",
"class NumpyLoader(DataLoader):\n",
" def __init__(self, dataset, batch_size=1,\n",
" shuffle=False, sampler=None,\n",
" batch_sampler=None, num_workers=0,\n",
" pin_memory=False, drop_last=False,\n",
" timeout=0, worker_init_fn=None):\n",
" super(self.__class__, self).__init__(dataset,\n",
" batch_size=batch_size,\n",
" shuffle=shuffle,\n",
" sampler=sampler,\n",
" batch_sampler=batch_sampler,\n",
" num_workers=num_workers,\n",
" collate_fn=numpy_collate,\n",
" pin_memory=pin_memory,\n",
" drop_last=drop_last,\n",
" timeout=timeout,\n",
" worker_init_fn=worker_init_fn)\n",
" \n",
"\n",
"class FlattenAndCast(object):\n",
" def __call__(self, pic):\n",
" return onp.ravel(onp.array(pic, dtype=np.float32))\n",
"\n",
"class SimDataset(Dataset):\n",
" def __init__(self, count, transform=None):\n",
" self.input_images, self.target_masks = simulation.generate_random_data(64, 64, count=count)\n",
" self.transform = transform\n",
"\n",
" def __len__(self):\n",
" return len(self.input_images)\n",
"\n",
" def __getitem__(self, idx):\n",
" image = self.input_images[idx]\n",
" mask = self.target_masks[idx]\n",
" if self.transform:\n",
" image = self.transform(image)\n",
"\n",
"\n",
" return [image, mask]\n",
"\n",
"\n",
"class SimDataset2(Dataset):\n",
" def __init__(self, count, transform=None):\n",
" self.input_images, self.target_masks = simulation.generate_random_data(64, 64, count=count)\n",
" self.target_masks = self.target_masks.transpose((0,2,3,1))\n",
" self.transform = transform\n",
"\n",
" def __len__(self):\n",
" return len(self.input_images)\n",
"\n",
" def __getitem__(self, idx):\n",
" image = self.input_images[idx]\n",
" mask = self.target_masks[idx]\n",
" if self.transform:\n",
" image = self.transform(image)\n",
"\n",
"\n",
" return [image, mask]\n",
"\n",
"# use the same transformations for train/val in this example\n",
"trans = transforms.Compose([\n",
" transforms.ToTensor(),\n",
" transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225]), # imagenet,\n",
"])\n",
"\n",
"\n",
"\n",
"nrmFcn = lambda x: (x + 0.0)/255.0\n",
"castFcn = lambda x: x.astype(onp.float32)\n",
"trans2 = transforms.Compose([ transforms.Lambda(nrmFcn), transforms.Lambda(castFcn) ])\n",
"\n",
"train_set = SimDataset2(2000, transform = trans2)\n",
"val_set = SimDataset2(200, transform = trans2)\n",
"\n",
"image_datasets = {\n",
" 'train': train_set, 'val': val_set\n",
"}\n",
"\n",
"batch_size = 25\n",
"\n",
"dataloaders = {\n",
" 'train': NumpyLoader(train_set, batch_size=batch_size, shuffle=True, num_workers=0),\n",
" 'val': NumpyLoader(val_set, batch_size=batch_size, shuffle=True, num_workers=0)\n",
"}\n",
"\n"
],
"execution_count": 0,
"outputs": []
},
{
"cell_type": "markdown",
"metadata": {
"id": "1bTj-wARZ3Kv",
"colab_type": "text"
},
"source": [
"# Build UNet with neuralTangent"
]
},
{
"cell_type": "code",
"metadata": {
"id": "3H1QrY2UZ2WT",
"colab_type": "code",
"colab": {}
},
"source": [
"from __future__ import print_function, division, absolute_import\n",
"import jax.numpy as np\n",
"from jax import grad, jit, vmap\n",
"from jax import random\n",
"\n",
"\n",
"from jax.experimental.stax import (LogSoftmax, sigmoid)"
],
"execution_count": 0,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"id": "paMBmaJ1aRjA",
"colab_type": "code",
"colab": {}
},
"source": [
"from neural_tangents import stax\n",
"from neural_tangents.stax import (Conv, Relu, AvgPool, FanOut, FanInSum, Identity, ConvTranspose)\n",
"\n",
"import jax.experimental.stax as ostax\n"
],
"execution_count": 0,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"id": "KRj-W3r4a_5K",
"colab_type": "code",
"colab": {}
},
"source": [
""
],
"execution_count": 0,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"id": "0oiufzgXfP8F",
"colab_type": "code",
"colab": {}
},
"source": [
"def double_conv(kernel_size, filters):\n",
" ks = kernel_size\n",
" filters1, filters2 = filters\n",
" Main = stax.serial(\n",
" Conv(filters1, (ks,ks),padding='SAME'), Relu(),\n",
" Conv(filters2, (ks,ks),padding='SAME')\n",
" )\n",
" return Main\n",
"\n",
"def upSample(filters, resize_rate=2):\n",
" \"filters : number of filters\"\n",
" return ConvTranspose(filters, (resize_rate,resize_rate), strides=(2,2))"
],
"execution_count": 0,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"id": "ncfZjDiJfeif",
"colab_type": "code",
"colab": {}
},
"source": [
"def UNetModel(n_class):\n",
" # model stuff\n",
" dconv_down1 = double_conv(3, [64, 64])\n",
" dconv_down2 = double_conv(3, [128, 128])\n",
" dconv_down3 = double_conv(3, [256, 256])\n",
" dconv_down4 = double_conv(3, [512, 512]) \n",
"\n",
" avgpool = AvgPool((2,2), strides=(2,2))\n",
"\n",
" dconv_up3 = double_conv(3, [256, 256]) # 256\n",
" dconv_up2 = double_conv(3, [128, 128]) # 128\n",
" dconv_up1 = double_conv(3, [64, 64])\n",
"\n",
" d1 = stax.serial( dconv_down1, avgpool ) # 64\n",
" d2 = stax.serial( dconv_down2, avgpool ) # 128\n",
" d3 = stax.serial( dconv_down3, avgpool ) # 256\n",
" d4 = stax.serial( dconv_down4, upSample(512, 2) ) # 512 \n",
"\n",
" conv_last = Conv(n_class, (1,1),padding='SAME')\n",
"\n",
"\n",
" # making the pipeline\n",
" enc4 = stax.serial(avgpool, d4) # 512\n",
" enc3 = stax.serial(avgpool, dconv_down3, Relu(), \n",
" FanOut(2), \n",
" stax.parallel(dconv_up3, stax.serial(enc4, dconv_up3)), \n",
" FanInSum(), Relu(), upSample( 256+512, 2) )\n",
" \n",
" enc2 = stax.serial(avgpool, dconv_down2, Relu(),\n",
" FanOut(2), \n",
" stax.parallel(dconv_up2, stax.serial(enc3, dconv_up2)), \n",
" FanInSum(), Relu(), upSample(128+256+512, 2) )\n",
" \n",
" enc1 = stax.serial(dconv_down1, Relu(),\n",
" FanOut(2),\n",
" stax.parallel(dconv_up1, stax.serial(enc2, dconv_up1)), \n",
" FanInSum(), Relu() )\n",
"\n",
" return stax.serial(enc1, conv_last)"
],
"execution_count": 0,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"id": "PoZUiAfQkwKW",
"colab_type": "code",
"colab": {}
},
"source": [
"rng = random.PRNGKey(0)\n",
"\n",
"step_size = 0.001\n",
"num_epochs = 10\n",
"batch_size = 128\n",
"momentum_mass = 0.9\n",
"\n",
"\n",
"init_params, predict, kernel_fn = UNetModel(6)\n",
"\n",
"opt_init, opt_update, get_params = optimizers.momentum(step_size, \n",
" mass=momentum_mass)\n",
"\n",
"\n",
"@jit\n",
"def update(i, opt_state, batch):\n",
" params = get_params(opt_state)\n",
" return opt_update(i, grad(loss)(params, batch), opt_state) \n",
"\n",
"@jit\n",
"def loss(params, batch):\n",
" inputs, targets = batch\n",
" #inputs = inputs.numpy().transpose((0,2,3,1))\n",
" #targets = targets.numpy().transpose((0,2,3,1))\n",
"\n",
" logits = predict(params, inputs )\n",
" logits = ostax.logsoftmax(logits) # log normalize\n",
" return -np.mean(np.sum(logits * targets, axis=[1,2,3])) # cross entropy loss\n",
"\n",
"\n",
"\n",
"def print_metrics(metrics, epoch_samples, phase): \n",
" outputs = []\n",
" for k in metrics[phase].keys():\n",
" outputs.append(\"{}: {:4f}\".format(k, metrics[phase][k] / epoch_samples))\n",
" \n",
" print(\"{}: {}\".format(phase, \", \".join(outputs))) \n",
"\n",
"\n",
"@jit\n",
"def binary_cross_entropy_with_logits(x, y):\n",
" # compute -y * log(sigmoid(x)) - (1 - y) * log(1 - sigmoid(x))\n",
" # Ref: https://www.tensorflow.org/api_docs/python/tf/nn/sigmoid_cross_entropy_with_logits\n",
" return np.clip(x, 0) + np.log1p(np.exp(-np.abs(x))) - x * y\n",
"\n",
"\n",
"@jit\n",
"def dice_loss(pred, target, smooth = 1.): \n",
"\n",
" #intersection = (pred * target).sum(dim=2).sum(dim=2)\n",
" intersection = (pred * target).sum(axis=1).sum(axis=1)\n",
" \n",
" loss = (1 - ((2. * intersection + smooth) / (pred.sum(axis=1).sum(axis=1) + target.sum(axis=1).sum(axis=1) + smooth)))\n",
" \n",
" return loss.mean()\n",
"\n",
"def makeReport(pred, target):\n",
" bce = binary_cross_entropy_with_logits(pred, target).sum()\n",
" \n",
" pred = sigmoid(pred)\n",
" dice = dice_loss(pred, target)\n",
" \n",
" return bce, dice"
],
"execution_count": 0,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"id": "9ZCffqqkl2Af",
"colab_type": "code",
"colab": {}
},
"source": [
"#init_testFcn, apply_testFcn = enc1\n",
"#init_testFcn, apply_testFcn = UNetModel(6)\n",
"\n",
"\n",
"random_image = inputs\n",
"random_image_masks = masks\n",
"\n",
"key = random.PRNGKey(1)\n",
"_, params = init_params(key, \n",
" input_shape=random_image.shape)\n",
"\n",
"y = predict(params, random_image )"
],
"execution_count": 0,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"id": "F6Qfx5r7nlZZ",
"colab_type": "code",
"outputId": "cf5bcab2-1709-448f-9cd4-7dc898a6ef01",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 285
}
},
"source": [
"plt.imshow(y[1,:,:,1])"
],
"execution_count": 15,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"<matplotlib.image.AxesImage at 0x7f6c93c25080>"
]
},
"metadata": {
"tags": []
},
"execution_count": 15
},
{
"output_type": "display_data",
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAPsAAAD7CAYAAACscuKmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n2A0jI/x/3d/t4aazpB4AAAAASUVORK5CYII=\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"tags": []
}
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "MsuJmFDUn0Xa",
"colab_type": "code",
"outputId": "88138147-842c-412d-832f-0220fd3d22d0",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 285
}
},
"source": [
"y_class = np.argmax(y, axis=-1)\n",
"plt.imshow(y_class[1,:,:])\n"
],
"execution_count": 16,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"<matplotlib.image.AxesImage at 0x7f6c93c88278>"
]
},
"metadata": {
"tags": []
},
"execution_count": 16
},
{
"output_type": "display_data",
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAPsAAAD7CAYAAACscuKmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"tags": []
}
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "In8xqbCEn528",
"colab_type": "code",
"outputId": "d97d106c-a489-4850-9765-bfc7849818a1",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 34
}
},
"source": [
"dice_loss(sigmoid(y), random_image_masks)"
],
"execution_count": 17,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"DeviceArray(0.99164575, dtype=float32)"
]
},
"metadata": {
"tags": []
},
"execution_count": 17
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "tv_A3LqHrrVI",
"colab_type": "code",
"colab": {}
},
"source": [
"batch = next(iter(dataloaders['train']))\n",
"#print(loss(params,batch))"
],
"execution_count": 0,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"id": "s0Y6eyjGtHPB",
"colab_type": "code",
"outputId": "c3e14e16-ef73-4ad1-92b3-14eb8d3904d6",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 34
}
},
"source": [
"loss(params,batch)"
],
"execution_count": 19,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"DeviceArray(178.7824, dtype=float32)"
]
},
"metadata": {
"tags": []
},
"execution_count": 19
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "yReBEENutlac",
"colab_type": "code",
"outputId": "39f1436a-509c-4a4d-8064-3768455cae1f",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 731
}
},
"source": [
"from collections import defaultdict\n",
"\n",
"num_epochs = 10\n",
"\n",
"random_image = inputs\n",
"key = random.PRNGKey(1)\n",
"_, params0 = init_params(key, \n",
" input_shape=random_image.shape)\n",
"\n",
"opt_state = opt_init(params0)\n",
"itercount = itertools.count()\n",
"\n",
"metrics = defaultdict(float)\n",
"metrics['train'] = defaultdict(float)\n",
"metrics['val'] = defaultdict(float)\n",
"metrics['train']['bce'] = []\n",
"metrics['train']['dice'] = []\n",
"metrics['train']['loss'] = []\n",
"metrics['val']['bce'] = []\n",
"metrics['val']['dice'] = []\n",
"metrics['val']['loss'] = []\n",
"\n",
"print(\"\\nStarting training...\")\n",
"for epoch in range(num_epochs):\n",
" print('Epoch {}/{}'.format(epoch, num_epochs - 1))\n",
" print('-' * 10)\n",
" \n",
" start_time = time.time()\n",
" for phase in ['train', 'val']:\n",
" epoch_samples = 0\n",
"\n",
" for inputs, labels in dataloaders[phase]:\n",
" #inputs = inputs.numpy().transpose((0,2,3,1))\n",
" #labels = labels.numpy().transpose((0,2,3,1))\n",
" \n",
" if phase=='train':\n",
" opt_state = update(next(itercount), \n",
" opt_state, \n",
" (inputs, labels) )\n",
" \n",
" params = get_params(opt_state)\n",
" outputs = predict(params, inputs )\n",
" bce,dice = makeReport(outputs, labels)\n",
" lossVal = loss(params, (inputs, labels))\n",
" metrics[phase]['bce'] = metrics[phase]['bce'] + [bce]\n",
" metrics[phase]['dice'] = metrics[phase]['dice'] + [dice]\n",
" metrics[phase]['loss'] = metrics[phase]['loss'] + [lossVal]\n",
"\n",
"\n",
" print(phase,'--',' bce :',bce, ' dice :',dice, ' loss :',lossVal)\n",
" epoch_time = time.time() - start_time\n"
],
"execution_count": 21,
"outputs": [
{
"output_type": "stream",
"text": [
"\n",
"Starting training...\n",
"Epoch 0/9\n",
"----------\n",
"train -- bce : 426937.25 dice : 0.99138594 loss : 162.02266\n",
"val -- bce : 426900.78 dice : 0.99175775 loss : 156.6451\n",
"Epoch 1/9\n",
"----------\n",
"train -- bce : 427174.97 dice : 0.99121785 loss : 156.95479\n",
"val -- bce : 427212.38 dice : 0.9911194 loss : 157.506\n",
"Epoch 2/9\n",
"----------\n",
"train -- bce : 427493.75 dice : 0.9902591 loss : 166.08307\n",
"val -- bce : 427358.75 dice : 0.9909472 loss : 152.44684\n",
"Epoch 3/9\n",
"----------\n",
"train -- bce : 427504.75 dice : 0.9907694 loss : 145.08008\n",
"val -- bce : 427508.94 dice : 0.9909479 loss : 147.8187\n",
"Epoch 4/9\n",
"----------\n",
"train -- bce : 427735.8 dice : 0.98985285 loss : 156.48497\n",
"val -- bce : 427636.06 dice : 0.990623 loss : 141.8883\n",
"Epoch 5/9\n",
"----------\n",
"train -- bce : 427768.88 dice : 0.9903689 loss : 135.5426\n",
"val -- bce : 427856.16 dice : 0.99005127 loss : 144.42183\n",
"Epoch 6/9\n",
"----------\n",
"train -- bce : 427990.53 dice : 0.99015534 loss : 129.97003\n",
"val -- bce : 427961.22 dice : 0.9903161 loss : 127.25979\n",
"Epoch 7/9\n",
"----------\n",
"train -- bce : 428401.0 dice : 0.98910147 loss : 127.563095\n",
"val -- bce : 428380.97 dice : 0.9895691 loss : 123.90769\n",
"Epoch 8/9\n",
"----------\n",
"train -- bce : 429082.22 dice : 0.9887206 loss : 109.136345\n",
"val -- bce : 429211.8 dice : 0.9885659 loss : 113.34527\n",
"Epoch 9/9\n",
"----------\n",
"train -- bce : 430841.78 dice : 0.9871861 loss : 98.39589\n",
"val -- bce : 430735.6 dice : 0.9876917 loss : 97.08817\n"
],
"name": "stdout"
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "kYEKfRdvurXp",
"colab_type": "code",
"outputId": "5633fabd-b938-4f94-c2ca-6845db5a897f",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 536
}
},
"source": [
"plt.imshow(random_image[1,:,:,:])\n",
"\n",
"y = predict(params, random_image )\n",
"y_class = np.argmax(y, axis=-1)\n",
"\n",
"\n",
"plt.figure()\n",
"plt.imshow(y_class[1,:,:])"
],
"execution_count": 22,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"<matplotlib.image.AxesImage at 0x7f6c82923748>"
]
},
"metadata": {
"tags": []
},
"execution_count": 22
},
{
"output_type": "display_data",
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAPsAAAD7CAYAAACscuKmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nMOxSIwy71AjDLjXCsEuN+D9Tdel84x5dWAAAAABJRU5ErkJggg==\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"tags": []
}
},
{
"output_type": "display_data",
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAPsAAAD7CAYAAACscuKmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJ6QncLET0hO42AnpCVzshPSE/wdj/vD8V210JwAAAABJRU5ErkJggg==\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"tags": []
}
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "DiaKkkWpUmf4",
"colab_type": "code",
"colab": {}
},
"source": [
"# def reverse_transform(inp):\n",
"# #inp = inp.transpose((2, 1, 0))\n",
"# mean = onp.array([0.485, 0.456, 0.406])\n",
"# std = onp.array([0.229, 0.224, 0.225])\n",
"# inp = std * inp + mean\n",
"# inp = onp.clip(inp, 0, 1)\n",
"# inp = (inp * 255).astype(np.uint8)"
],
"execution_count": 0,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"id": "gFJGkC_ex3QG",
"colab_type": "code",
"outputId": "a919b256-5854-4735-89fa-31d71e302f5c",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 1000
}
},
"source": [
"\n",
"\n",
"params = get_params(opt_state)\n",
"\n",
"y = predict(params, inputs )\n",
"y_class = np.argmax(y, axis=-1)\n",
"\n",
"y_class2 = onp.zeros((y_class.shape[0],\n",
" y_class.shape[1],\n",
" y_class.shape[2],\n",
" 6))\n",
"\n",
"for i in range(6):\n",
" y_class2[:,:,:,i] = y_class==i\n",
"\n",
"#input_images_rgb = [reverse_transform(x) for x in inputs]\n",
"\n",
"# Map each channel (i.e. class) to each color\n",
"target_masks_rgb = [helper.masks_to_colorimg(x) for x in labels.transpose((0,3,1,2))]\n",
"pred_masks_rgb = [helper.masks_to_colorimg(x) for x in y_class2.transpose((0,3,1,2))]\n",
"\n",
"\n",
"# Left: Input image, Right: Target mask (Ground-truth)\n",
"#helper.plot_side_by_side([input_images_rgb, pred_masks_rgb, target_masks_rgb])\n",
"helper.plot_side_by_side([ pred_masks_rgb, target_masks_rgb])"
],
"execution_count": 24,
"outputs": [
{
"output_type": "display_data",
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAcsAABVdCAYAAAD/rP7EAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAQAo+P+yH6DqDhq45wAAAABJRU5ErkJggg==\n",
"text/plain": [
"<Figure size 576x7200 with 50 Axes>"
]
},
"metadata": {
"tags": []
}
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "GhwuYAncx9Fk",
"colab_type": "code",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 34
},
"outputId": "f4f2b3a4-9084-475c-bc2a-7dc4cbe16926"
},
"source": [
"ls"
],
"execution_count": 25,
"outputs": [
{
"output_type": "stream",
"text": [
"\u001b[0m\u001b[01;34mneural-tangents\u001b[0m/ \u001b[01;34mpytorch-unet\u001b[0m/ \u001b[01;34msample_data\u001b[0m/\n"
],
"name": "stdout"
}
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "BOf6DSz_opyS",
"colab_type": "text"
},
"source": [
"# Applying Kernel"
]
},
{
"cell_type": "code",
"metadata": {
"id": "LpocmT_rW3tv",
"colab_type": "code",
"colab": {}
},
"source": [
"from neural_tangents.utils.kernel import Marginalisation as M\n",
"\n",
"@jit\n",
"def mykernel(x,y):\n",
" return kernel_fn(x,y,'nngp')\n",
"\n",
"# kernel = kernel_fn(random_image[:10],random_image[:10], \n",
"# 'nngp', marginalization={'marginal': M.OVER_ALL, 'cross': M.OVER_PIXELS})\n"
],
"execution_count": 0,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"id": "BTxadc9wT1sv",
"colab_type": "code",
"outputId": "ddd7f5c5-ee00-42c3-a3e2-2374ca976b36",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 1000
}
},
"source": [
"kernel = mykernel(random_image[:10],random_image[:10])"
],
"execution_count": 27,
"outputs": [
{
"output_type": "stream",
"text": [
"/usr/local/lib/python3.6/dist-packages/jax/lax/lax.py:4571: UserWarning: Explicitly requested dtype <class 'jax.numpy.lax_numpy.float64'> requested in astype is not available, and will be truncated to dtype float32. To enable more dtypes, set the jax_enable_x64 configuration option or the JAX_ENABLE_X64 shell environment variable. See https://github.com/google/jax#current-gotchas for more.\n",
" warnings.warn(msg.format(dtype, fun_name , truncated_dtype))\n"
],
"name": "stderr"
},
{
"output_type": "error",
"ename": "RuntimeError",
"evalue": "ignored",
"traceback": [
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
"\u001b[0;31mRuntimeError\u001b[0m Traceback (most recent call last)",
"\u001b[0;32m<ipython-input-27-053c34ab30f7>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mkernel\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmykernel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mrandom_image\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;36m10\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mrandom_image\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;36m10\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
"\u001b[0;32m/usr/local/lib/python3.6/dist-packages/jax/api.py\u001b[0m in \u001b[0;36mf_jitted\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m 147\u001b[0m \u001b[0m_check_args\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0margs_flat\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 148\u001b[0m \u001b[0mflat_fun\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mout_tree\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mflatten_fun\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mf\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0min_tree\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 149\u001b[0;31m \u001b[0mout\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mxla\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mxla_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mflat_fun\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs_flat\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdevice\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mdevice\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbackend\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mbackend\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 150\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0mtree_unflatten\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mout_tree\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mout\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 151\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;32m/usr/local/lib/python3.6/dist-packages/jax/core.py\u001b[0m in \u001b[0;36mcall_bind\u001b[0;34m(primitive, f, *args, **params)\u001b[0m\n\u001b[1;32m 600\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0mtop_trace\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 601\u001b[0m \u001b[0;32mwith\u001b[0m \u001b[0mnew_sublevel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 602\u001b[0;31m \u001b[0mouts\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mprimitive\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mimpl\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mf\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mparams\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 603\u001b[0m \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 604\u001b[0m \u001b[0mtracers\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmap\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtop_trace\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfull_raise\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0margs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;32m/usr/local/lib/python3.6/dist-packages/jax/interpreters/xla.py\u001b[0m in \u001b[0;36m_xla_call_impl\u001b[0;34m(fun, *args, **params)\u001b[0m\n\u001b[1;32m 440\u001b[0m \u001b[0mdevice\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mparams\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'device'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 441\u001b[0m \u001b[0mbackend\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mparams\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'backend'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 442\u001b[0;31m \u001b[0mcompiled_fun\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_xla_callable\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfun\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdevice\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbackend\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0mmap\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0marg_spec\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0margs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 443\u001b[0m \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 444\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0mcompiled_fun\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;32m/usr/local/lib/python3.6/dist-packages/jax/linear_util.py\u001b[0m in \u001b[0;36mmemoized_fun\u001b[0;34m(fun, *args)\u001b[0m\n\u001b[1;32m 221\u001b[0m \u001b[0mfun\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpopulate_stores\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstores\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 222\u001b[0m \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 223\u001b[0;31m \u001b[0mans\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcall\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfun\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 224\u001b[0m \u001b[0mcache\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mans\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfun\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstores\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 225\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0mans\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;32m/usr/local/lib/python3.6/dist-packages/jax/interpreters/xla.py\u001b[0m in \u001b[0;36m_xla_callable\u001b[0;34m(fun, device, backend, *arg_specs)\u001b[0m\n\u001b[1;32m 497\u001b[0m options = xb.get_compile_options(\n\u001b[1;32m 498\u001b[0m num_replicas=nreps, device_assignment=(device.id,) if device else None)\n\u001b[0;32m--> 499\u001b[0;31m \u001b[0mcompiled\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mbuilt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mCompile\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mcompile_options\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0moptions\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbackend\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mxb\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_backend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbackend\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 500\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 501\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0mnreps\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;32m/usr/local/lib/python3.6/dist-packages/jaxlib/xla_client.py\u001b[0m in \u001b[0;36mCompile\u001b[0;34m(self, argument_shapes, compile_options, backend)\u001b[0m\n\u001b[1;32m 607\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0margument_shapes\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 608\u001b[0m \u001b[0mcompile_options\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0margument_layouts\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0margument_shapes\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 609\u001b[0;31m \u001b[0;32mreturn\u001b[0m \u001b[0mbackend\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcompile\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcomputation\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcompile_options\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 610\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 611\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mGetProgramShape\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;32m/usr/local/lib/python3.6/dist-packages/jaxlib/tpu_client.py\u001b[0m in \u001b[0;36mcompile\u001b[0;34m(self, c_computation, compile_options)\u001b[0m\n\u001b[1;32m 103\u001b[0m \u001b[0mcompile_options\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0margument_layouts\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 104\u001b[0m \u001b[0moptions\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclient\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 105\u001b[0;31m compile_options.device_assignment)\n\u001b[0m\u001b[1;32m 106\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 107\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mget_default_device_assignment\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnum_replicas\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;31mRuntimeError\u001b[0m: Resource exhausted: Ran out of memory in memory space hbm. Used 25.99G of 7.48G hbm. Exceeded hbm capacity by 18.50G.\n\nTotal hbm usage >= 26.50G:\n reserved 529.00M \n program 25.99G \n arguments unknown size \n\nOutput size unknown.\n\nProgram hbm requirement 25.99G:\n reserved 4.0K\n global 36.0K\n HLO temp 25.99G (74.4% utilization: Unpadded (19.34G) Padded (25.98G), 0.0% fragmentation (10.31M))\n\n Largest program allocations in hbm:\n\n 1. Size: 12.50G\n Operator: op_type=\"conv_general_dilated\"\n Shape: f32[409600,1,64,64]{0,1,3,2:T(2,128)}\n Unpadded size: 6.25G\n Extra memory due to padding: 6.25G (2.0x expansion)\n XLA label: %convolution.5785 = f32[409600,1,64,64]{0,1,3,2:T(2,128)} convolution(bf16[409600,1,64,64]{0,1,3,2:T(4,128)(2,1)} %reshape.1452, bf16[3,3,1,1]{3,2,1,0:T(4,128)(2,1)} %constant.2723), window={size=3x3 pad=1_1x1_1}, dim_labels=bf01_01io->bf01, metadata={op_t...\n Allocation type: HLO temp\n ==========================\n\n 2. Size: 6.25G\n Shape: f32[409600,1,64,64]{0,3,2,1}\n Unpadded size: 6.25G\n XLA label: %copy.1516 = f32[409600,1,64,64]{0,3,2,1} copy(f32[409600,1,64,64]{0,1,3,2:T(2,128)} %convolution.4620)\n Allocation type: HLO temp\n ==========================\n\n 3. Size: 6.25G\n Shape: f32[409600,1,64,64]{0,3,2,1}\n Unpadded size: 6.25G\n XLA label: %copy.1540 = f32[409600,1,64,64]{0,3,2,1} copy(f32[409600,1,64,64]{0,1,3,2:T(2,128)} %convolution.5785)\n Allocation type: HLO temp\n ==========================\n\n 4. Size: 640.00M\n Operator: op_type=\"reshape\"\n Shape: bf16[10,64,64,64,64]{2,1,0,4,3:T(8,128)(2,1)}\n Unpadded size: 320.00M\n Extra memory due to padding: 320.00M (2.0x expansion)\n XLA label: %reshape.753 = bf16[10,64,64,64,64]{2,1,0,4,3:T(8,128)(2,1)} reshape(bf16[40960,1,64,64]{0,1,3,2:T(4,128)(2,1)} %fusion.420), metadata={op_type=\"reshape\"}\n Allocation type: HLO temp\n ==========================\n\n 5. Size: 160.00M\n Operator: op_type=\"transpose\"\n Shape: bf16[10,64,64,32,32]{2,1,0,4,3:T(8,128)(2,1)}\n Unpadded size: 80.00M\n Extra memory due to padding: 80.00M (2.0x expansion)\n XLA label: %copy.1153 = bf16[10,64,64,32,32]{2,1,0,4,3:T(8,128)(2,1)} copy(bf16[10,64,64,32,32]{2,1,4,3,0:T(8,128)(2,1)} %bitcast.127), metadata={op_type=\"transpose\"}\n Allocation type: HLO temp\n ==========================\n\n 6. Size: 100.00M\n Shape: f32[409600,64]{0,1:T(8,128)}\n Unpadded size: 100.00M\n XLA label: %reshape.1326 = f32[409600,64]{0,1:T(8,128)} reshape(f32[10,10,64,64,64]{3,2,1,0,4:T(8,128)} %broadcast.1682.remat)\n Allocation type: HLO temp\n ==========================\n\n 7. Size: 100.00M\n Shape: f32[409600,64]{0,1:T(8,128)}\n Unpadded size: 100.00M\n XLA label: %reshape.1332 = f32[409600,64]{0,1:T(8,128)} reshape(f32[10,10,64,64,64]{3,2,1,0,4:T(8,128)} %broadcast.2053)\n Allocation type: HLO temp\n ==========================\n\n 8. Size: 256.0K\n Operator: op_type=\"slice\"\n Shape: f32[10,4096]{1,0:T(8,128)}\n Unpadded size: 160.0K\n Extra memory due to padding: 96.0K (1.6x expansion)\n XLA label: %fusion.671 = f32[10,4096]{1,0:T(8,128)} fusion(f32[10,4096,4096]{2,1,0:T(8,128)} %reshape.4392, pred[4096,4096]{1,0:T(8,128)E(32)} %fusion.1076.remat), kind=kLoop, calls=%fused_computation.591, metadata={op_type=\"slice\"}\n Allocation type: HLO temp\n ==========================\n\n 9. Size: 9.0K\n Shape: bf16[3,3,1,1]{3,2,1,0:T(4,128)(2,1)}\n Unpadded size: 18B\n Extra memory due to padding: 9.0K (512.0x expansion)\n XLA label: constant literal\n Allocation type: global\n ==========================\n\n 10. Size: 4.0K\n XLA label: profiler\n Allocation type: reserved\n ==========================\n\n 11. Size: 4.0K\n Shape: bf16[2,2,1,1]{3,2,1,0:T(4,128)(2,1)}\n Unpadded size: 8B\n Extra memory due to padding: 4.0K (512.0x expansion)\n XLA label: constant literal\n Allocation type: global\n ==========================\n\n 12. Size: 4.0K\n Shape: u32[8,128]{1,0}\n Unpadded size: 4.0K\n XLA label: constant literal\n Allocation type: global\n ==========================\n\n"
]
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "Yq_3VENCpC_q",
"colab_type": "code",
"colab": {}
},
"source": [
""
],
"execution_count": 0,
"outputs": []
}
]
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment