lr_finder_example.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%reload_ext autoreload\n",
    "%autoreload 2\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from fastai.text import *"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Retrieval"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "bs=64"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "path = untar_data(URLs.IMDB)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_lm = (TextList.from_folder(path)\n",
    "           #Inputs: all the text files in path\n",
    "            .filter_by_folder(include=['train', 'test', 'unsup']) \n",
    "           #We may have other temp folders that contain text files so we only keep what's in train and test\n",
    "            .split_by_rand_pct(0.1)\n",
    "           #We randomly split and keep 10% (10,000 reviews) for validation\n",
    "            .label_for_lm()           \n",
    "           #We want to do a language model so we label accordingly\n",
    "            .databunch(bs=bs))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# LRFinder Initialization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn = language_model_learner(data_lm, AWD_LSTM, drop_mult=0.3)\n",
    "learn.unfreeze()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch.nn.functional as F\n",
    "from torch.optim import Adam\n",
    "\n",
    "model = learn.model\n",
    "opt = Adam(model.parameters())\n",
    "train_dl = learn.data.train_dl\n",
    "def loss_fn(model, xb, yb):\n",
    "    logits = model(xb)[0]\n",
    "    logits_size = logits.size()\n",
    "    return F.cross_entropy(logits.view([logits_size[0] * logits_size[1], -1]), yb.view(-1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ubuntu/pytorch/torch/optim/lr_scheduler.py:72: UserWarning: Detected call of `lr_scheduler.step()` before `optimizer.step()`. In PyTorch 1.1.0 and later, you should call them in the opposite order: `optimizer.step()` before `lr_scheduler.step()`.  Failure to do this will result in PyTorch skipping the first value of the learning rate schedule.See more details at https://pytorch.org/docs/stable/optim.html#how-to-adjust-learning-rate\n",
      "  \"https://pytorch.org/docs/stable/optim.html#how-to-adjust-learning-rate\", UserWarning)\n"
     ]
    }
   ],
   "source": [
    "from torch.utils.tensorboard.lr_finder import LRFinder\n",
    "lr_finder = LRFinder(model, train_dl, opt, loss_fn)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# LRFinder Usage"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "lr_finder.find()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Set `push_to_tensorboard` to `False` in order to view the plot directly."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEKCAYAAAD9xUlFAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzt3X+cXHV97/HXZ2dmZ39vks3mh0lICEZQQDBZKciVC0Wt8LBSFVt86BWhj6ZYi7U+9F57fdTb6vVHa63FcgtFW26tSh/lV/0FitpLoSJiEvIDCZIfJLD5udlN9vfO7Mx87h/n7DAMu8km2TNzZvf9fHAeO3POmTmfPWTnPed7vud7zN0REREBqKt2ASIiEh8KBRERKVIoiIhIkUJBRESKFAoiIlKkUBARkSKFgoiIFCkURESkSKEgIiJFyWoXcLIWLlzoq1atqnYZIiI1ZePGjUfcvfNE69VcKKxatYoNGzZUuwwRkZpiZnuns56aj0REpEihICIiRQoFEREpUiiIiEiRQkFERIoUCiIiUqRQEBGRIoWCiEgNuOXHO3h0R0/k21EoiIjEXKHg3PKTZ3niub7It6VQEBGJuWOj4xQcOprrI9+WQkFEJOb6hjMALGhJR74thYKISMwdGcoCOlIQERGgbzgMhRaFgojInNcbhsICHSmIiEjvUHBOYX6TQkFEZM7rG84yrylFKhH9R7ZCQUQk5nqHsxVpOgKFgohI7PUOZSrS8wgUCiIisdc3nKWjOfprFEChICISe33DWRZUoDsqKBRERGKtUPDwSKHGQ8HMzjazzSXTgJl9pGydy82sv2SdT0VVj4hILarkuEcAyaje2N1/BVwIYGYJYB9w/ySrPurub4uqDhGRWlbJcY+gcs1HVwK73H1vhbYnIjIrVHLcI6hcKFwH3DXFskvMbIuZPWhm51aoHhGRmtBXwSEuoAKhYGb1wNuBuydZvAlY6e4XAH8L/NsU77HezDaY2YaenujvPCQiEhe9FRwMDypzpHAVsMndD5UvcPcBdx8KHz8ApMxs4STr3eHuXe7e1dnZGX3FIiIxUclxj6AyofAepmg6MrMlZmbh44vCenorUJOISE3oG87S3liZcY8gwt5HAGbWBLwZ+P2SeTcBuPvtwLXAB80sB4wC17m7R1mTiEgt6R3OVqzpCCIOBXcfATrK5t1e8vhW4NYoaxARqWWVHPcIdEWziEis9VVwhFRQKIiIxFrfcJaOCl24BgoFEZHYqvS4R6BQEBGJrYlxj9R8JCIixXGP1HwkIiIVH/cIFAoiIrFV6XGPQKEgIhJblR73CBQKIiKxVelxj0ChICISW5Ue9wgUCiIisVXpcY9AoSAiEluVHvcIFAoiIrFV6XGPQKEgIhJblR73CBQKIiKxVI1xj0ChICISS9UY9wgUCiIisVSNcY9AoSAiEkvVGPcIFAoiIrFUjXGPQKEgIhJL1Rj3CBQKIiKxVI1xj0ChICISS9UY9wgUCiIisVSNcY9AoSAiEkvVGPcIIgwFMzvbzDaXTANm9pGydczMvmJmO81sq5mtjaoeEZFacngww6LWhopvN7JQcPdfufuF7n4hsA4YAe4vW+0qYE04rQdui6oeEZFa4e4c7B9jcdssCoUyVwK73H1v2fxrgK974HFgnpktrVBNIiKxNJjJMZLNs6S9slczQ+VC4TrgrknmLwNeKHneHc4TEZmzDvWPAbCkvbHi2448FMysHng7cPdkiyeZ55O8x3oz22BmG3p6ema6RBGRWDk4EIbCLG0+ugrY5O6HJlnWDawoeb4c2F++krvf4e5d7t7V2dkZUZkiIvFwoH92h8J7mLzpCOA7wPvDXkgXA/3ufqACNYmIxNZE89GitsqfU0hG+eZm1gS8Gfj9knk3Abj77cADwNXAToLeSTdEWY+ISC04MDDGguZ6GlKJim870lBw9xGgo2ze7SWPHfhQlDWIiNSaQ1Xqjgq6ollEJHYODoyxtF2hICIiULUL10ChICISK5lcnt7hbFV6HoFCQUQkVg4PBPdRUPORiIgUL1xbrFAQEZGD4TUKOlIQEZFiKOhEs4iIcHBgjMZUgraGSC8jm5JCQUQkRiauUTCbbLzQ6CkURERipJrXKIBCQUQkVg72j7GkSieZQaEgIhIbhYJzeFChICIiQO9wlvG8V+1qZlAoiIjExqGB6nZHBYWCiEhsHKjyhWugUBARiY3ivZkVCiIicqh/jESdsbCl8rfhnKBQEBGJiQP9YyxqTZOoq86Fa6BQEBGJjUMD1b1wDRQKIiKxcaB/tKonmUGhICISG4cGMjpSEBERGBwbZyiTq2rPI1AoiIjEwsSFa2o+EhERDvYH92ae1c1HZjbPzO4xs2fMbLuZXVK2/HIz6zezzeH0qSjrERGJq+KFa1UOhahv7XML8AN3v9bM6oGmSdZ51N3fFnEdIiKxdrB/FKju1cwQYSiYWRtwGfABAHfPAtmoticiUsv2948xvylFQypR1TqibD5aDfQAd5rZk2b2NTNrnmS9S8xsi5k9aGbnRliPiEhs7e4ZYnVnS7XLiDQUksBa4DZ3fx0wDHyibJ1NwEp3vwD4W+DfJnsjM1tvZhvMbENPT0+EJYuIVMeunmHO6pzse3NlRRkK3UC3u/88fH4PQUgUufuAuw+Fjx8AUma2sPyN3P0Od+9y967Ozs4ISxYRqbz+0XF6BjOcNZuPFNz9IPCCmZ0dzroSeLp0HTNbYmYWPr4orKc3qppEROJoV88QQCxCIereRzcD3wx7Hu0GbjCzmwDc/XbgWuCDZpYDRoHr3N0jrklEJFZ2HQ5DYdEsDwV33wx0lc2+vWT5rcCtUdYgIhJ3u3qGSSWMFfMbq12KrmgWEam2XT1DrOpoJpmo/kdy9SsQEZnjdvUMxeJ8AkwzFMzsLDNLh48vN7MPm9m8aEsTEZn9xvMFnu8d4axF1e+OCtM/UrgXyJvZK4F/AM4EvhVZVSIic8Te3hFyBa+tIwWg4O454B3A37j7HwNLoytLRGRuiFN3VJh+KIyb2XuA64HvhfNS0ZQkIjJ3TITC6hhczQzTD4UbgEuAz7r7c2Z2JvCN6MoSEZkbdh0eZnFbmtaGeHzPntZ1Cu7+NPBhADObD7S6+xeiLExEZC6IU88jmH7vo4fNrM3MFgBbCEY+/etoSxMRmd3cvTZDAWh39wHgncCd7r4OeFN0ZYmIzH49QxkGx3KxGB11wnRDIWlmS4Hf5sUTzSIichp2HR4G4jHm0YTphsKngR8Cu9z9F2a2GtgRXVkiIrNf3LqjwvRPNN8N3F3yfDfwrqiKEhGZC3b1DNFUn2Bple/LXGq6J5qXm9n9ZnbYzA6Z2b1mtjzq4kREZrPgbmsthLeViYXpNh/dCXwHeAWwDPhuOE9ERE7RrsNDsTrJDNMPhU53v9Pdc+H0fwHdF1NE5BSNZvPsOzYaq/MJMP1QOGJm7zOzRDi9D902U0TklO0+Ep+7rZWabijcSNAd9SBwgOA2mjdEVZSIyGy383C8xjyaMK1QcPfn3f3t7t7p7ovc/bcILmQTEZFT8PT+AeoTdaxeWJtHCpP56IxVISIyx2zb1885S1upT8brBpinU018+lCJiNQQd2fbvn7OX9Ze7VJe5nRCwWesChGROWRv7wiDY7lYhsJxr2g2s0Em//A3oDGSikREZrmt+/oBOK/WQsHdWytViIjIXPHUvn7qk3W8anH8PmIjPcNhZvPM7B4ze8bMtpvZJWXLzcy+YmY7zWyrma2Nsh4RkTjY1t3Pq5fE7yQzRBwKwC3AD9z9HOACYHvZ8quANeG0Hrgt4npERKqqUHCe2tfP+cvj13QEEYaCmbUBlwH/AODuWXc/VrbaNcDXPfA4MC+8b4OIyKy0t2+EwUw8TzJDtEcKq4Eeglt3PmlmXzOz8kv3lgEvlDzvDueJiMxKW7uD78ZxPMkM0YZCElgL3OburwOGgU+UrTPZtQ4v6+1kZuvNbIOZbejp6Zn5SkVEKiTOJ5kh2lDoBrrd/efh83sIQqJ8nRUlz5cD+8vfyN3vcPcud+/q7NTgrCJSu7bt6+fVS9tIJeJ3khkiDAV3Pwi8YGZnh7OuBJ4uW+07wPvDXkgXA/3ufiCqmkREqik4yTzA+cvaql3KlKZ1O87TcDPwTTOrB3YDN5jZTQDufjvwAHA1sBMYQSOvisgstqd3mKFMjtcum1ftUqYUaSi4+2agq2z27SXLHfhQlDWIiMTFthhfyTwhno1aIiKz0LbuftLJOtYsjtdw2aUUCiIiFRL3k8ygUBARqYhCwfnl/oHYXrQ2QaEgIlIBu48MMRTjK5knKBRERCrgpzt7Afi11QuqXMnxKRRERCrg0R09nLGgiZUd5aP9xItCQUQkYtlcgZ/t6uWyVy2sdiknpFAQEYnYpuePMpzN88Y18R+mR6EgIhKxR3f0kKgzLjmro9qlnJBCQUQkYo/uOMLaM+bR1pCqdiknpFAQEYlQ33CWbfv6a6LpCBQKIiKR+s+dR3CHN66J/0lmUCiIiETq0Wd7aG9M8drl8R0ZtZRCQUQkIu7OIzt6+C+vXEiibrIbTcaPQkFEJCI7Dg9xaCBTM01HoFAQEYnMI88G95R/46tq4yQzKBRERCLzyI4jnNXZzLJ5jdUuZdoUCiIiERjK5Pj57t6a6Yo6QaEgIhKBB7cdIJMr8JsXLK12KSdFoSAiEoF7NnZz5sJm1p4xv9qlnBSFgojIDHuhb4SfP9fHO1+3DLPa6Io6QaEgIjLD7tu0DzN457rl1S7lpCkURERmkLtz76ZuLlndUVO9jiYoFEREZtAv9hzl+b4Rrq3BowRQKIiIzKh7N3bTXJ/grectqXYppyQZ5Zub2R5gEMgDOXfvKlt+OfBt4Llw1n3u/ukoaxIRicpoNs/3tx3gqvOX0lQf6cdrZCpR9RXufuQ4yx9197dVoA4RkUj98JcHGcrkarbpCNR8JCIyY/51wwssn9/IRasWVLuUUxZ1KDjwkJltNLP1U6xziZltMbMHzezcyVYws/VmtsHMNvT09ERXrYjIKdr8wjEe29XL+y5eSV2NDJM9maibjy519/1mtgj4kZk94+6PlCzfBKx09yEzuxr4N2BN+Zu4+x3AHQBdXV0ecc0iIift1n/fybymFO+7eGW1SzktkR4puPv+8Odh4H7gorLlA+4+FD5+AEiZWe0MPC4iAjy9f4Afbz/EjZeeSUu6Nk8wT4gsFMys2cxaJx4DbwGeKltniYXXgJvZRWE9vVHVJCIShf/z8E5a00muf8Oqapdy2qKMtMXA/eFnfhL4lrv/wMxuAnD324FrgQ+aWQ4YBa5zdzUPiUjN2Hl4iAe2HeAPLj+L9sZUtcs5bZGFgrvvBi6YZP7tJY9vBW6NqgYRkaj93cM7aUgmuPHSM6tdyoxQl1QRkVP0fO8I3968n/f+2hl0tKSrXc6MUCiIiJyiW36yg0Sd8XuXra52KTNGoSAicgo27Onj3k3d3HDpKha3NVS7nBmjUBAROUm5fIE//fYvWdrewId//WWXVtU0hYKIyEn6xuN72X5ggD9922torvHrEsopFERETkLPYIYvPfQsb1yzkKtqdHjs41EoiIichM8/uJ2xXJ4/f/u5NXf/5elQKIiITNPPdvVy36Z9rL9sNas7W6pdTiQUCiIi03B0OMtH/3Uzqzqa+NAVr6x2OZGZXWdIREQi4O587O4t9A5lue8P3lCzd1WbDh0piIicwD/+dA8/eeYwf3L1OZy3rL3a5URKoSAichxbu4/xhQe386ZXL+YDs2AU1BNRKIiITKF/dJyb73qShS1pvnjta2dlb6Nys7dhTETkNGRzBT74jY3sOzrKXesvZn5zfbVLqgiFgohIGXfnf96/jcd29fJX776A169aUO2SKkbNRyIiZW79953cs7GbD1+5hmvXLa92ORWlUBARKfHtzfv40o+e5R2vW8Yfv2l2DXY3HQoFEZHQ97ce4GN3b+GiMxfwhXedPydOLJdTKIiIAP/yxPPcfNcmLlwxj69d30U6mah2SVWhE80iMuf9/X/s4vMPPsPlZ3dy23vX0Vg/NwMBFAoiMsd9+UfPcstPdvC21y7lr3/7QuqTc7sBRaEgInPWfZu6ueUnO3j3uuV84V2vJVE3984hlJvbkSgic9aWF47xifu2cfHqBXzunecrEEIKBRGZcw4PjLH+nzewqDXN3713HamEPgonRLonzGyPmW0zs81mtmGS5WZmXzGznWa21czWRlmPiEgml+f3v7GRgdEcX31/FwvmyPAV01WJcwpXuPuRKZZdBawJp18Dbgt/iohE4n9/bztPPn+M2967llcvbat2ObFT7WOma4Cve+BxYJ6ZLa1yTSIySz226wj//Phebrz0TK46Xx81k4k6FBx4yMw2mtn6SZYvA14oed4dznsJM1tvZhvMbENPT09EpYrIbDaSzfGJe7exsqOJj//G2dUuJ7aiDoVL3X0tQTPRh8zssrLlk53u95fNcL/D3bvcvauzszOKOkVklvurHz7L830j/MW7XjunL047kUhDwd33hz8PA/cDF5Wt0g2sKHm+HNgfZU0iMvds3NvHnY89x3+7eCUXr+6odjmxFlkomFmzmbVOPAbeAjxVttp3gPeHvZAuBvrd/UAU9TzfO8JXH9nN47t7GcrkotiEiMTQ2Hiej9+zlVe0N/I/rjqn2uXEXpS9jxYD94ejDCaBb7n7D8zsJgB3vx14ALga2AmMADdEVczG5/v47APbATCD1QubWdXRTCZXYCSbY3S8AEBTfaI4NaeTtKaTtDQkaUmnWNrewPL5jSyb38ii1gZd7CISc6PZPDff9SS7e4b5+o0X0ZLWIA4nYu4va8KPta6uLt+w4WWXPEzLkaEM2/b1s627n63d/ew/NkpjfYLGVIKGVAJwRrL5cMoxnMkzlMkxlMmRL7x0PyXqjLaGJO2NKdoaUzSmEpjBxO5MJozGVILG+iSNqToSdUahAI5TcCi4Q/AfE/8PJrZQcBjPFRjPF8jmCzTVJzhzYQurFzZzZmczqUQd/aPjDIyOMzA2TqGkNgcy4wXGxvNkcsHP0Ykpm2c879RZEIx1ZiQTddQn6kin6kgn61jYkmZxWwNL2hpY3JamOZ2kqT7YP+lk3ZwcSlhqU+9Qht/9pw1s6T7Gn/3muVz/hlXVLqmqzGyju3edaL05FZsLW9JccfYirjh70Um9zj0IiwP9Y3QfHWHfsVEOHBvj2GiWgdEc/aPjjGRzmBsE/zE2XuDo8Dij40HAuL/4QWyAmWHhh7MRPAaKy+oTdaSSRipRx4H+Mf79mcOM508uwOsMGlKJMJyCn8lEHe6Oh8GULziZXCGYxvMMHqdprc6gqT5JY3gklawzcgUnl3ey+QJ1BvXJIGTqkwla00naGlPMa0rR2pAkURIoefdiaGVyBXBoSSdpbQiOzJrrk6RTdTQkE8HPVLDNid+lqT4Iq8b6BE3h7yUyYc+RYa6/8wkO9o9x23vX8dbzllS7pJoxp0LhVJkZzekkr1zUwisXtVSlhly+wL5jo+w+Mkyh4LQ1poKjlIbUS5qxzCCdrCOdTJBK2El/sx8bz3NoYIwD/WMcHswwksmFwRYcaYxk84yO5xjJ5snlnWQiCK5UIjgSyuYLZHMFMrk8g2M5uo+O8PT+cQbGcpQelZoZDamgzoZUHQ4MjQVHZSPZ/Envn9aGJPOb6pnfXE9bQ5KG8OivMVUXHAnWJ2hKJWmsr6OubJ/Ma6qno7mejpZ65jfV09qQpDmd1NAHNernu3v54Dc34e586/cuZt3K+dUuqaYoFGpEMlHHyo5mVnY0R7qdhlSiIts5nly+wGh4BFFsAsvmGRvPF5v3JoJpJGzi6x8d5+hIlqMjQbNaz2CGTK7AaPbF5rNsrnBSdaSTdTSnk6STdcXms7aGFAvCAOlorqe1IUVTOkFLOklTfZLO1jSL29IsbEkrVCrM3fn6z/byme89zYoFTfzD9V2s7qzOl7haplCQ2Ekm6mhN1NE6w++byxcYyxWC8zkETXWFAvSPjnNkOEPvUJajw9nieaThTI7hbI6x8RfDaWB0nF09QzyxJ8vRkSxTnZIzgwVN9cWmsJZ0kuZ0gpaGFC3pJC3pBAua07xiXgPL5gWdFxQkp25sPM8n73+Kezd1c+U5i/jydRfS1pCqdlk1SaEgc0YyUUfLJB+67U0pzuhoOun3yxec4WyO0Wye4TBIegYzHBrIcGhgjJ6hTBAsmaDTwpGhLHt7RxjM5BgcG2ds/OVHLvOaUnQ017OwJc0ZC5o4s7OZMzuCDgYrFzTroqsy7s5Pd/byuQe28/SBAf7oyjX80ZVrqFPPwFOmUBA5RUEPtNQpfyMdyuQ4cGyUfeF0ZDDLkaEMvcMZegYz/MezPdy9sfslr1na3sDKjiZWdTSzbF4jrwiPMlYsaGJpW8Oc+TB0dx7b1cvf/PhZfrHnKEvbG/jq+7t482sWV7u0mqdQEKmSlnSSNYtbWbN46oayoUyOPUeG2X1kmD3h9FzvMD/efogjQ9mXrNuYSrBqYTOrO5uZ35QiWVdHss5IJevoaK5nUVsDi1rTLGpN09Gcpq0xWXNdjI8MZfj+1gPc9+Q+trxwjCVtDXzmmnP57devIJ3UUdRMUCiIxFhLOsl5y9o5b1n7y5aNjefZHx5l7O0dYXfPMLuPDPHUvn4Gx3KM5wvk8h78LLz85EeyzpjXVM/ClnqWtjewdF4jr2hvYFFbAx3NQU+ujuZ6FjTX05KuToDk8gWeOTjIpueP8uPth/npziPkC845S1r59DXn8jsKgxmnUBCpUQ2pBKs7W1jd2cIb10y9nrszMJrj8GDQzfjw4Bi9Q1n6hoOpZzDDgf4xtnT30zecnfQ96hN1zG9OMb8pON/R2ZpmYUs9HS1p2hpSxetLWtLJ4rUl6WTQHbilIRle3Dl5qBQKzrGwx9jzfSPs7R1mb+8Izx4aZGt3P6PjQRflFQsauem/rubtFyzj7CUz3Q1BJigURGY5M6O9KUV7U+q4TVUQHH30DGboHc7SF/bIOjYyTt9I0DOrdzg477F37zCHB4Juv9NRZ8FRTzqVIFUXXEmfrDMGxsbpG85SfiDT1pBkdWcLv/P6FaxdOZ+1Z8xj2bzGmmvuqkUKBREpakglWLGgiRULTtwby90ZzuYZGgt6Uw2GPa0yYRfeTC64pmSiZ9bgWI5MOHxLLl9gvOC0NaSKFw52hD2uVnU0Ma9Jt8isFoWCiJwSMwuvuUiypL2h2uXIDNGVMiIiUqRQEBGRIoWCiIgUKRRERKRIoSAiIkUKBRERKVIoiIhIkUJBRESKzKe6S0hMmVk/sGOSRe1A/3HmlS+feD7ZOguBI6dQ3mQ1THedqeqb7Plkj0+39uPVdqLl2vcvzouq/lPd9+XPo/i3U819X/p4Lu7749VXvnylu3ee8N2Cm7jXzgTcMd35pfPKl088n2wdYMNM1nYq9R/v+RQ1n1bt06lf+/74+z7K+k9131fi3041930l6o/zvj+d+qeaarH56LsnMf+7x1n+3Wmsc7Km8/rp1n+855M9Pt3ap/Me2ve1t+/Ln0dRfzX3/XS3fzy1vO+n8x4ntY2aaz6qBDPb4O5d1a7jVNRy7aD6q6mWa4farj9OtdfikUIl3FHtAk5DLdcOqr+aarl2qO36Y1O7jhRERKRIRwoiIlI060PBzP7RzA6b2VOn8Np1ZrbNzHaa2Ves5LZPZnazmf3KzH5pZn85s1UXtzHjtZvZn5nZPjPbHE5Xz3zlxRoi2ffh8o+ZmZvZwpmr+CXvH8W+/4yZbQ33+0Nm9oqZr7xYQxT1f9HMngl/h/vNbN7MVx5Z7e8O/1YLZhZJ2/3p1D3F+11vZjvC6fqS+cf92zhtp9oNqlYm4DJgLfDUKbz2CeASwIAHgavC+VcAPwbS4fNFNVT7nwEfq9V9Hy5bAfwQ2AssrJXagbaSdT4M3F5L+x54C5AMH/8F8Bc1VPurgbOBh4GuONUd1rSqbN4CYHf4c374eP7xfseZmmb9kYK7PwL0lc4zs7PM7AdmttHMHjWzc8pfZ2ZLCf6If+bB/4mvA78VLv4g8AV3z4TbOFxDtVdMhPV/GfjvQGQnxKKo3d0HSlZtrsH6H3L3XLjq48DyGqp9u7v/Kop6T7fuKfwG8CN373P3o8CPgLdW4m971ofCFO4Abnb3dcDHgL+bZJ1lQHfJ8+5wHsCrgDea2c/N7D/M7PWRVvtSp1s7wB+GTQD/aGbzoyt1UqdVv5m9Hdjn7luiLnQSp73vzeyzZvYC8F7gUxHWOpmZ+Lcz4UaCb6mVMpO1V9J06p7MMuCFkucTv0vkv+Ocu0ezmbUAbwDuLmmKS0+26iTzJr7ZJQkO6S4GXg/8q5mtDpM7MjNU+23AZ8LnnwG+RPAHHrnTrd/MmoBPEjRjVNQM7Xvc/ZPAJ83sT4A/BP7XDJc6qZmqP3yvTwI54JszWeNUZrL2Sjpe3WZ2A/BH4bxXAg+YWRZ4zt3fwdS/S+S/45wLBYKjo2PufmHpTDNLABvDp98h+PAsPTxeDuwPH3cD94Uh8ISZFQjGLumJsnBmoHZ3P1Tyuq8C34uy4DKnW/9ZwJnAlvCPbDmwycwucveDMa+93LeA71OhUGCG6g9PeL4NuDLqL0ElZnrfV8qkdQO4+53AnQBm9jDwAXffU7JKN3B5yfPlBOceuon6d4zihEvcJmAVJSd/gMeAd4ePDbhgitf9guBoYOKEztXh/JuAT4ePX0VwmGc1UvvSknX+GPiXWtr3ZevsIaITzRHt+zUl69wM3FNL+x54K/A00Bll3VH+uyHCE82nWjdTn2h+jqBFYn74eMF0fsfT/h2i/p9b7Qm4CzgAjBOk7O8SfNv8AbAl/Ef+qSle2wU8BewCbuXFi/3qgW+EyzYBv15Dtf8zsA3YSvDtamkUtUdVf9k6e4iu91EU+/7ecP5WgvFoltXSvgd2EnwB2hxOkfSeiqj2d4TvlQEOAT+MS91MEgrh/BvDfb4TuOFk/jZOZ9IVzSIiUjRXex+JiMgkFAoiIlKkUBARkSKFgoiIFCkURESkSKEgs4KZDVV4e18zs9fM0HvlLRg59Skz++6JRh81s3n01aZKAAAC5klEQVRm9gczsW2RcuqSKrOCmQ25e8sMvl/SXxz8LVKltZvZPwHPuvtnj7P+KuB77n5eJeqTuUVHCjJrmVmnmd1rZr8Ip0vD+ReZ2WNm9mT48+xw/gfM7G4z+y7wkJldbmYPm9k9FtxH4JsTY9eH87vCx0PhQHdbzOxxM1sczj8rfP4LM/v0NI9mfsaLg/+1mNlPzGyTBePnXxOu8wXgrPDo4ovhuh8Pt7PVzP58BnejzDEKBZnNbgG+7O6vB94FfC2c/wxwmbu/jmCk0s+VvOYS4Hp3//Xw+euAjwCvAVYDl06ynWbgcXe/AHgE+L2S7d8Sbv+E49OEY/lcSXClOcAY8A53X0twD48vhaH0CWCXu1/o7h83s7cAa4CLgAuBdWZ22Ym2JzKZuTggnswdbwJeUzJCZZuZtQLtwD+Z2RqCESZTJa/5kbuXjon/hLt3A5jZZoKxbf6zbDtZXhxYcCPw5vDxJbw41v23gL+aos7GkvfeSDB2PgRj23wu/IAvEBxBLJ7k9W8JpyfD5y0EIfHIFNsTmZJCQWazOuASdx8tnWlmfwv8P3d/R9g+/3DJ4uGy98iUPM4z+d/MuL94cm6qdY5n1N0vNLN2gnD5EPAVgnsudALr3H3czPYADZO83oDPu/vfn+R2RV5GzUcymz1EcM8CAMxsYgjjdmBf+PgDEW7/cYJmK4DrTrSyu/cT3KbzY2aWIqjzcBgIVwArw1UHgdaSl/4QuDEcvx8zW2Zmi2bod5A5RqEgs0WTmXWXTB8l+IDtCk++Pk0w5DnAXwKfN7OfAokIa/oI8FEzewJYCvSf6AXu/iTBiJrXEdzEpsvMNhAcNTwTrtML/DTswvpFd3+IoHnqZ2a2DbiHl4aGyLSpS6pIRMI7xY26u5vZdcB73P2aE71OpJp0TkEkOuuAW8MeQ8eo0G1PRU6HjhRERKRI5xRERKRIoSAiIkUKBRERKVIoiIhIkUJBRESKFAoiIlL0/wFqEIk97YVI/AAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "lr_finder.plot(push_to_tensorboard=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can use the `skip_start` and `skip_end` parameters in order to zoom in on the area around the minimum value."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "lr_finder.plot(skip_start=20, skip_end=20, push_to_tensorboard=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In general, you want to pick to pick a learning rate (or a max learning rate, if you're using a learning rate scheduler) that is as high as possible while still being reasonably far away from the minimum value. Ideally, you would want to pick the point that has the greatest rate of change as well. For this example, it looks like 1e-3 would be a good learning rate."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now you can push the plot to tensorboard under the `Images` section."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "lr_finder.plot(skip_start=20, skip_end=20)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Environment (conda_pytorch_p36)",
   "language": "python",
   "name": "conda_pytorch_p36"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}