fig_utils/dino_transfer.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import re\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "fewshot_dataset = ['aricraft', 'caltech101', 'cars', 'cifar10', 'cifar100', 'dtd', 'flowers', 'food', 'pets', 'sun397']\n",
    "\n",
    "def get_head(path):\n",
    "    if 'freezetext' in path or 'freezehead' in path or 'dino-txt' in path:\n",
    "        if 'wit400m' in path or 'dino-txt' in path:\n",
    "            return 'Frozen CLIP'\n",
    "        elif 'roberta' in path:\n",
    "            return 'Frozen Roberta'\n",
    "        else:\n",
    "            raise NotImplementedError\n",
    "    elif 'inittext' in path:\n",
    "        if 'wit400m' in path:\n",
    "            return 'CLIP init'\n",
    "        elif 'roberta' in path:\n",
    "            return 'Roberta init'\n",
    "        else:\n",
    "            raise NotImplementedError\n",
    "    else:\n",
    "        return 'Rand init'\n",
    "\n",
    "def get_num(path, key='sample', type=int, return_str=False):\n",
    "    if type==float:\n",
    "        res = re.findall(r'{}\\d+\\.\\d+'.format(key), path)\n",
    "    else:\n",
    "        res = re.findall(r'{}[0-9]+'.format(key), path)\n",
    "    if len(res) > 0:\n",
    "        val = res[0].replace(key, '')\n",
    "    else:\n",
    "        val = str(-1)\n",
    "    if return_str:\n",
    "        return val\n",
    "    else:\n",
    "        return type(val)\n",
    "\n",
    "def get_metrics(paths, task='fewshot'):\n",
    "    metrics = []\n",
    "    for path in paths:\n",
    "        path = os.path.join(path, '{}.csv'.format(task))\n",
    "        if os.path.exists(path):\n",
    "            df = pd.read_csv(path, sep='\\t')\n",
    "            df['model'] = 'CLIP' if 'open_clip' in path else 'DINO' if 'dino' in path else 'SL'\n",
    "            df['pretrain'] = 'LAIONet' if 'laionet' in path else 'ImageNet'\n",
    "            df['head'] = get_head(path)\n",
    "            subsample = get_num(path, key='samplehead' if df.loc[0, 'model'] == 'SL' else 'sample', type=int)\n",
    "            proto = get_num(path, key='proto', type=int)\n",
    "            proto = 65536 if proto == -1 else proto\n",
    "            df['vocab'] = 1000 if subsample == -1 and df.loc[0, 'model'] == 'SL' else proto if subsample == -1 and df.loc[0, 'model'] == 'DINO' else subsample\n",
    "            df['sample method'] = 'All' if subsample == -1 else 'Uniform' if 'unfm' in path else 'Square Root' if 'sqrt' in path else 'Linear'\n",
    "            if task == 'fewshot' and df[df['dataset'] == 'all'].size > 0:\n",
    "                df['dataset'] = fewshot_dataset[:len(df[df['dataset'] == 'all'])]\n",
    "            metrics.append(df)\n",
    "    metrics = pd.concat(metrics)\n",
    "    metrics = metrics.reset_index(drop=True)\n",
    "    return metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/xwen/miniconda3/lib/python3.10/site-packages/seaborn/_base.py:948: FutureWarning: When grouping with a length-1 list-like, you will need to pass a length-1 tuple to get_group in a future version of pandas. Pass `(name,)` instead of `name` to silence this warning.\n",
      "  data_subset = grouped_data.get_group(pd_key)\n",
      "/Users/xwen/miniconda3/lib/python3.10/site-packages/seaborn/_base.py:948: FutureWarning: When grouping with a length-1 list-like, you will need to pass a length-1 tuple to get_group in a future version of pandas. Pass `(name,)` instead of `name` to silence this warning.\n",
      "  data_subset = grouped_data.get_group(pd_key)\n",
      "/Users/xwen/miniconda3/lib/python3.10/site-packages/seaborn/_base.py:948: FutureWarning: When grouping with a length-1 list-like, you will need to pass a length-1 tuple to get_group in a future version of pandas. Pass `(name,)` instead of `name` to silence this warning.\n",
      "  data_subset = grouped_data.get_group(pd_key)\n",
      "/Users/xwen/miniconda3/lib/python3.10/site-packages/seaborn/_base.py:948: FutureWarning: When grouping with a length-1 list-like, you will need to pass a length-1 tuple to get_group in a future version of pandas. Pass `(name,)` instead of `name` to silence this warning.\n",
      "  data_subset = grouped_data.get_group(pd_key)\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "paths = [\n",
    "    '../exps_dino/output/dino_imagenet_rn50_ep100_bs512/',\n",
    "    '../exps_dino/output/dino_laionet_rn50_ep100_bs1024/',\n",
    "    '../exps_dino/output/dino_laionet_rn50_ep100_bs1024_sample1024/',\n",
    "    '../exps_dino/output/dino_laionet_rn50_ep100_bs1024_sample4096/',\n",
    "    '../exps_dino/output/dino_laionet_rn50_ep100_bs1024_sample16384/',\n",
    "    '../exps_dino/output/dino_laionet_rn50_ep100_bs1024_proto16384/',\n",
    "]\n",
    "\n",
    "df = get_metrics(paths, task='linear')\n",
    "baseline = df[df['pretrain'] == 'ImageNet']\n",
    "df = df[df['pretrain'] != 'ImageNet'].copy()\n",
    "plot_df = df[(df['temp'] == 0) & ((df['sample method'] != 'All') | (df['vocab'] == 65536))].copy()\n",
    "delta_accs = [acc - baseline[baseline['dataset'] == ds]['test_acc'].item() for acc, ds in zip(plot_df['test_acc'].to_list(), plot_df['dataset'].to_list())]\n",
    "plot_df.loc[:, 'delta_acc'] = delta_accs\n",
    "plot_df = plot_df.sort_values(by=['dataset'])\n",
    "\n",
    "fig, ax = plt.subplots(1, 1, figsize=(10, 4))\n",
    "ax.yaxis.grid(color='gray', linestyle='-', linewidth=0.7, alpha=0.3, zorder=-1)\n",
    "ax.axhline(y=0, color='black', linestyle='-', linewidth=1, zorder=4)\n",
    "\n",
    "order = ['birdsnap', 'pets', 'food', 'caltech101', 'voc2007', 'cifar100', 'cifar10', 'dtd', 'sun397', 'flowers', 'aircraft', 'cars'][::-1]\n",
    "datasets = ['Birdsnap', 'OxfordPets', 'Food101', 'Caltech101', 'PascalVOC2007', 'CIFAR100', 'CIFAR10', 'DTD', 'SUN397', 'Flowers102', 'FGVCAircraft', 'StanfordCars'][::-1]\n",
    "positions = [.7] * 10 + [.68, .64]\n",
    "positions[5] = .73\n",
    "positions[7] = .65\n",
    "positions[8] = .72\n",
    "vas = ['bottom'] * 7 + ['top'] + ['bottom'] * 2 + ['top'] * 2\n",
    "has = ['left'] * 7  + ['right'] + ['left'] * 2 + ['right'] * 2\n",
    "shifts = [.15] * 7 + [-.2, .15, .25] + [-.2] * 2\n",
    "\n",
    "positions = positions[::-1]\n",
    "vas = vas[::-1]\n",
    "has = has[::-1]\n",
    "shifts = shifts[::-1]\n",
    "\n",
    "blues = sns.color_palette('Blues', n_colors=5)[1:]\n",
    "reds = sns.color_palette('Oranges', n_colors=5)\n",
    "palette = blues[:-1] + [reds[3]]\n",
    "sns.barplot(plot_df, x='dataset', y='delta_acc', hue='vocab', order=order, ax=ax, palette=palette, alpha=1, saturation=1, ec='w', linewidth=1, zorder=3)\n",
    "legend1 = ax.legend(title='Vocabulary size', alignment='center', loc='lower right', ncol=2, bbox_to_anchor=(1, .015), fontsize=11)\n",
    "\n",
    "delta_accs = [acc - baseline[baseline['dataset'] == ds]['test_acc'].item() for acc, ds in zip(df['test_acc'].to_list(), df['dataset'].to_list())]\n",
    "plot_df = df.copy()\n",
    "plot_df.loc[:, 'delta_acc'] = delta_accs\n",
    "plot_df2 = plot_df[(plot_df['vocab'] == 16384)].copy()\n",
    "plot_df2 = plot_df2.sort_values(by=['dataset'])\n",
    "plot_df2 = plot_df2[(plot_df2['model'] == 'DINO') & (plot_df2['sample method'] == 'All')]\n",
    "\n",
    "ys = [plot_df2[plot_df2['dataset'] == ds]['delta_acc'].item() if ds in plot_df2['dataset'].to_list() else 0 for ds in order]\n",
    "xs = list(np.arange(len(ys)) + .1)\n",
    "# ys[-1] -= .1\n",
    "ax.bar([xs[2]]+xs[5:], [ys[2]]+ys[5:], width=.2, color=reds[2], alpha=1, ec='w', linewidth=1, zorder=2)\n",
    "ax.bar(xs[:2]+xs[3:5], ys[:2]+ys[3:5], width=.2, color=reds[2], alpha=1, ec='w', linewidth=1, zorder=3, label='Vanilla DINO')\n",
    "\n",
    "handles, labels = ax.get_legend_handles_labels()\n",
    "legend2 = plt.legend(handles=[handles[-1], handles[1]], labels=[labels[-1], '+ Sub voc.'], title='Method', loc='lower right', bbox_to_anchor=(.7, .015), fontsize=11)\n",
    "ax.add_artist(legend1)\n",
    "\n",
    "for i, dataset in enumerate(datasets):\n",
    "    ax.text(i+shifts[i], positions[i], dataset, va=vas[i], ha=has[i], transform=ax.get_xaxis_transform(), rotation=45, fontsize=10.5)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(False)\n",
    "ax.set_xlim(xmin=len(datasets)-.55, xmax=-.45)\n",
    "ax.set_xticks([])\n",
    "ax.set_xlabel('')\n",
    "ax.set_ylabel('Linear Probe $\\\\Delta$ Score (%)', fontsize=13)\n",
    "plt.title('DINO variants on LAIONet vs. vanilla DINO on ImageNet', fontsize=13)\n",
    "plt.savefig('figs_new/improved_dino_laionet.pdf', bbox_inches='tight', dpi=300)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_46582/3048884805.py:15: FutureWarning: In future versions `DataFrame.to_latex` is expected to utilise the base implementation of `Styler.to_latex` for formatting and rendering. The arguments signature may therefore change. It is recommended instead to use `DataFrame.style.to_latex` which also contains additional functionality.\n",
      "  df_dataset_as_cols.to_latex('texfiles/tmp.tex')\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>dataset</th>\n",
       "      <th>aircraft</th>\n",
       "      <th>birdsnap</th>\n",
       "      <th>caltech101</th>\n",
       "      <th>cars</th>\n",
       "      <th>cifar10</th>\n",
       "      <th>cifar100</th>\n",
       "      <th>dtd</th>\n",
       "      <th>flowers</th>\n",
       "      <th>food</th>\n",
       "      <th>pets</th>\n",
       "      <th>sun397</th>\n",
       "      <th>voc2007</th>\n",
       "      <th>mean</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>model</th>\n",
       "      <th>pretrain</th>\n",
       "      <th>head</th>\n",
       "      <th>vocab</th>\n",
       "      <th>sample method</th>\n",
       "      <th>temp</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">DINO</th>\n",
       "      <th>ImageNet</th>\n",
       "      <th>Rand init</th>\n",
       "      <th>65536</th>\n",
       "      <th>All</th>\n",
       "      <th>0</th>\n",
       "      <td>27.0</td>\n",
       "      <td>37.1</td>\n",
       "      <td>82.3</td>\n",
       "      <td>23.6</td>\n",
       "      <td>86.4</td>\n",
       "      <td>62.9</td>\n",
       "      <td>68.7</td>\n",
       "      <td>80.8</td>\n",
       "      <td>55.8</td>\n",
       "      <td>66.4</td>\n",
       "      <td>57.0</td>\n",
       "      <td>81.6</td>\n",
       "      <td>60.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"4\" valign=\"top\">LAIONet</th>\n",
       "      <th rowspan=\"4\" valign=\"top\">Rand init</th>\n",
       "      <th>1024</th>\n",
       "      <th>Linear</th>\n",
       "      <th>0</th>\n",
       "      <td>30.8</td>\n",
       "      <td>27.2</td>\n",
       "      <td>78.6</td>\n",
       "      <td>23.8</td>\n",
       "      <td>83.9</td>\n",
       "      <td>61.4</td>\n",
       "      <td>68.1</td>\n",
       "      <td>80.5</td>\n",
       "      <td>50.7</td>\n",
       "      <td>57.7</td>\n",
       "      <td>56.0</td>\n",
       "      <td>77.2</td>\n",
       "      <td>58.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4096</th>\n",
       "      <th>Linear</th>\n",
       "      <th>0</th>\n",
       "      <td>30.3</td>\n",
       "      <td>30.1</td>\n",
       "      <td>78.9</td>\n",
       "      <td>24.8</td>\n",
       "      <td>84.6</td>\n",
       "      <td>63.4</td>\n",
       "      <td>69.5</td>\n",
       "      <td>77.7</td>\n",
       "      <td>53.3</td>\n",
       "      <td>61.0</td>\n",
       "      <td>56.9</td>\n",
       "      <td>78.6</td>\n",
       "      <td>59.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16384</th>\n",
       "      <th>Linear</th>\n",
       "      <th>0</th>\n",
       "      <td>32.2</td>\n",
       "      <td>31.2</td>\n",
       "      <td>79.4</td>\n",
       "      <td>25.2</td>\n",
       "      <td>85.4</td>\n",
       "      <td>63.9</td>\n",
       "      <td>70.2</td>\n",
       "      <td>79.1</td>\n",
       "      <td>54.3</td>\n",
       "      <td>62.2</td>\n",
       "      <td>57.7</td>\n",
       "      <td>79.0</td>\n",
       "      <td>60.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65536</th>\n",
       "      <th>All</th>\n",
       "      <th>0</th>\n",
       "      <td>26.7</td>\n",
       "      <td>24.6</td>\n",
       "      <td>77.8</td>\n",
       "      <td>24.6</td>\n",
       "      <td>83.5</td>\n",
       "      <td>60.0</td>\n",
       "      <td>67.1</td>\n",
       "      <td>79.3</td>\n",
       "      <td>48.6</td>\n",
       "      <td>55.5</td>\n",
       "      <td>55.4</td>\n",
       "      <td>77.3</td>\n",
       "      <td>56.7</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "dataset                                            aircraft  birdsnap  \\\n",
       "model pretrain head      vocab sample method temp                       \n",
       "DINO  ImageNet Rand init 65536 All           0         27.0      37.1   \n",
       "      LAIONet  Rand init 1024  Linear        0         30.8      27.2   \n",
       "                         4096  Linear        0         30.3      30.1   \n",
       "                         16384 Linear        0         32.2      31.2   \n",
       "                         65536 All           0         26.7      24.6   \n",
       "\n",
       "dataset                                            caltech101  cars  cifar10  \\\n",
       "model pretrain head      vocab sample method temp                              \n",
       "DINO  ImageNet Rand init 65536 All           0           82.3  23.6     86.4   \n",
       "      LAIONet  Rand init 1024  Linear        0           78.6  23.8     83.9   \n",
       "                         4096  Linear        0           78.9  24.8     84.6   \n",
       "                         16384 Linear        0           79.4  25.2     85.4   \n",
       "                         65536 All           0           77.8  24.6     83.5   \n",
       "\n",
       "dataset                                            cifar100   dtd  flowers  \\\n",
       "model pretrain head      vocab sample method temp                            \n",
       "DINO  ImageNet Rand init 65536 All           0         62.9  68.7     80.8   \n",
       "      LAIONet  Rand init 1024  Linear        0         61.4  68.1     80.5   \n",
       "                         4096  Linear        0         63.4  69.5     77.7   \n",
       "                         16384 Linear        0         63.9  70.2     79.1   \n",
       "                         65536 All           0         60.0  67.1     79.3   \n",
       "\n",
       "dataset                                            food  pets  sun397  \\\n",
       "model pretrain head      vocab sample method temp                       \n",
       "DINO  ImageNet Rand init 65536 All           0     55.8  66.4    57.0   \n",
       "      LAIONet  Rand init 1024  Linear        0     50.7  57.7    56.0   \n",
       "                         4096  Linear        0     53.3  61.0    56.9   \n",
       "                         16384 Linear        0     54.3  62.2    57.7   \n",
       "                         65536 All           0     48.6  55.5    55.4   \n",
       "\n",
       "dataset                                            voc2007  mean  \n",
       "model pretrain head      vocab sample method temp                 \n",
       "DINO  ImageNet Rand init 65536 All           0        81.6  60.8  \n",
       "      LAIONet  Rand init 1024  Linear        0        77.2  58.0  \n",
       "                         4096  Linear        0        78.6  59.1  \n",
       "                         16384 Linear        0        79.0  60.0  \n",
       "                         65536 All           0        77.3  56.7  "
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "paths = [\n",
    "    '../exps_dino/output/dino_imagenet_rn50_ep100_bs512/',\n",
    "    '../exps_dino/output/dino_laionet_rn50_ep100_bs1024/',\n",
    "    '../exps_dino/output/dino_laionet_rn50_ep100_bs1024_sample1024/',\n",
    "    '../exps_dino/output/dino_laionet_rn50_ep100_bs1024_sample4096/',\n",
    "    '../exps_dino/output/dino_laionet_rn50_ep100_bs1024_sample16384/',\n",
    "]\n",
    "\n",
    "df = get_metrics(paths, task='linear')\n",
    "df_dataset_as_cols = df.pivot(index=['model', 'pretrain', 'head', 'vocab', 'sample method', 'temp'], columns='dataset', values='test_acc')\n",
    "# add a col that averages across all datasets\n",
    "df_dataset_as_cols['mean'] = df_dataset_as_cols.mean(axis=1)\n",
    "# format values to 2 decimal places\n",
    "df_dataset_as_cols = df_dataset_as_cols.round(1)\n",
    "df_dataset_as_cols.to_latex('texfiles/tmp.tex')\n",
    "df_dataset_as_cols"
   ]
  }
 ],
 "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": "3.10.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}