Skip to content

Instantly share code, notes, and snippets.

@regonn

regonn/JuliaMXNet.ipynb

Created November 11, 2017 10:01
Show Gist options

  • Select an option

    Learn more about clone URLs
  • Save regonn/e6629c28fdf6f9e2e0cc57a6c2a5bc80 to your computer and use it in GitHub Desktop.

Select an option

Learn more about clone URLs
Save regonn/e6629c28fdf6f9e2e0cc57a6c2a5bc80 to your computer and use it in GitHub Desktop.
Download ZIP
Julia MXNet Titanic 2017-11-11
Display the source blob
Display the rendered blob
Raw
JuliaMXNet.ipynb
{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"using MXNet\n",
"using DataFrames"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {
"scrolled": true
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"The columns in the train set are:\n",
"\n",
"Symbol[:PassengerId, :Survived, :Pclass, :Name, :Sex, :Age, :SibSp, :Parch, :Ticket, :Fare, :Cabin, :Embarked]\n",
"\n",
"There are 891 rows in the training set\n",
"The columns in the train set are:\n",
"\n",
"Symbol[:PassengerId, :Survived, :Pclass, :Name, :Sex, :Age, :SibSp, :Parch, :Ticket, :Fare, :Cabin, :Embarked]\n"
]
},
{
"data": {
"text/html": [
"<table class=\"data-frame\"><thead><tr><th></th><th>PassengerId</th><th>Pclass</th><th>Name</th><th>Sex</th><th>Age</th><th>SibSp</th><th>Parch</th><th>Ticket</th><th>Fare</th><th>Cabin</th><th>Embarked</th></tr></thead><tbody><tr><th>1</th><td>892</td><td>3</td><td>Kelly, Mr. James</td><td>male</td><td>34.5</td><td>0</td><td>0</td><td>330911</td><td>7.8292</td><td>NA</td><td>Q</td></tr><tr><th>2</th><td>893</td><td>3</td><td>Wilkes, Mrs. James (Ellen Needs)</td><td>female</td><td>47.0</td><td>1</td><td>0</td><td>363272</td><td>7.0</td><td>NA</td><td>S</td></tr><tr><th>3</th><td>894</td><td>2</td><td>Myles, Mr. Thomas Francis</td><td>male</td><td>62.0</td><td>0</td><td>0</td><td>240276</td><td>9.6875</td><td>NA</td><td>Q</td></tr><tr><th>4</th><td>895</td><td>3</td><td>Wirz, Mr. Albert</td><td>male</td><td>27.0</td><td>0</td><td>0</td><td>315154</td><td>8.6625</td><td>NA</td><td>S</td></tr><tr><th>5</th><td>896</td><td>3</td><td>Hirvonen, Mrs. Alexander (Helga E Lindqvist)</td><td>female</td><td>22.0</td><td>1</td><td>1</td><td>3101298</td><td>12.2875</td><td>NA</td><td>S</td></tr><tr><th>6</th><td>897</td><td>3</td><td>Svensson, Mr. Johan Cervin</td><td>male</td><td>14.0</td><td>0</td><td>0</td><td>7538</td><td>9.225</td><td>NA</td><td>S</td></tr><tr><th>7</th><td>898</td><td>3</td><td>Connolly, Miss. Kate</td><td>female</td><td>30.0</td><td>0</td><td>0</td><td>330972</td><td>7.6292</td><td>NA</td><td>Q</td></tr><tr><th>8</th><td>899</td><td>2</td><td>Caldwell, Mr. Albert Francis</td><td>male</td><td>26.0</td><td>1</td><td>1</td><td>248738</td><td>29.0</td><td>NA</td><td>S</td></tr><tr><th>9</th><td>900</td><td>3</td><td>Abrahim, Mrs. Joseph (Sophie Halaut Easu)</td><td>female</td><td>18.0</td><td>0</td><td>0</td><td>2657</td><td>7.2292</td><td>NA</td><td>C</td></tr><tr><th>10</th><td>901</td><td>3</td><td>Davies, Mr. John Samuel</td><td>male</td><td>21.0</td><td>2</td><td>0</td><td>A/4 48871</td><td>24.15</td><td>NA</td><td>S</td></tr><tr><th>11</th><td>902</td><td>3</td><td>Ilieff, Mr. Ylio</td><td>male</td><td>NA</td><td>0</td><td>0</td><td>349220</td><td>7.8958</td><td>NA</td><td>S</td></tr><tr><th>12</th><td>903</td><td>1</td><td>Jones, Mr. Charles Cresson</td><td>male</td><td>46.0</td><td>0</td><td>0</td><td>694</td><td>26.0</td><td>NA</td><td>S</td></tr><tr><th>13</th><td>904</td><td>1</td><td>Snyder, Mrs. John Pillsbury (Nelle Stevenson)</td><td>female</td><td>23.0</td><td>1</td><td>0</td><td>21228</td><td>82.2667</td><td>B45</td><td>S</td></tr><tr><th>14</th><td>905</td><td>2</td><td>Howard, Mr. Benjamin</td><td>male</td><td>63.0</td><td>1</td><td>0</td><td>24065</td><td>26.0</td><td>NA</td><td>S</td></tr><tr><th>15</th><td>906</td><td>1</td><td>Chaffee, Mrs. Herbert Fuller (Carrie Constance Toogood)</td><td>female</td><td>47.0</td><td>1</td><td>0</td><td>W.E.P. 5734</td><td>61.175</td><td>E31</td><td>S</td></tr><tr><th>16</th><td>907</td><td>2</td><td>del Carlo, Mrs. Sebastiano (Argenia Genovesi)</td><td>female</td><td>24.0</td><td>1</td><td>0</td><td>SC/PARIS 2167</td><td>27.7208</td><td>NA</td><td>C</td></tr><tr><th>17</th><td>908</td><td>2</td><td>Keane, Mr. Daniel</td><td>male</td><td>35.0</td><td>0</td><td>0</td><td>233734</td><td>12.35</td><td>NA</td><td>Q</td></tr><tr><th>18</th><td>909</td><td>3</td><td>Assaf, Mr. Gerios</td><td>male</td><td>21.0</td><td>0</td><td>0</td><td>2692</td><td>7.225</td><td>NA</td><td>C</td></tr><tr><th>19</th><td>910</td><td>3</td><td>Ilmakangas, Miss. Ida Livija</td><td>female</td><td>27.0</td><td>1</td><td>0</td><td>STON/O2. 3101270</td><td>7.925</td><td>NA</td><td>S</td></tr><tr><th>20</th><td>911</td><td>3</td><td>Assaf Khalil, Mrs. Mariana (Miriam\")\"</td><td>female</td><td>45.0</td><td>0</td><td>0</td><td>2696</td><td>7.225</td><td>NA</td><td>C</td></tr><tr><th>21</th><td>912</td><td>1</td><td>Rothschild, Mr. Martin</td><td>male</td><td>55.0</td><td>1</td><td>0</td><td>PC 17603</td><td>59.4</td><td>NA</td><td>C</td></tr><tr><th>22</th><td>913</td><td>3</td><td>Olsen, Master. Artur Karl</td><td>male</td><td>9.0</td><td>0</td><td>1</td><td>C 17368</td><td>3.1708</td><td>NA</td><td>S</td></tr><tr><th>23</th><td>914</td><td>1</td><td>Flegenheim, Mrs. Alfred (Antoinette)</td><td>female</td><td>NA</td><td>0</td><td>0</td><td>PC 17598</td><td>31.6833</td><td>NA</td><td>S</td></tr><tr><th>24</th><td>915</td><td>1</td><td>Williams, Mr. Richard Norris II</td><td>male</td><td>21.0</td><td>0</td><td>1</td><td>PC 17597</td><td>61.3792</td><td>NA</td><td>C</td></tr><tr><th>25</th><td>916</td><td>1</td><td>Ryerson, Mrs. Arthur Larned (Emily Maria Borie)</td><td>female</td><td>48.0</td><td>1</td><td>3</td><td>PC 17608</td><td>262.375</td><td>B57 B59 B63 B66</td><td>C</td></tr><tr><th>26</th><td>917</td><td>3</td><td>Robins, Mr. Alexander A</td><td>male</td><td>50.0</td><td>1</td><td>0</td><td>A/5. 3337</td><td>14.5</td><td>NA</td><td>S</td></tr><tr><th>27</th><td>918</td><td>1</td><td>Ostby, Miss. Helene Ragnhild</td><td>female</td><td>22.0</td><td>0</td><td>1</td><td>113509</td><td>61.9792</td><td>B36</td><td>C</td></tr><tr><th>28</th><td>919</td><td>3</td><td>Daher, Mr. Shedid</td><td>male</td><td>22.5</td><td>0</td><td>0</td><td>2698</td><td>7.225</td><td>NA</td><td>C</td></tr><tr><th>29</th><td>920</td><td>1</td><td>Brady, Mr. John Bertram</td><td>male</td><td>41.0</td><td>0</td><td>0</td><td>113054</td><td>30.5</td><td>A21</td><td>S</td></tr><tr><th>30</th><td>921</td><td>3</td><td>Samaan, Mr. Elias</td><td>male</td><td>NA</td><td>2</td><td>0</td><td>2662</td><td>21.6792</td><td>NA</td><td>C</td></tr><tr><th>&vellip;</th><td>&vellip;</td><td>&vellip;</td><td>&vellip;</td><td>&vellip;</td><td>&vellip;</td><td>&vellip;</td><td>&vellip;</td><td>&vellip;</td><td>&vellip;</td><td>&vellip;</td><td>&vellip;</td></tr></tbody></table>"
],
"text/plain": [
"418×11 DataFrames.DataFrame\n",
"│ Row │ PassengerId │ Pclass │\n",
"├─────┼─────────────┼────────┤\n",
"│ 1 │ 892 │ 3 │\n",
"│ 2 │ 893 │ 3 │\n",
"│ 3 │ 894 │ 2 │\n",
"│ 4 │ 895 │ 3 │\n",
"│ 5 │ 896 │ 3 │\n",
"│ 6 │ 897 │ 3 │\n",
"│ 7 │ 898 │ 3 │\n",
"│ 8 │ 899 │ 2 │\n",
"│ 9 │ 900 │ 3 │\n",
"│ 10 │ 901 │ 3 │\n",
"│ 11 │ 902 │ 3 │\n",
"⋮\n",
"│ 407 │ 1298 │ 2 │\n",
"│ 408 │ 1299 │ 1 │\n",
"│ 409 │ 1300 │ 3 │\n",
"│ 410 │ 1301 │ 3 │\n",
"│ 411 │ 1302 │ 3 │\n",
"│ 412 │ 1303 │ 1 │\n",
"│ 413 │ 1304 │ 3 │\n",
"│ 414 │ 1305 │ 3 │\n",
"│ 415 │ 1306 │ 1 │\n",
"│ 416 │ 1307 │ 3 │\n",
"│ 417 │ 1308 │ 3 │\n",
"│ 418 │ 1309 │ 3 │\n",
"\n",
"│ Row │ Name │ Sex │ Age │\n",
"├─────┼───────────────────────────────────────────────────┼──────────┼──────┤\n",
"│ 1 │ \"Kelly, Mr. James\" │ \"male\" │ 34.5 │\n",
"│ 2 │ \"Wilkes, Mrs. James (Ellen Needs)\" │ \"female\" │ 47.0 │\n",
"│ 3 │ \"Myles, Mr. Thomas Francis\" │ \"male\" │ 62.0 │\n",
"│ 4 │ \"Wirz, Mr. Albert\" │ \"male\" │ 27.0 │\n",
"│ 5 │ \"Hirvonen, Mrs. Alexander (Helga E Lindqvist)\" │ \"female\" │ 22.0 │\n",
"│ 6 │ \"Svensson, Mr. Johan Cervin\" │ \"male\" │ 14.0 │\n",
"│ 7 │ \"Connolly, Miss. Kate\" │ \"female\" │ 30.0 │\n",
"│ 8 │ \"Caldwell, Mr. Albert Francis\" │ \"male\" │ 26.0 │\n",
"│ 9 │ \"Abrahim, Mrs. Joseph (Sophie Halaut Easu)\" │ \"female\" │ 18.0 │\n",
"│ 10 │ \"Davies, Mr. John Samuel\" │ \"male\" │ 21.0 │\n",
"│ 11 │ \"Ilieff, Mr. Ylio\" │ \"male\" │ NA │\n",
"⋮\n",
"│ 407 │ \"Ware, Mr. William Jeffery\" │ \"male\" │ 23.0 │\n",
"│ 408 │ \"Widener, Mr. George Dunton\" │ \"male\" │ 50.0 │\n",
"│ 409 │ \"Riordan, Miss. Johanna Hannah\\\"\\\"\" │ \"female\" │ NA │\n",
"│ 410 │ \"Peacock, Miss. Treasteall\" │ \"female\" │ 3.0 │\n",
"│ 411 │ \"Naughton, Miss. Hannah\" │ \"female\" │ NA │\n",
"│ 412 │ \"Minahan, Mrs. William Edward (Lillian E Thorpe)\" │ \"female\" │ 37.0 │\n",
"│ 413 │ \"Henriksson, Miss. Jenny Lovisa\" │ \"female\" │ 28.0 │\n",
"│ 414 │ \"Spector, Mr. Woolf\" │ \"male\" │ NA │\n",
"│ 415 │ \"Oliva y Ocana, Dona. Fermina\" │ \"female\" │ 39.0 │\n",
"│ 416 │ \"Saether, Mr. Simon Sivertsen\" │ \"male\" │ 38.5 │\n",
"│ 417 │ \"Ware, Mr. Frederick\" │ \"male\" │ NA │\n",
"│ 418 │ \"Peter, Master. Michael J\" │ \"male\" │ NA │\n",
"\n",
"│ Row │ SibSp │ Parch │ Ticket │ Fare │ Cabin │ Embarked │\n",
"├─────┼───────┼───────┼──────────────────────┼─────────┼────────┼──────────┤\n",
"│ 1 │ 0 │ 0 │ \"330911\" │ 7.8292 │ NA │ \"Q\" │\n",
"│ 2 │ 1 │ 0 │ \"363272\" │ 7.0 │ NA │ \"S\" │\n",
"│ 3 │ 0 │ 0 │ \"240276\" │ 9.6875 │ NA │ \"Q\" │\n",
"│ 4 │ 0 │ 0 │ \"315154\" │ 8.6625 │ NA │ \"S\" │\n",
"│ 5 │ 1 │ 1 │ \"3101298\" │ 12.2875 │ NA │ \"S\" │\n",
"│ 6 │ 0 │ 0 │ \"7538\" │ 9.225 │ NA │ \"S\" │\n",
"│ 7 │ 0 │ 0 │ \"330972\" │ 7.6292 │ NA │ \"Q\" │\n",
"│ 8 │ 1 │ 1 │ \"248738\" │ 29.0 │ NA │ \"S\" │\n",
"│ 9 │ 0 │ 0 │ \"2657\" │ 7.2292 │ NA │ \"C\" │\n",
"│ 10 │ 2 │ 0 │ \"A/4 48871\" │ 24.15 │ NA │ \"S\" │\n",
"│ 11 │ 0 │ 0 │ \"349220\" │ 7.8958 │ NA │ \"S\" │\n",
"⋮\n",
"│ 407 │ 1 │ 0 │ \"28666\" │ 10.5 │ NA │ \"S\" │\n",
"│ 408 │ 1 │ 1 │ \"113503\" │ 211.5 │ \"C80\" │ \"C\" │\n",
"│ 409 │ 0 │ 0 │ \"334915\" │ 7.7208 │ NA │ \"Q\" │\n",
"│ 410 │ 1 │ 1 │ \"SOTON/O.Q. 3101315\" │ 13.775 │ NA │ \"S\" │\n",
"│ 411 │ 0 │ 0 │ \"365237\" │ 7.75 │ NA │ \"Q\" │\n",
"│ 412 │ 1 │ 0 │ \"19928\" │ 90.0 │ \"C78\" │ \"Q\" │\n",
"│ 413 │ 0 │ 0 │ \"347086\" │ 7.775 │ NA │ \"S\" │\n",
"│ 414 │ 0 │ 0 │ \"A.5. 3236\" │ 8.05 │ NA │ \"S\" │\n",
"│ 415 │ 0 │ 0 │ \"PC 17758\" │ 108.9 │ \"C105\" │ \"C\" │\n",
"│ 416 │ 0 │ 0 │ \"SOTON/O.Q. 3101262\" │ 7.25 │ NA │ \"S\" │\n",
"│ 417 │ 0 │ 0 │ \"359309\" │ 8.05 │ NA │ \"S\" │\n",
"│ 418 │ 1 │ 1 │ \"2668\" │ 22.3583 │ NA │ \"C\" │"
]
},
"execution_count": 3,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# Load in the data\n",
"train = readtable(\"./input/train.csv\")\n",
"\n",
"println(\"The columns in the train set are:\\n\")\n",
"println(names(train))\n",
"println(@sprintf(\"\\nThere are %d rows in the training set\", nrow(train)))\n",
"# It's yours to take from here!\n",
"test = readtable(\"./input/test.csv\")\n",
"# Load in the data\n",
"train = readtable(\"./input/train.csv\")\n",
"\n",
"println(\"The columns in the train set are:\\n\")\n",
"println(names(train))\n",
"# It's yours to take from here!\n",
"test = readtable(\"./input/test.csv\")"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"\"female3\""
]
},
"execution_count": 4,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"#read the non NA labels\n",
"xTrain_df = train[:,[3,7,8,10]]\n",
"xTest_df = test[:,[2,6,7]]\n",
"yTrain_df = train[:2]\n",
"\n",
"# new column Pclass x Sex\n",
"train[:PclassSex] = \"\"\n",
"train[(train[:Sex] .== \"male\").&(train[:Pclass] .== 1), :PclassSex] = \"male1\"\n",
"train[(train[:Sex] .== \"male\").&(train[:Pclass] .== 2), :PclassSex] = \"male2\"\n",
"train[(train[:Sex] .== \"male\").&(train[:Pclass] .== 3), :PclassSex] = \"male3\"\n",
"train[(train[:Sex] .== \"female\").&(train[:Pclass] .== 1), :PclassSex] = \"female1\"\n",
"train[(train[:Sex] .== \"female\").&(train[:Pclass] .== 2), :PclassSex] = \"female2\"\n",
"train[(train[:Sex] .== \"female\").&(train[:Pclass] .== 3), :PclassSex] = \"female3\"\n",
"\n",
"# age_mean\n",
"male_pclass1_age_mean = mean(dropna(train[(train[:Sex] .== \"male\") .& (train[:Pclass] .== 1), :Age]))\n",
"male_pclass2_age_mean = mean(dropna(train[(train[:Sex] .== \"male\") .& (train[:Pclass] .== 2), :Age]))\n",
"male_pclass3_age_mean = mean(dropna(train[(train[:Sex] .== \"male\") .& (train[:Pclass] .== 3), :Age]))\n",
"female_pclass1_age_mean = mean(dropna(train[(train[:Sex] .== \"female\") .& (train[:Pclass] .== 1), :Age]))\n",
"female_pclass2_age_mean = mean(dropna(train[(train[:Sex] .== \"female\") .& (train[:Pclass] .== 2), :Age]))\n",
"female_pclass3_age_mean = mean(dropna(train[(train[:Sex] .== \"female\") .& (train[:Pclass] .== 3), :Age]))\n",
"\n",
"train[(isna.(train[:Age]).&(train[:Sex] .== \"male\").&(train[:Pclass] .== 1)), :Age] = male_pclass1_age_mean\n",
"train[(isna.(train[:Age]).&(train[:Sex] .== \"male\").&(train[:Pclass] .== 2)), :Age] = male_pclass2_age_mean\n",
"train[(isna.(train[:Age]).&(train[:Sex] .== \"male\").&(train[:Pclass] .== 3)), :Age] = male_pclass3_age_mean\n",
"train[(isna.(train[:Age]).&(train[:Sex] .== \"female\").&(train[:Pclass] .== 1)), :Age] = female_pclass1_age_mean\n",
"train[(isna.(train[:Age]).&(train[:Sex] .== \"female\").&(train[:Pclass] .== 2)), :Age] = female_pclass2_age_mean\n",
"train[(isna.(train[:Age]).&(train[:Sex] .== \"female\").&(train[:Pclass] .== 3)), :Age] = female_pclass3_age_mean\n",
"\n",
"test[(isna.(test[:Age]).&(test[:Sex] .== \"male\").&(test[:Pclass] .== 1)), :Age] = male_pclass1_age_mean\n",
"test[(isna.(test[:Age]).&(test[:Sex] .== \"male\").&(test[:Pclass] .== 2)), :Age] = male_pclass2_age_mean\n",
"test[(isna.(test[:Age]).&(test[:Sex] .== \"male\").&(test[:Pclass] .== 3)), :Age] = male_pclass3_age_mean\n",
"test[(isna.(test[:Age]).&(test[:Sex] .== \"female\").&(test[:Pclass] .== 1)), :Age] = female_pclass1_age_mean\n",
"test[(isna.(test[:Age]).&(test[:Sex] .== \"female\").&(test[:Pclass] .== 2)), :Age] = female_pclass2_age_mean\n",
"test[(isna.(test[:Age]).&(test[:Sex] .== \"female\").&(test[:Pclass] .== 3)), :Age] = female_pclass3_age_mean\n",
"\n",
"train[:AgeBand] = 0\n",
"train[train[:Age].<= 16, :AgeBand] = 0\n",
"train[16 .< train[:Age] .<= 32, :AgeBand] = 1\n",
"train[32 .< train[:Age] .<= 48, :AgeBand] = 2\n",
"train[48 .< train[:Age] .<= 64, :AgeBand] = 3\n",
"train[train[:Age].> 64, :AgeBand] = 4\n",
"\n",
"test[:AgeBand] = 0\n",
"test[test[:Age].<= 16, :AgeBand] = 0\n",
"test[16 .< test[:Age] .<= 32, :AgeBand] = 1\n",
"test[32 .< test[:Age] .<= 48, :AgeBand] = 2\n",
"test[48 .< test[:Age] .<= 64, :AgeBand] = 3\n",
"test[test[:Age].> 64, :AgeBand] = 4\n",
"\n",
"#convert Int\n",
"xTrain_df[:Fare] = round.(Int, xTrain_df[:Fare])\n",
"\n",
"#transform into arrays\n",
"xTrain_array = Array(xTrain_df)\n",
"xTest_array = Array(xTest_df)\n",
"yTrain_array = Array(yTrain_df)\n",
"\n",
"#read the NAs age, sex and fare (fare only for test sample)\n",
"sexTrain_df = train[:Sex]\n",
"sexTest_df = test[:Sex]\n",
"fareTest_df = test[:Fare]\n",
"\n",
"#calculate the average of the ages and fares\n",
"fareTest_avg = mean(train[:Fare])\n",
"\n",
"#convert sex into integer\n",
"sexTrain_int_array = Int.(map(x->x.== \"male\",sexTrain_df))\n",
"sexTest_int_array = Int.(map(x->x.== \"male\",sexTest_df))\n",
"\n",
"#convert into arrays and fill NAs with mean values \n",
"ageTrain_array = convert(Array,train[:AgeBand]);\n",
"ageTest_array = convert(Array,test[:AgeBand]);\n",
"fareTest_array = convert(Array,fareTest_df,fareTest_avg);\n",
"\n",
"#glue the original arrays and the modified ones\n",
"xTrain_array = [xTrain_array ageTrain_array sexTrain_int_array]\n",
"xTest_array = [xTest_array fareTest_array ageTest_array sexTest_int_array]"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"100-element Array{Int64,1}:\n",
" 0\n",
" 0\n",
" 0\n",
" 0\n",
" 0\n",
" 1\n",
" 1\n",
" 0\n",
" 0\n",
" 0\n",
" 1\n",
" 1\n",
" 1\n",
" ⋮\n",
" 1\n",
" 1\n",
" 0\n",
" 0\n",
" 0\n",
" 0\n",
" 0\n",
" 0\n",
" 1\n",
" 0\n",
" 1\n",
" 0"
]
},
"execution_count": 12,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"xTrain = xTrain_array[1:791]\n",
"yTrain = yTrain_array[1:791]\n",
"xValidation = xTrain_array[792:891]\n",
"yValidation = yTrain_array[792:891]"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"MXNet.mx.ArrayDataProvider(Array[Float32[2.0 3.0 … 1.0 3.0]], Symbol[:data], Array[Float32[0.0 0.0 … 1.0 0.0]], Symbol[:softmax_label], 20, 100, true, 0.0f0, 0.0f0, MXNet.mx.NDArray[20 mx.NDArray{Float32} @ CPU0:\n",
" 4.99718f-36\n",
" 0.0 \n",
" 0.0 \n",
" 0.0 \n",
" 0.0 \n",
" 0.0 \n",
" 0.0 \n",
" 0.0 \n",
" 0.0 \n",
" 0.0 \n",
" 0.0 \n",
" 0.0 \n",
" 0.0 \n",
" 0.0 \n",
" 0.0 \n",
" 0.0 \n",
" 0.0 \n",
" 0.0 \n",
" 0.0 \n",
" 0.0 ], MXNet.mx.NDArray[20 mx.NDArray{Float32} @ CPU0:\n",
" 4.96016f-36\n",
" 0.0 \n",
" 3.19135f-36\n",
" 0.0 \n",
" 3.19223f-36\n",
" 0.0 \n",
" 3.19135f-36\n",
" 0.0 \n",
" 3.19223f-36\n",
" 0.0 \n",
" 3.19227f-36\n",
" 0.0 \n",
" 1.12104f-44\n",
" 0.0 \n",
" 3.06936f-36\n",
" 0.0 \n",
" 2.52234f-44\n",
" 0.0 \n",
" 2.52234f-44\n",
" 0.0 ])"
]
},
"execution_count": 13,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"batchsize = 20 # can adjust this later, but must be defined now for next line\n",
"trainprovider = mx.ArrayDataProvider(:data => xTrain, batch_size=batchsize, shuffle=true, :softmax_label => yTrain)\n",
"validationprovider = mx.ArrayDataProvider(:data => xValidation, batch_size=batchsize, shuffle=true, :softmax_label => yValidation)"
]
},
{
"cell_type": "code",
"execution_count": 61,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"MXNet.mx.SymbolicNode softmax"
]
},
"execution_count": 61,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"data = mx.Variable(:data)\n",
"label = mx.Variable(:label)\n",
"mlp = @mx.chain mx.Variable(:data) =>\n",
" mx.FullyConnected(name=:fc1, num_hidden=6) =>\n",
" mx.Activation(name=:relu1, act_type=:relu) =>\n",
" mx.FullyConnected(name=:fc2, num_hidden=10) =>\n",
" mx.Activation(name=:relu2, act_type=:relu) =>\n",
" mx.FullyConnected(name=:fc3, num_hidden=2) =>\n",
" mx.SoftmaxOutput(name=:softmax)"
]
},
{
"cell_type": "code",
"execution_count": 62,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"MXNet.mx.FeedForward(MXNet.mx.SymbolicNode softmax, MXNet.mx.Context[GPU0], #undef, #undef, #undef)"
]
},
"execution_count": 62,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"model = mx.FeedForward(mlp, context=mx.gpu(0))"
]
},
{
"cell_type": "code",
"execution_count": 63,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"MXNet.mx.SGD(MXNet.mx.SGDOptions(0.1, 0.9, 0, 0.0001, MXNet.mx.LearningRate.Fixed(0.1), MXNet.mx.Momentum.Fixed(0.9)), #undef)"
]
},
"execution_count": 63,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"optimizer = mx.SGD(lr=0.1, momentum=0.9)"
]
},
{
"cell_type": "code",
"execution_count": 64,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36mStart training on MXNet.mx.Context[GPU0]\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36mInitializing parameters...\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36mCreating KVStore...\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36mTempSpace: Total 0 MB allocated on GPU0\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36mStart training...\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 001/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1207 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 002/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1039 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 003/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1086 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 004/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1220 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 005/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1055 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 006/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.5775\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1048 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 007/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1042 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 008/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1189 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 009/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1058 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 010/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1048 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 011/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6050\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1211 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 012/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1065 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 013/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1064 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 014/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6050\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1186 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 015/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1066 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 016/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1062 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 017/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1086 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 018/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1069 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 019/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.0995 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 020/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1197 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 021/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1062 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 022/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1066 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 023/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1167 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 024/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1074 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 025/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.0987 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 026/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1171 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 027/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6000\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1078 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 028/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1075 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 029/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1180 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 030/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1064 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 031/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.0993 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 032/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1183 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 033/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.5825\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1060 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 034/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1076 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 035/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.5825\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1200 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 036/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1056 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 037/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1007 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 038/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1206 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 039/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6050\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1059 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 040/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1078 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 041/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1004 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 042/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1126 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 043/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1009 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 044/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.0940 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 045/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1140 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 046/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1058 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 047/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.5825\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1060 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 048/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1117 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 049/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1053 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m== Epoch 050/050 ==========\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Training summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6175\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m time = 0.1072 seconds\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m## Validation summary\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36m accuracy = 0.6400\n",
"\u001b[39m\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36mFinish training on MXNet.mx.Context[GPU0]\n",
"\u001b[39m"
]
}
],
"source": [
"mx.fit(model, optimizer, trainprovider, eval_data=validationprovider, n_epoch = 50)"
]
},
{
"cell_type": "code",
"execution_count": 60,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"\u001b[1m\u001b[36mINFO: \u001b[39m\u001b[22m\u001b[36mTempSpace: Total 0 MB allocated on GPU0\n",
"\u001b[39m"
]
},
{
"data": {
"text/plain": [
"2×100 Array{Float32,2}:\n",
" 0.6184 0.6184 0.6184 0.6184 0.6184 … 0.6184 0.6184 0.6184 0.6184\n",
" 0.3816 0.3816 0.3816 0.3816 0.3816 0.3816 0.3816 0.3816 0.3816"
]
},
"execution_count": 60,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"fit = mx.predict(model, validationprovider)"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Julia 0.6.1",
"language": "julia",
"name": "julia-0.6"
},
"language_info": {
"file_extension": ".jl",
"mimetype": "application/julia",
"name": "julia",
"version": "0.6.1"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment