From 4208b85a3d1693bb787d471e4c7a08ad1e056e24 Mon Sep 17 00:00:00 2001
From: tanayag <tanay_agrawal@hotmail.com>
Date: Tue, 23 Oct 2018 14:21:17 +0530
Subject: [PATCH] add

---
 test_one.ipynb | 720 +++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 720 insertions(+)
 create mode 100644 test_one.ipynb

diff --git a/test_one.ipynb b/test_one.ipynb
new file mode 100644
index 0000000..d221ba4
--- /dev/null
+++ b/test_one.ipynb
@@ -0,0 +1,720 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import numpy as np\n",
+    "import matplotlib.pyplot as plt\n",
+    "import torch\n",
+    "import torch.nn as nn\n",
+    "import torch.optim as optim\n",
+    "from torch.autograd import Variable"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# generating dataset for discriminator\n",
+    "def get_dataset():\n",
+    "    return torch.Tensor(np.random.normal(4,1.25,(1,100)))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# generating noise for generator\n",
+    "def make_noise():\n",
+    "    return torch.Tensor(np.random.uniform(0,0,(1,50)))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "class generator(nn.Module):\n",
+    "    \n",
+    "    def __init__(self,inp,out):\n",
+    "        super(generator,self).__init__()\n",
+    "        self.net = nn.Sequential(nn.Linear(inp,300),\n",
+    "                                 nn.ReLU(inplace=True),\n",
+    "                                nn.Linear(300,300),\n",
+    "                                 nn.ReLU(inplace=True),\n",
+    "                                nn.Linear(300,out)\n",
+    "                               )\n",
+    "        \n",
+    "    def forward(self,x):\n",
+    "        x = self.net(x)\n",
+    "        return x\n",
+    "        "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "class discriminator(nn.Module):\n",
+    "    \n",
+    "    def __init__(self,inp,out):\n",
+    "        super(discriminator,self).__init__()\n",
+    "        self.net = nn.Sequential(nn.Linear(inp,300),\n",
+    "                                 nn.ReLU(inplace=True),\n",
+    "                                 nn.Linear(300,300),\n",
+    "                                 nn.ReLU(inplace=True),\n",
+    "                                 nn.Linear(300,out),\n",
+    "                                 nn.Sigmoid()\n",
+    "                                )\n",
+    "        \n",
+    "    def forward(self,x):\n",
+    "        x = self.net(x)\n",
+    "        return x"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def stats(array):\n",
+    "    array = array.detach().numpy()\n",
+    "    return [np.mean(array),np.std(array)]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "gen = generator(50,100)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "generator(\n",
+       "  (net): Sequential(\n",
+       "    (0): Linear(in_features=50, out_features=300, bias=True)\n",
+       "    (1): ReLU(inplace)\n",
+       "    (2): Linear(in_features=300, out_features=300, bias=True)\n",
+       "    (3): ReLU(inplace)\n",
+       "    (4): Linear(in_features=300, out_features=100, bias=True)\n",
+       "  )\n",
+       ")"
+      ]
+     },
+     "execution_count": 8,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "gen"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 9,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "x = make_noise()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "x1 = gen(x)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 11,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "[0.0018841184, 0.036860824]"
+      ]
+     },
+     "execution_count": 11,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "stats(x1)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 12,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "discrim = discriminator(100,1)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 13,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "discriminator(\n",
+       "  (net): Sequential(\n",
+       "    (0): Linear(in_features=100, out_features=300, bias=True)\n",
+       "    (1): ReLU(inplace)\n",
+       "    (2): Linear(in_features=300, out_features=300, bias=True)\n",
+       "    (3): ReLU(inplace)\n",
+       "    (4): Linear(in_features=300, out_features=1, bias=True)\n",
+       "    (5): Sigmoid()\n",
+       "  )\n",
+       ")"
+      ]
+     },
+     "execution_count": 13,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "discrim"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 14,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "x2 = discrim(x1)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 15,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "tensor([[ 0.5205]])"
+      ]
+     },
+     "execution_count": 15,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "x2"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 16,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "epochs = 500"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 17,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "\n",
+    "d_step = 10\n",
+    "g_step = 8\n",
+    "criteriond1 = nn.BCELoss()\n",
+    "optimizerd1 = optim.SGD(discrim.parameters(), lr=0.001, momentum=0.9)\n",
+    "\n",
+    "criteriond2 = nn.BCELoss()\n",
+    "optimizerd2 = optim.SGD(gen.parameters(), lr=0.001, momentum=0.9)\n",
+    "\n",
+    "printing_steps = 20"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 18,
+   "metadata": {
+    "scrolled": true
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "('Epoch:', 0)\n",
+      "[0.0018841184, 0.036860824]\n",
+      "[0.001885428, 0.036860917]\n",
+      "[0.0018879148, 0.03686111]\n",
+      "[0.0018914625, 0.036861386]\n",
+      "[0.0018959647, 0.036861762]\n",
+      "[0.0019013253, 0.036862247]\n",
+      "[0.0019074584, 0.03686284]\n",
+      "[0.0019142872, 0.036863554]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 20)\n",
+      "[0.007975619, 0.038056422]\n",
+      "[0.008030179, 0.03808928]\n",
+      "[0.008086704, 0.03812305]\n",
+      "[0.008145271, 0.03815762]\n",
+      "[0.008205742, 0.038193222]\n",
+      "[0.008268582, 0.038229816]\n",
+      "[0.008333553, 0.038267855]\n",
+      "[0.008401263, 0.038307115]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 40)\n",
+      "[0.025739962, 0.07039434]\n",
+      "[0.025967699, 0.07082409]\n",
+      "[0.026195178, 0.071255475]\n",
+      "[0.026425453, 0.071689904]\n",
+      "[0.026660535, 0.072128154]\n",
+      "[0.026900362, 0.07257016]\n",
+      "[0.027144287, 0.07301512]\n",
+      "[0.027392678, 0.07346259]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 60)\n",
+      "[0.08450457, 0.14448352]\n",
+      "[0.08509155, 0.14511523]\n",
+      "[0.08568327, 0.14574906]\n",
+      "[0.08627997, 0.14638585]\n",
+      "[0.0868816, 0.14702617]\n",
+      "[0.08748844, 0.1476706]\n",
+      "[0.08809981, 0.1483197]\n",
+      "[0.08871749, 0.1489742]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 80)\n",
+      "[0.28840774, 0.27236685]\n",
+      "[0.29072142, 0.2731846]\n",
+      "[0.29304278, 0.27398548]\n",
+      "[0.2953758, 0.27477252]\n",
+      "[0.2977213, 0.27554947]\n",
+      "[0.30008143, 0.27632013]\n",
+      "[0.30245793, 0.27708793]\n",
+      "[0.30485335, 0.27785522]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 100)\n",
+      "[1.1765242, 0.47627425]\n",
+      "[1.1888257, 0.48015204]\n",
+      "[1.2012541, 0.48397332]\n",
+      "[1.2138283, 0.4878198]\n",
+      "[1.2265671, 0.49175856]\n",
+      "[1.2395148, 0.49582666]\n",
+      "[1.2526832, 0.5000789]\n",
+      "[1.2660835, 0.50456303]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 120)\n",
+      "[3.7521708, 1.1674445]\n",
+      "[3.7650826, 1.1632067]\n",
+      "[3.7773826, 1.1575049]\n",
+      "[3.7894604, 1.1513976]\n",
+      "[3.8016295, 1.1458958]\n",
+      "[3.8140812, 1.1418389]\n",
+      "[3.8267963, 1.1399107]\n",
+      "[3.839567, 1.14043]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 140)\n",
+      "[5.4304433, 0.95063484]\n",
+      "[5.4390655, 0.9454818]\n",
+      "[5.4475994, 0.9381844]\n",
+      "[5.45611, 0.9292164]\n",
+      "[5.4646544, 0.9190324]\n",
+      "[5.473289, 0.90804404]\n",
+      "[5.4820514, 0.8966615]\n",
+      "[5.490971, 0.88520867]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 160)\n",
+      "[6.554171, 1.0230845]\n",
+      "[6.5443854, 1.0079885]\n",
+      "[6.5336504, 0.9934966]\n",
+      "[6.5223026, 0.98004055]\n",
+      "[6.510661, 0.968119]\n",
+      "[6.4991875, 0.95708686]\n",
+      "[6.488184, 0.9479753]\n",
+      "[6.4779205, 0.94164]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 180)\n",
+      "[6.6743093, 1.1352086]\n",
+      "[6.674415, 1.1340271]\n",
+      "[6.6746874, 1.1323054]\n",
+      "[6.6751537, 1.1302996]\n",
+      "[6.675848, 1.1281751]\n",
+      "[6.6767945, 1.1260948]\n",
+      "[6.678016, 1.1241832]\n",
+      "[6.67953, 1.1225513]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 200)\n",
+      "[6.7470884, 1.0527247]\n",
+      "[6.7447414, 1.0537938]\n",
+      "[6.742381, 1.0543174]\n",
+      "[6.740056, 1.0544192]\n",
+      "[6.73781, 1.0542215]\n",
+      "[6.735685, 1.0538297]\n",
+      "[6.733716, 1.0533464]\n",
+      "[6.731936, 1.0528711]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 220)\n",
+      "[5.3740454, 1.1091442]\n",
+      "[5.3554115, 1.1135103]\n",
+      "[5.336738, 1.1160842]\n",
+      "[5.318118, 1.1171281]\n",
+      "[5.299648, 1.116884]\n",
+      "[5.281429, 1.1155953]\n",
+      "[5.2635565, 1.1134976]\n",
+      "[5.246128, 1.1108178]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 240)\n",
+      "[3.563117, 0.7772749]\n",
+      "[3.5572462, 0.7718341]\n",
+      "[3.5509315, 0.76590574]\n",
+      "[3.5443897, 0.7597089]\n",
+      "[3.5378225, 0.75346625]\n",
+      "[3.5314343, 0.74733096]\n",
+      "[3.5253472, 0.74146765]\n",
+      "[3.5196652, 0.73601395]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 260)\n",
+      "[3.762761, 0.44140473]\n",
+      "[3.766048, 0.44176507]\n",
+      "[3.769457, 0.44208202]\n",
+      "[3.772982, 0.442378]\n",
+      "[3.7766001, 0.44267818]\n",
+      "[3.7803097, 0.44300264]\n",
+      "[3.7840986, 0.44338825]\n",
+      "[3.7879653, 0.44384944]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 280)\n",
+      "[4.547168, 0.5041005]\n",
+      "[4.5512414, 0.50436765]\n",
+      "[4.5553017, 0.50460684]\n",
+      "[4.5593777, 0.5048857]\n",
+      "[4.5634933, 0.50526166]\n",
+      "[4.5676703, 0.505786]\n",
+      "[4.571926, 0.5064988]\n",
+      "[4.5762706, 0.5074317]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 300)\n",
+      "[4.714465, 0.60834765]\n",
+      "[4.7135544, 0.6082357]\n",
+      "[4.712129, 0.607926]\n",
+      "[4.71026, 0.6074627]\n",
+      "[4.7080126, 0.60689396]\n",
+      "[4.705456, 0.6062651]\n",
+      "[4.702648, 0.6056212]\n",
+      "[4.699642, 0.6050036]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 320)\n",
+      "[4.1752644, 0.49929926]\n",
+      "[4.1692877, 0.5012012]\n",
+      "[4.1635256, 0.5029179]\n",
+      "[4.157988, 0.5044922]\n",
+      "[4.1526837, 0.5059638]\n",
+      "[4.1476197, 0.50736845]\n",
+      "[4.1428027, 0.5087379]\n",
+      "[4.1382384, 0.51009995]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 340)\n",
+      "[3.9004126, 0.49423027]\n",
+      "[3.899606, 0.49115932]\n",
+      "[3.8988214, 0.48803943]\n",
+      "[3.8981094, 0.48495686]\n",
+      "[3.8975136, 0.4819867]\n",
+      "[3.897069, 0.47919214]\n",
+      "[3.8968034, 0.47662398]\n",
+      "[3.8967361, 0.47432172]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 360)\n",
+      "[3.845068, 0.45620793]\n",
+      "[3.8458064, 0.45662042]\n",
+      "[3.846777, 0.45696673]\n",
+      "[3.847946, 0.45725322]\n",
+      "[3.8492954, 0.4574869]\n",
+      "[3.850814, 0.45767534]\n",
+      "[3.8524847, 0.4578276]\n",
+      "[3.8542979, 0.45795205]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 380)\n",
+      "[4.4157295, 0.50118345]\n",
+      "[4.4171925, 0.5007066]\n",
+      "[4.418768, 0.5001312]\n",
+      "[4.420454, 0.49948642]\n",
+      "[4.4222555, 0.49879527]\n",
+      "[4.4241753, 0.49808162]\n",
+      "[4.4262133, 0.49736917]\n",
+      "[4.4283686, 0.49667883]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 400)\n",
+      "[4.5400233, 0.5564293]\n",
+      "[4.540456, 0.5564975]\n",
+      "[4.5409265, 0.5565482]\n",
+      "[4.5414352, 0.5565863]\n",
+      "[4.541979, 0.55661666]\n",
+      "[4.542558, 0.55664325]\n",
+      "[4.5431724, 0.55666983]\n",
+      "[4.54382, 0.55670017]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 420)\n",
+      "[4.692982, 0.5231943]\n",
+      "[4.692985, 0.5231961]\n",
+      "[4.6929874, 0.5231977]\n",
+      "[4.6929893, 0.5231991]\n",
+      "[4.6929913, 0.5232004]\n",
+      "[4.6929936, 0.52320147]\n",
+      "[4.692995, 0.5232025]\n",
+      "[4.6929965, 0.5232035]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 440)\n",
+      "[5.0075383, 0.43208402]\n",
+      "[5.007661, 0.43213037]\n",
+      "[5.007836, 0.4321994]\n",
+      "[5.008058, 0.43229046]\n",
+      "[5.0083237, 0.43240368]\n",
+      "[5.0086274, 0.43253857]\n",
+      "[5.008966, 0.43269572]\n",
+      "[5.0093365, 0.43287548]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 460)\n",
+      "[5.029313, 0.44687128]\n",
+      "[5.0294237, 0.44697717]\n",
+      "[5.0295234, 0.4470743]\n",
+      "[5.0296135, 0.4471632]\n",
+      "[5.0296936, 0.4472443]\n",
+      "[5.029766, 0.4473183]\n",
+      "[5.0298314, 0.44738564]\n",
+      "[5.02989, 0.44744688]\n",
+      "\n",
+      "\n",
+      "\n",
+      "('Epoch:', 480)\n",
+      "[5.0598216, 0.46177983]\n",
+      "[5.0598273, 0.461782]\n",
+      "[5.059832, 0.46178398]\n",
+      "[5.0598364, 0.46178573]\n",
+      "[5.059841, 0.4617873]\n",
+      "[5.0598445, 0.4617887]\n",
+      "[5.059849, 0.46178997]\n",
+      "[5.059851, 0.4617911]\n",
+      "\n",
+      "\n",
+      "\n"
+     ]
+    }
+   ],
+   "source": [
+    "for epoch in range(epochs):\n",
+    "    \n",
+    "    if epoch%printing_steps==0:\n",
+    "        print(\"Epoch:\", epoch)\n",
+    "    \n",
+    "    # training discriminator\n",
+    "    for d_i in range(d_step):\n",
+    "        discrim.zero_grad()\n",
+    "        \n",
+    "        #real\n",
+    "        data_d_real = Variable(get_dataset())\n",
+    "        data_d_real_pred = discrim(data_d_real)\n",
+    "        data_d_real_loss = criteriond1(data_d_real_pred,Variable(torch.ones(1,1)))\n",
+    "        data_d_real_loss.backward()\n",
+    "        \n",
+    "        \n",
+    "        #fake\n",
+    "        data_d_noise = Variable(make_noise())\n",
+    "        data_d_gen_out = gen(data_d_noise).detach()\n",
+    "        data_fake_dicrim_out = discrim(data_d_gen_out)\n",
+    "        data_fake_d_loss = criteriond1(data_fake_dicrim_out,Variable(torch.zeros(1,1)))\n",
+    "        data_fake_d_loss.backward()\n",
+    "        \n",
+    "        optimizerd1.step()\n",
+    "        \n",
+    "    for g_i in range(g_step):\n",
+    "        \n",
+    "        gen.zero_grad()\n",
+    "        \n",
+    "        data_noise_gen = Variable(make_noise())\n",
+    "        data_g_gen_out = gen(data_noise_gen)\n",
+    "        data_g_dis_out = discrim(data_g_gen_out)\n",
+    "        data_g_loss = criteriond2(data_g_dis_out,Variable(torch.ones(1,1)))\n",
+    "        data_g_loss.backward()\n",
+    "        \n",
+    "        optimizerd2.step()\n",
+    "        \n",
+    "        if epoch%printing_steps==0:\n",
+    "            print(stats(data_g_gen_out))\n",
+    "    \n",
+    "    if epoch%printing_steps==0:\n",
+    "        print(\"\\n\\n\")\n",
+    "    "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 19,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "generator(\n",
+       "  (net): Sequential(\n",
+       "    (0): Linear(in_features=50, out_features=300, bias=True)\n",
+       "    (1): ReLU(inplace)\n",
+       "    (2): Linear(in_features=300, out_features=300, bias=True)\n",
+       "    (3): ReLU(inplace)\n",
+       "    (4): Linear(in_features=300, out_features=100, bias=True)\n",
+       "  )\n",
+       ")"
+      ]
+     },
+     "execution_count": 19,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "gen"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 20,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# torch.save(gen.state_dict(),\"d_step_5000_g_step_2000\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 21,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "y = gen(make_noise())"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 22,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "[5.0801945, 0.45474356]"
+      ]
+     },
+     "execution_count": 22,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "stats(y)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 2",
+   "language": "python",
+   "name": "python2"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 2
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython2",
+   "version": "2.7.15"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}