forVerify.ipynb

{
 "metadata": {
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.10"
  },
  "orig_nbformat": 4,
  "kernelspec": {
   "name": "python3",
   "display_name": "Python 3.7.10 64-bit ('mmlab-cu110': conda)"
  },
  "interpreter": {
   "hash": "b28049a6882f1b8a7e915a2aa52e1234edf04cda286b30c820db76a58a4ad7bf"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2,
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "source": [
    "import cv2\n",
    "import os\n",
    "import os.path as osp\n",
    "import numpy as np\n",
    "import torch\n",
    "from torch import distributions"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "source": [
    "a = torch.zeros(size=(10, 10))\n",
    "time_sampler = distributions.Categorical(torch.tensor([0.5, 0.5]))\n",
    "time_sampler.sample([a.shape[0]])"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "tensor([1, 1, 1, 0, 0, 0, 1, 1, 1, 0])"
      ]
     },
     "metadata": {},
     "execution_count": 31
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "source": [
    "alpha = 1\n",
    "sampler = distributions.Beta(alpha, alpha)\n",
    "sampler.sample((13,2))"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "tensor([[0.2248, 0.0675],\n",
       "        [0.5345, 0.0096],\n",
       "        [0.7611, 0.3541],\n",
       "        [0.1827, 0.0382],\n",
       "        [0.3131, 0.9250],\n",
       "        [0.5426, 0.8808],\n",
       "        [0.5400, 0.0893],\n",
       "        [0.1646, 0.5771],\n",
       "        [0.5305, 0.0797],\n",
       "        [0.6338, 0.7853],\n",
       "        [0.7720, 0.9447],\n",
       "        [0.2451, 0.1928],\n",
       "        [0.1335, 0.9116]])"
      ]
     },
     "metadata": {},
     "execution_count": 8
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "source": [
    "lr = 0.00006\n",
    "warmup_iters = 1500\n",
    "warmup_ratio = 1e-6\n",
    "power = 1.0\n",
    "min_lr = 0.0\n",
    "start_lr = lr * warmup_ratio\n",
    "step = (lr-start_lr)/warmup_iters\n",
    "start_lr, step, start_lr + step*49"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "(6e-11, 3.999996e-08, 1.96005804e-06)"
      ]
     },
     "metadata": {},
     "execution_count": 5
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "source": [
    "path = r'show_dir/34.png'\n",
    "img = cv2.imread(path)\n",
    "img.dtype, img.shape, np.unique(img)"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "(dtype('uint8'), (1024, 1024, 3), array([  0, 255], dtype=uint8))"
      ]
     },
     "metadata": {},
     "execution_count": 2
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "source": [
    "a = np.array([98.73, 0.85, 0.39, 0.04])\n",
    "b = 1 / a\n",
    "c = b/b.sum()\n",
    "a, b, c, c.sum()"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "(array([9.873e+01, 8.500e-01, 3.900e-01, 4.000e-02]),\n",
       " array([1.01286336e-02, 1.17647059e+00, 2.56410256e+00, 2.50000000e+01]),\n",
       " array([3.52291701e-04, 4.09197173e-02, 8.91839992e-02, 8.69543992e-01]),\n",
       " 1.0)"
      ]
     },
     "metadata": {},
     "execution_count": 10
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "source": [
    "a = dict(asd=1, we=2)"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "source": [
    "a.get('asd', []), a.get('sdf', [])"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "(1, [])"
      ]
     },
     "metadata": {},
     "execution_count": 6
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "source": [
    "path = r'/home/csl/code/rsaicp_CD/data/ade/ADEChallengeData2016/annotations/training/ADE_train_00000011.png'\n",
    "a = cv2.imread(path, -1)"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "source": [
    "a.shape"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "(247, 356)"
      ]
     },
     "metadata": {},
     "execution_count": 6
    }
   ],
   "metadata": {}
  }
 ]
}