Archivos TFG

c95dc6af · osciban · 4ef30f42 · c95dc6af · c95dc6af · c95dc6af
Commit c95dc6af authored Jun 23, 2021 by osciban
--- a/Capítulo 5/Arquitecturas y número de épocas/AlexNet10.ipynb
+++ b/Capítulo 5/Arquitecturas y número de épocas/AlexNet10.ipynb
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "id": "logical-edition",
+   "metadata": {},
+   "source": [
+    "# AlexNet"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "early-single",
+   "metadata": {
+    "scrolled": true
+   },
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "C:\\Users\\Tottem\\.conda\\envs\\fastai\\lib\\site-packages\\fastbook\\__init__.py:18: UserWarning: Missing `graphviz` - please run `conda install fastbook`\n",
+      "  except ModuleNotFoundError: warn(\"Missing `graphviz` - please run `conda install fastbook`\")\n"
+     ]
+    }
+   ],
+   "source": [
+    "import fastai\n",
+    "import fastbook\n",
+    "from fastbook import *\n",
+    "from fastai.vision.widgets import *\n",
+    "fastbook.setup_book()\n",
+    "from PIL import Image"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "final-complaint",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "path ='Balones'"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "becoming-vulnerability",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "balls = DataBlock(\n",
+    "    blocks=(ImageBlock, CategoryBlock), \n",
+    "    get_items=get_image_files, \n",
+    "    splitter=RandomSplitter(valid_pct=0.2, seed=42),\n",
+    "    get_y=parent_label,\n",
+    "    item_tfms=Resize(224))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "id": "preceding-locking",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "balls = balls.new(item_tfms=Resize(224, ResizeMethod.Pad, pad_mode='zeros'))\n",
+    "dls = balls.dataloaders(path,num_workers=0)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "id": "swiss-carry",
+   "metadata": {
+    "scrolled": false
+   },
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 864x216 with 4 Axes>"
+      ]
+     },
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "dls.valid.show_batch(max_n=4, nrows=1)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "id": "friendly-israeli",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: left;\">\n",
+       "      <th>epoch</th>\n",
+       "      <th>train_loss</th>\n",
+       "      <th>valid_loss</th>\n",
+       "      <th>error_rate</th>\n",
+       "      <th>time</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <td>0</td>\n",
+       "      <td>2.413877</td>\n",
+       "      <td>1.608709</td>\n",
+       "      <td>0.603448</td>\n",
+       "      <td>00:07</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>"
+      ],
+      "text/plain": [
+       "<IPython.core.display.HTML object>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/html": [
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: left;\">\n",
+       "      <th>epoch</th>\n",
+       "      <th>train_loss</th>\n",
+       "      <th>valid_loss</th>\n",
+       "      <th>error_rate</th>\n",
+       "      <th>time</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <td>0</td>\n",
+       "      <td>1.578609</td>\n",
+       "      <td>1.139119</td>\n",
+       "      <td>0.500000</td>\n",
+       "      <td>00:05</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <td>1</td>\n",
+       "      <td>1.533084</td>\n",
+       "      <td>0.811250</td>\n",
+       "      <td>0.258621</td>\n",
+       "      <td>00:05</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <td>2</td>\n",
+       "      <td>1.275637</td>\n",
+       "      <td>0.631664</td>\n",
+       "      <td>0.224138</td>\n",
+       "      <td>00:05</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <td>3</td>\n",
+       "      <td>1.076173</td>\n",
+       "      <td>0.583295</td>\n",
+       "      <td>0.155172</td>\n",
+       "      <td>00:05</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <td>4</td>\n",
+       "      <td>0.933389</td>\n",
+       "      <td>0.531890</td>\n",
+       "      <td>0.155172</td>\n",
+       "      <td>00:05</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <td>5</td>\n",
+       "      <td>0.817017</td>\n",
+       "      <td>0.521254</td>\n",
+       "      <td>0.155172</td>\n",
+       "      <td>00:05</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <td>6</td>\n",
+       "      <td>0.719973</td>\n",
+       "      <td>0.515153</td>\n",
+       "      <td>0.137931</td>\n",
+       "      <td>00:05</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <td>7</td>\n",
+       "      <td>0.644289</td>\n",
+       "      <td>0.515008</td>\n",
+       "      <td>0.120690</td>\n",
+       "      <td>00:05</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <td>8</td>\n",
+       "      <td>0.574483</td>\n",
+       "      <td>0.519688</td>\n",
+       "      <td>0.120690</td>\n",
+       "      <td>00:05</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <td>9</td>\n",
+       "      <td>0.526094</td>\n",
+       "      <td>0.526192</td>\n",
+       "      <td>0.120690</td>\n",
+       "      <td>00:05</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>"
+      ],
+      "text/plain": [
+       "<IPython.core.display.HTML object>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "learn = cnn_learner(dls, alexnet, metrics=error_rate, cbs=CSVLogger(\"resultados/AlexNet10.csv\"))\n",
+    "learn.fine_tune(10)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "id": "tracked-scratch",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 432x288 with 1 Axes>"
+      ]
+     },
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "learn.recorder.plot_loss()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "id": "cross-referral",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/html": [],
+      "text/plain": [
+       "<IPython.core.display.HTML object>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 432x288 with 1 Axes>"
+      ]
+     },
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "interp = ClassificationInterpretation.from_learner(learn)\n",
+    "interp.plot_confusion_matrix()"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.8.5"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}
+%% Cell type:markdown id:logical-edition tags:
+# AlexNet
+%% Cell type:code id:early-single tags:
+``` python
+import fastai
+import fastbook
+from fastbook import *
+from fastai.vision.widgets import *
+fastbook.setup_book()
+from PIL import Image
+```
+%% Output
+    C:\Users\Tottem\.conda\envs\fastai\lib\site-packages\fastbook\__init__.py:18: UserWarning: Missing `graphviz` - please run `conda install fastbook`
+      except ModuleNotFoundError: warn("Missing `graphviz` - please run `conda install fastbook`")
+%% Cell type:code id:final-complaint tags:
+``` python
+path ='Balones'
+```
+%% Cell type:code id:becoming-vulnerability tags:
+``` python
+balls = DataBlock(
+    blocks=(ImageBlock, CategoryBlock),
+    get_items=get_image_files,
+    splitter=RandomSplitter(valid_pct=0.2, seed=42),
+    get_y=parent_label,
+    item_tfms=Resize(224))
+```
+%% Cell type:code id:preceding-locking tags:
+``` python
+balls = balls.new(item_tfms=Resize(224, ResizeMethod.Pad, pad_mode='zeros'))
+dls = balls.dataloaders(path,num_workers=0)
+```
+%% Cell type:code id:swiss-carry tags:
+``` python
+dls.valid.show_batch(max_n=4, nrows=1)
+```
+%% Output
+%% Cell type:code id:friendly-israeli tags:
+``` python
+learn = cnn_learner(dls, alexnet, metrics=error_rate, cbs=CSVLogger("resultados/AlexNet10.csv"))
+learn.fine_tune(10)
+```
+%% Output
+%% Cell type:code id:tracked-scratch tags:
+``` python
+learn.recorder.plot_loss()
+```
+%% Output
+%% Cell type:code id:cross-referral tags:
+``` python
+interp = ClassificationInterpretation.from_learner(learn)
+interp.plot_confusion_matrix()
+```
+%% Output
--- a/Capítulo 5/Arquitecturas y número de épocas/AlexNet20.ipynb
+++ b/Capítulo 5/Arquitecturas y número de épocas/AlexNet20.ipynb
--- a/Capítulo 5/Arquitecturas y número de épocas/AlexNet30.ipynb
+++ b/Capítulo 5/Arquitecturas y número de épocas/AlexNet30.ipynb
--- a/Capítulo 5/Arquitecturas y número de épocas/AlexNet40.ipynb
+++ b/Capítulo 5/Arquitecturas y número de épocas/AlexNet40.ipynb
--- a/Capítulo 5/Arquitecturas y número de épocas/AlexNet50.ipynb
+++ b/Capítulo 5/Arquitecturas y número de épocas/AlexNet50.ipynb
--- a/Capítulo 5/Arquitecturas y número de épocas/DenseNet12110.ipynb
+++ b/Capítulo 5/Arquitecturas y número de épocas/DenseNet12110.ipynb
--- a/Capítulo 5/Arquitecturas y número de épocas/DenseNet12120.ipynb
+++ b/Capítulo 5/Arquitecturas y número de épocas/DenseNet12120.ipynb
--- a/Capítulo 5/Arquitecturas y número de épocas/DenseNet12130.ipynb
+++ b/Capítulo 5/Arquitecturas y número de épocas/DenseNet12130.ipynb
--- a/Capítulo 5/Arquitecturas y número de épocas/DenseNet12140.ipynb
+++ b/Capítulo 5/Arquitecturas y número de épocas/DenseNet12140.ipynb
--- a/Capítulo 5/Arquitecturas y número de épocas/DenseNet12150.ipynb
+++ b/Capítulo 5/Arquitecturas y número de épocas/DenseNet12150.ipynb
--- a/Capítulo 5/Arquitecturas y número de épocas/DenseNet16110.ipynb
+++ b/Capítulo 5/Arquitecturas y número de épocas/DenseNet16110.ipynb
--- a/Capítulo 5/Arquitecturas y número de épocas/DenseNet16120.ipynb
+++ b/Capítulo 5/Arquitecturas y número de épocas/DenseNet16120.ipynb
--- a/Capítulo 5/Arquitecturas y número de épocas/DenseNet16130.ipynb
+++ b/Capítulo 5/Arquitecturas y número de épocas/DenseNet16130.ipynb
--- a/Capítulo 5/Arquitecturas y número de épocas/DenseNet16140.ipynb
+++ b/Capítulo 5/Arquitecturas y número de épocas/DenseNet16140.ipynb
--- a/Capítulo 5/Arquitecturas y número de épocas/DenseNet16150.ipynb
+++ b/Capítulo 5/Arquitecturas y número de épocas/DenseNet16150.ipynb
--- a/Capítulo 5/Arquitecturas y número de épocas/DenseNet16910.ipynb
+++ b/Capítulo 5/Arquitecturas y número de épocas/DenseNet16910.ipynb
--- a/Capítulo 5/Arquitecturas y número de épocas/DenseNet16920.ipynb
+++ b/Capítulo 5/Arquitecturas y número de épocas/DenseNet16920.ipynb
--- a/Capítulo 5/Arquitecturas y número de épocas/DenseNet16930.ipynb
+++ b/Capítulo 5/Arquitecturas y número de épocas/DenseNet16930.ipynb
--- a/Capítulo 5/Arquitecturas y número de épocas/DenseNet16940.ipynb
+++ b/Capítulo 5/Arquitecturas y número de épocas/DenseNet16940.ipynb
--- a/Capítulo 5/Arquitecturas y número de épocas/DenseNet16950.ipynb
+++ b/Capítulo 5/Arquitecturas y número de épocas/DenseNet16950.ipynb