examples/pannuke_cellpose_benchmark.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# In this notebook, we will demo how to benchmark a trained model\n",
    "\n",
    "We will be using [fvcore](https://github.com/facebookresearch/fvcore) package for computing the FLOPS and [pandas](https://pandas.pydata.org/docs/) to display the results nicely."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# !pip install cellseg-models-pytorch[all]\n",
    "# !pip install pandas\n",
    "# !pip install -U fvcore"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First we need to run the inference on the test set! Depending on the dataset, you need to decide whether to use the `SlidingWindowInferer` or `ResizeInferer`. Here, we have trained a cellpose model (without any bells and whistles) for 20 epochs on the [Pannuke](https://arxiv.org/abs/2003.10778) folds 2 and 3 and are going to use fold 1 as the test set. Pannuke dataset contains 256x256 patches so we are going to use the `ResizeInferer`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from cellseg_models_pytorch.models import get_model\n",
    "from cellseg_models_pytorch.inference import ResizeInferer\n",
    "\n",
    "# the pannuke classes\n",
    "type_classes = {\n",
    "    \"background\": 0,\n",
    "    \"neoplastic\": 1,\n",
    "    \"inflammatory\": 2,\n",
    "    \"connective\": 3,\n",
    "    \"dead\": 4,\n",
    "    \"epithelial\": 5,\n",
    "}\n",
    "\n",
    "# Initializee the model and input and ckpt paths\n",
    "model = get_model(\"cellpose\", type=\"base\", ntypes=len(type_classes))\n",
    "input_images = \"/path/to/pannuke/fold1/images/\"  # modify this path\n",
    "ckpt_path = \"/path/to/fold23_weights.ckpt\"  # modify this path\n",
    "\n",
    "# Initialize the output head args\n",
    "out_activations = {\"type\": \"softmax\", \"cellpose\": None}\n",
    "out_boundary_weights = {\"type\": False, \"cellpose\": False}\n",
    "\n",
    "# Initialize the inferer\n",
    "inferer = ResizeInferer(\n",
    "    model=model,\n",
    "    input_path=input_images,\n",
    "    checkpoint_path=ckpt_path,\n",
    "    out_activations=out_activations,\n",
    "    out_boundary_weights=out_boundary_weights,\n",
    "    resize=(256, 256),  # not actually resizing anything\n",
    "    instance_postproc=\"cellpose\",\n",
    "    normalization=\"minmax\",  # minmax normalization was used for training\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Running inference: 100%|██████████| 332/332 [02:47<00:00,  1.98batch/s, post-processing]\n"
     ]
    }
   ],
   "source": [
    "# Run the inference\n",
    "inferer.infer(\n",
    "    save_intermediate=True\n",
    ")  # save_intermediate needs to be set to True for latency benchmarks."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Let's plot a few result segmentation masks before benchmarking\n",
    "\n",
    "Some of the results look reasonable some maybe less so. This indicates that we should have trained longer and maybe should have used different sampling strategies or other bells and whistles."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7f80054a9130>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x1000 with 16 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from skimage.color import label2rgb\n",
    "\n",
    "samples = list(inferer.out_masks.keys())\n",
    "randints = list(np.random.randint(low=0, high=len(samples), size=8))\n",
    "keys = [samples[i] for i in randints]\n",
    "\n",
    "fig, ax = plt.subplots(4, 4, figsize=(10, 10))\n",
    "ax = ax.flatten()\n",
    "ax[0].imshow(label2rgb(inferer.out_masks[keys[0]][\"inst\"], bg_label=0))\n",
    "ax[1].imshow(label2rgb(inferer.out_masks[keys[0]][\"type\"], bg_label=0))\n",
    "ax[2].imshow(label2rgb(inferer.out_masks[keys[1]][\"inst\"], bg_label=0))\n",
    "ax[3].imshow(label2rgb(inferer.out_masks[keys[1]][\"type\"], bg_label=0))\n",
    "ax[4].imshow(label2rgb(inferer.out_masks[keys[2]][\"inst\"], bg_label=0))\n",
    "ax[5].imshow(label2rgb(inferer.out_masks[keys[2]][\"type\"], bg_label=0))\n",
    "ax[6].imshow(label2rgb(inferer.out_masks[keys[3]][\"inst\"], bg_label=0))\n",
    "ax[7].imshow(label2rgb(inferer.out_masks[keys[3]][\"type\"], bg_label=0))\n",
    "ax[8].imshow(label2rgb(inferer.out_masks[keys[4]][\"inst\"], bg_label=0))\n",
    "ax[9].imshow(label2rgb(inferer.out_masks[keys[4]][\"type\"], bg_label=0))\n",
    "ax[10].imshow(label2rgb(inferer.out_masks[keys[5]][\"inst\"], bg_label=0))\n",
    "ax[11].imshow(label2rgb(inferer.out_masks[keys[5]][\"type\"], bg_label=0))\n",
    "ax[12].imshow(label2rgb(inferer.out_masks[keys[6]][\"inst\"], bg_label=0))\n",
    "ax[13].imshow(label2rgb(inferer.out_masks[keys[6]][\"type\"], bg_label=0))\n",
    "ax[14].imshow(label2rgb(inferer.out_masks[keys[7]][\"inst\"], bg_label=0))\n",
    "ax[15].imshow(label2rgb(inferer.out_masks[keys[7]][\"type\"], bg_label=0))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Segmentation Performance Benchmarks\n",
    "\n",
    "Next, we are going to benchmark the segmentation results with the panoptic quality metric (PQ).\n",
    "We are going to compute the metric per each cell type class. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Running neoplastic inst benchmark: 100%|██████████| 266/266 [00:13<00:00, 20.05it/s]\n",
      "Running inflammatory inst benchmark: 100%|██████████| 266/266 [00:10<00:00, 25.65it/s]\n",
      "Running connective inst benchmark: 100%|██████████| 266/266 [00:11<00:00, 22.77it/s]\n",
      "Running dead inst benchmark: 100%|██████████| 266/266 [00:04<00:00, 53.21it/s]\n",
      "Running epithelial inst benchmark: 100%|██████████| 266/266 [00:08<00:00, 32.88it/s]\n"
     ]
    }
   ],
   "source": [
    "from cellseg_models_pytorch.utils.seg_benchmark import SegBenchmarker\n",
    "\n",
    "in_dir = \"/path/to/pannuke/fold1/labels\"  # modify this path\n",
    "bm = SegBenchmarker(\n",
    "    true_path=in_dir,\n",
    "    inferer=inferer,\n",
    "    type_classes=type_classes,\n",
    ")\n",
    "\n",
    "benchmark_res = bm.run_inst_benchmark(how=\"multi\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Sneak peak on the metrics result table\n",
    "\n",
    "Some of the metrics are clamped to -1.0. This means that neither the ground truth labels or the resulting segmentation masks contained the corresponding class. These are filtered out in the next steps."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "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>pq</th>\n",
       "      <th>sq</th>\n",
       "      <th>dq</th>\n",
       "      <th>name</th>\n",
       "      <th>type</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.629760</td>\n",
       "      <td>0.845205</td>\n",
       "      <td>0.745098</td>\n",
       "      <td>Adrenal_gland_fold1_0</td>\n",
       "      <td>neoplastic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.704438</td>\n",
       "      <td>0.858534</td>\n",
       "      <td>0.820513</td>\n",
       "      <td>Adrenal_gland_fold1_1</td>\n",
       "      <td>neoplastic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.531160</td>\n",
       "      <td>0.849855</td>\n",
       "      <td>0.625000</td>\n",
       "      <td>Adrenal_gland_fold1_10</td>\n",
       "      <td>neoplastic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.780535</td>\n",
       "      <td>0.902494</td>\n",
       "      <td>0.864865</td>\n",
       "      <td>Adrenal_gland_fold1_100</td>\n",
       "      <td>neoplastic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.000000</td>\n",
       "      <td>-1.000000</td>\n",
       "      <td>-1.000000</td>\n",
       "      <td>Adrenal_gland_fold1_101</td>\n",
       "      <td>neoplastic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13275</th>\n",
       "      <td>-1.000000</td>\n",
       "      <td>-1.000000</td>\n",
       "      <td>-1.000000</td>\n",
       "      <td>Uterus_fold1_5</td>\n",
       "      <td>epithelial</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13276</th>\n",
       "      <td>-1.000000</td>\n",
       "      <td>-1.000000</td>\n",
       "      <td>-1.000000</td>\n",
       "      <td>Uterus_fold1_6</td>\n",
       "      <td>epithelial</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13277</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>Uterus_fold1_7</td>\n",
       "      <td>epithelial</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13278</th>\n",
       "      <td>-1.000000</td>\n",
       "      <td>-1.000000</td>\n",
       "      <td>-1.000000</td>\n",
       "      <td>Uterus_fold1_8</td>\n",
       "      <td>epithelial</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13279</th>\n",
       "      <td>-1.000000</td>\n",
       "      <td>-1.000000</td>\n",
       "      <td>-1.000000</td>\n",
       "      <td>Uterus_fold1_9</td>\n",
       "      <td>epithelial</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>13280 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             pq        sq        dq                     name        type\n",
       "0      0.629760  0.845205  0.745098    Adrenal_gland_fold1_0  neoplastic\n",
       "1      0.704438  0.858534  0.820513    Adrenal_gland_fold1_1  neoplastic\n",
       "2      0.531160  0.849855  0.625000   Adrenal_gland_fold1_10  neoplastic\n",
       "3      0.780535  0.902494  0.864865  Adrenal_gland_fold1_100  neoplastic\n",
       "4     -1.000000 -1.000000 -1.000000  Adrenal_gland_fold1_101  neoplastic\n",
       "...         ...       ...       ...                      ...         ...\n",
       "13275 -1.000000 -1.000000 -1.000000           Uterus_fold1_5  epithelial\n",
       "13276 -1.000000 -1.000000 -1.000000           Uterus_fold1_6  epithelial\n",
       "13277  0.000000  0.000000  0.000000           Uterus_fold1_7  epithelial\n",
       "13278 -1.000000 -1.000000 -1.000000           Uterus_fold1_8  epithelial\n",
       "13279 -1.000000 -1.000000 -1.000000           Uterus_fold1_9  epithelial\n",
       "\n",
       "[13280 rows x 5 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "res_df = pd.DataFrame(benchmark_res)\n",
    "res_df"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, we'll compute the mean PQ over the per-class PQs. As you can see, it's not very good."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "pq      0.354983\n",
       "sq      0.550885\n",
       "dq      0.441920\n",
       "name         NaN\n",
       "type         NaN\n",
       "dtype: float64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# This is the mPQ over all per-class PQs.\n",
    "res_df[res_df[[\"pq\", \"dq\", \"sq\"]] >= 0].mean(axis=0)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The Pannuke dataset contains lot of images that contain only a few cells. The models can easily fail to detect any of these which results in PQ=0. If we also filter out examples with PQ=0 we get a drastic increase."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "pq      0.513649\n",
       "sq      0.797112\n",
       "dq      0.639444\n",
       "name         NaN\n",
       "type         NaN\n",
       "dtype: float64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# If we filter out also the examples that got\n",
    "res_df[res_df[[\"pq\", \"dq\", \"sq\"]] > 0].mean(axis=0)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can also look at the mean PQ per cell-type to get a clearer picture of the hard to detect classes. Turns out that the dead cell class is very hard to detect. Reason for this is that there are very few examples of dead cells in the training set.. All in all, the results are far from being SOTA so there is still a lot of proper training to do to get a matching performance. On the bright side, the model learned a decent mapping with only a handful of training epochs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "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>pq</th>\n",
       "      <th>sq</th>\n",
       "      <th>dq</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>type</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>connective</th>\n",
       "      <td>0.315990</td>\n",
       "      <td>0.543382</td>\n",
       "      <td>0.408232</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dead</th>\n",
       "      <td>0.009412</td>\n",
       "      <td>0.081074</td>\n",
       "      <td>0.014467</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>epithelial</th>\n",
       "      <td>0.370907</td>\n",
       "      <td>0.549539</td>\n",
       "      <td>0.468822</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>inflammatory</th>\n",
       "      <td>0.306563</td>\n",
       "      <td>0.467061</td>\n",
       "      <td>0.377805</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>neoplastic</th>\n",
       "      <td>0.468043</td>\n",
       "      <td>0.668050</td>\n",
       "      <td>0.563139</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    pq        sq        dq\n",
       "type                                      \n",
       "connective    0.315990  0.543382  0.408232\n",
       "dead          0.009412  0.081074  0.014467\n",
       "epithelial    0.370907  0.549539  0.468822\n",
       "inflammatory  0.306563  0.467061  0.377805\n",
       "neoplastic    0.468043  0.668050  0.563139"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "res_df.loc[res_df[\"pq\"] >= 0].groupby(\"type\").mean(numeric_only=True)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Latency and throughput Benchmarks\n",
    "\n",
    "Next we will benchmark the model latency. These models typically have an NN backbone that spits out intermediate outputs that need to be post-processed. Thus, we will benchmark the latencies and throughputs for the post-processing and NN backbone forward pass separately. This benchmarking was done on a laptop with:\n",
    "\n",
    "- a worn-out NVIDIA GeForce RTX 2080 Mobile (8Gb VRAM)\n",
    "- Intel i7-9750H 6 x 6 cores @ 2.60GHz (32 GiB RAM)\n",
    "\n",
    "**NOTE**:\n",
    "- NN backbone latency = time in seconds for one forward pass of the model\n",
    "- post-processing latency = time in seconds for one pass of the post-processing function."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "from cellseg_models_pytorch.utils.latency_benchmark import LatencyBenchmarker\n",
    "\n",
    "lat_bm = LatencyBenchmarker(inferer=inferer)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First, let's compute the model forward pass mean latency over 100 samples (in seconds).\n",
    "\n",
    "We also report the input size and batch size since those have an effect on the result."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 100/100 [00:07<00:00, 13.85rep/s]\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>batch_size</th>\n",
       "      <th>input_shape</th>\n",
       "      <th>mean latency(s)</th>\n",
       "      <th>std latency(s)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>8</td>\n",
       "      <td>(3, 256, 256)</td>\n",
       "      <td>0.071261</td>\n",
       "      <td>0.000297</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   batch_size    input_shape  mean latency(s)  std latency(s)\n",
       "0           8  (3, 256, 256)         0.071261        0.000297"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_latency = lat_bm.model_latency(reps=100, input_size=(256, 256), batch_size=8)\n",
    "pd.DataFrame(model_latency)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then, let's get the post-processing latencies. This cellpose model's post-processing pipeline contains instance segmentation post-processing and cell type segmentation post-processing.\n",
    "\n",
    "Note that, the post-processing latencies depend on the number of cells in each image, thus, we are going to compute the latencies over the whole pannuke fold3. We are going to run the post-processing 10 times per image and report the mean for each image.\n",
    "\n",
    "**NOTE**: This takes a while to finsh."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 2656/2656 [18:19<00:00,  2.41it/s]\n",
      "100%|██████████| 2656/2656 [02:35<00:00, 17.10it/s]\n"
     ]
    }
   ],
   "source": [
    "inst_post_proc_lat = lat_bm.postproc_latency(\"inst\", reps_per_img=10)\n",
    "inst_post_proc_lat = pd.DataFrame(inst_post_proc_lat)\n",
    "type_post_proc_lat = lat_bm.postproc_latency(\"type\", reps_per_img=10)\n",
    "type_post_proc_lat = pd.DataFrame(type_post_proc_lat)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Sneak peak to latency benchmark tables\n",
    "\n",
    "The Pannuke images are so small (256x256px) that the latencies are small as well. However, for larger images, the post-procesing methods can become a bottleneck. \n",
    "\n",
    "Note that the number of cells and number of pixels non-zero pixels are reported as well since they have an effect on the latency."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "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>repetitions</th>\n",
       "      <th>mean latency(s)</th>\n",
       "      <th>std latency(s)</th>\n",
       "      <th>input_shape</th>\n",
       "      <th>ncells</th>\n",
       "      <th>npixels</th>\n",
       "      <th>name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10</td>\n",
       "      <td>0.033668</td>\n",
       "      <td>0.011313</td>\n",
       "      <td>(256, 256)</td>\n",
       "      <td>30</td>\n",
       "      <td>14036</td>\n",
       "      <td>Adrenal_gland_fold1_0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10</td>\n",
       "      <td>0.029491</td>\n",
       "      <td>0.000736</td>\n",
       "      <td>(256, 256)</td>\n",
       "      <td>22</td>\n",
       "      <td>10700</td>\n",
       "      <td>Adrenal_gland_fold1_1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10</td>\n",
       "      <td>0.031859</td>\n",
       "      <td>0.002062</td>\n",
       "      <td>(256, 256)</td>\n",
       "      <td>26</td>\n",
       "      <td>9238</td>\n",
       "      <td>Adrenal_gland_fold1_10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10</td>\n",
       "      <td>0.033126</td>\n",
       "      <td>0.002427</td>\n",
       "      <td>(256, 256)</td>\n",
       "      <td>21</td>\n",
       "      <td>11581</td>\n",
       "      <td>Adrenal_gland_fold1_100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10</td>\n",
       "      <td>0.030292</td>\n",
       "      <td>0.002052</td>\n",
       "      <td>(256, 256)</td>\n",
       "      <td>5</td>\n",
       "      <td>1773</td>\n",
       "      <td>Adrenal_gland_fold1_101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2651</th>\n",
       "      <td>10</td>\n",
       "      <td>0.024961</td>\n",
       "      <td>0.000670</td>\n",
       "      <td>(256, 256)</td>\n",
       "      <td>8</td>\n",
       "      <td>4200</td>\n",
       "      <td>Uterus_fold1_5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2652</th>\n",
       "      <td>10</td>\n",
       "      <td>0.033751</td>\n",
       "      <td>0.004854</td>\n",
       "      <td>(256, 256)</td>\n",
       "      <td>28</td>\n",
       "      <td>22435</td>\n",
       "      <td>Uterus_fold1_6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2653</th>\n",
       "      <td>10</td>\n",
       "      <td>0.029906</td>\n",
       "      <td>0.002054</td>\n",
       "      <td>(256, 256)</td>\n",
       "      <td>5</td>\n",
       "      <td>1108</td>\n",
       "      <td>Uterus_fold1_7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2654</th>\n",
       "      <td>10</td>\n",
       "      <td>0.032594</td>\n",
       "      <td>0.000991</td>\n",
       "      <td>(256, 256)</td>\n",
       "      <td>31</td>\n",
       "      <td>9882</td>\n",
       "      <td>Uterus_fold1_8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2655</th>\n",
       "      <td>10</td>\n",
       "      <td>0.034824</td>\n",
       "      <td>0.001039</td>\n",
       "      <td>(256, 256)</td>\n",
       "      <td>40</td>\n",
       "      <td>16600</td>\n",
       "      <td>Uterus_fold1_9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2656 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      repetitions  mean latency(s)  std latency(s) input_shape  ncells  \\\n",
       "0              10         0.033668        0.011313  (256, 256)      30   \n",
       "1              10         0.029491        0.000736  (256, 256)      22   \n",
       "2              10         0.031859        0.002062  (256, 256)      26   \n",
       "3              10         0.033126        0.002427  (256, 256)      21   \n",
       "4              10         0.030292        0.002052  (256, 256)       5   \n",
       "...           ...              ...             ...         ...     ...   \n",
       "2651           10         0.024961        0.000670  (256, 256)       8   \n",
       "2652           10         0.033751        0.004854  (256, 256)      28   \n",
       "2653           10         0.029906        0.002054  (256, 256)       5   \n",
       "2654           10         0.032594        0.000991  (256, 256)      31   \n",
       "2655           10         0.034824        0.001039  (256, 256)      40   \n",
       "\n",
       "      npixels                     name  \n",
       "0       14036    Adrenal_gland_fold1_0  \n",
       "1       10700    Adrenal_gland_fold1_1  \n",
       "2        9238   Adrenal_gland_fold1_10  \n",
       "3       11581  Adrenal_gland_fold1_100  \n",
       "4        1773  Adrenal_gland_fold1_101  \n",
       "...       ...                      ...  \n",
       "2651     4200           Uterus_fold1_5  \n",
       "2652    22435           Uterus_fold1_6  \n",
       "2653     1108           Uterus_fold1_7  \n",
       "2654     9882           Uterus_fold1_8  \n",
       "2655    16600           Uterus_fold1_9  \n",
       "\n",
       "[2656 rows x 7 columns]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inst_post_proc_lat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "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>repetitions</th>\n",
       "      <th>mean latency(s)</th>\n",
       "      <th>std latency(s)</th>\n",
       "      <th>img_shape</th>\n",
       "      <th>ncells</th>\n",
       "      <th>npixels</th>\n",
       "      <th>name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10</td>\n",
       "      <td>0.005883</td>\n",
       "      <td>0.001243</td>\n",
       "      <td>(256, 256)</td>\n",
       "      <td>30</td>\n",
       "      <td>14036</td>\n",
       "      <td>Adrenal_gland_fold1_0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10</td>\n",
       "      <td>0.006105</td>\n",
       "      <td>0.001933</td>\n",
       "      <td>(256, 256)</td>\n",
       "      <td>22</td>\n",
       "      <td>10700</td>\n",
       "      <td>Adrenal_gland_fold1_1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10</td>\n",
       "      <td>0.004274</td>\n",
       "      <td>0.000409</td>\n",
       "      <td>(256, 256)</td>\n",
       "      <td>26</td>\n",
       "      <td>9238</td>\n",
       "      <td>Adrenal_gland_fold1_10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10</td>\n",
       "      <td>0.003436</td>\n",
       "      <td>0.000439</td>\n",
       "      <td>(256, 256)</td>\n",
       "      <td>21</td>\n",
       "      <td>11581</td>\n",
       "      <td>Adrenal_gland_fold1_100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10</td>\n",
       "      <td>0.003133</td>\n",
       "      <td>0.000218</td>\n",
       "      <td>(256, 256)</td>\n",
       "      <td>5</td>\n",
       "      <td>1773</td>\n",
       "      <td>Adrenal_gland_fold1_101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2651</th>\n",
       "      <td>10</td>\n",
       "      <td>0.005212</td>\n",
       "      <td>0.000586</td>\n",
       "      <td>(256, 256)</td>\n",
       "      <td>8</td>\n",
       "      <td>4200</td>\n",
       "      <td>Uterus_fold1_5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2652</th>\n",
       "      <td>10</td>\n",
       "      <td>0.007198</td>\n",
       "      <td>0.000778</td>\n",
       "      <td>(256, 256)</td>\n",
       "      <td>28</td>\n",
       "      <td>22435</td>\n",
       "      <td>Uterus_fold1_6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2653</th>\n",
       "      <td>10</td>\n",
       "      <td>0.005758</td>\n",
       "      <td>0.000429</td>\n",
       "      <td>(256, 256)</td>\n",
       "      <td>5</td>\n",
       "      <td>1108</td>\n",
       "      <td>Uterus_fold1_7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2654</th>\n",
       "      <td>10</td>\n",
       "      <td>0.006258</td>\n",
       "      <td>0.000985</td>\n",
       "      <td>(256, 256)</td>\n",
       "      <td>31</td>\n",
       "      <td>9882</td>\n",
       "      <td>Uterus_fold1_8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2655</th>\n",
       "      <td>10</td>\n",
       "      <td>0.008745</td>\n",
       "      <td>0.002565</td>\n",
       "      <td>(256, 256)</td>\n",
       "      <td>40</td>\n",
       "      <td>16600</td>\n",
       "      <td>Uterus_fold1_9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2656 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      repetitions  mean latency(s)  std latency(s)   img_shape  ncells  \\\n",
       "0              10         0.005883        0.001243  (256, 256)      30   \n",
       "1              10         0.006105        0.001933  (256, 256)      22   \n",
       "2              10         0.004274        0.000409  (256, 256)      26   \n",
       "3              10         0.003436        0.000439  (256, 256)      21   \n",
       "4              10         0.003133        0.000218  (256, 256)       5   \n",
       "...           ...              ...             ...         ...     ...   \n",
       "2651           10         0.005212        0.000586  (256, 256)       8   \n",
       "2652           10         0.007198        0.000778  (256, 256)      28   \n",
       "2653           10         0.005758        0.000429  (256, 256)       5   \n",
       "2654           10         0.006258        0.000985  (256, 256)      31   \n",
       "2655           10         0.008745        0.002565  (256, 256)      40   \n",
       "\n",
       "      npixels                     name  \n",
       "0       14036    Adrenal_gland_fold1_0  \n",
       "1       10700    Adrenal_gland_fold1_1  \n",
       "2        9238   Adrenal_gland_fold1_10  \n",
       "3       11581  Adrenal_gland_fold1_100  \n",
       "4        1773  Adrenal_gland_fold1_101  \n",
       "...       ...                      ...  \n",
       "2651     4200           Uterus_fold1_5  \n",
       "2652    22435           Uterus_fold1_6  \n",
       "2653     1108           Uterus_fold1_7  \n",
       "2654     9882           Uterus_fold1_8  \n",
       "2655    16600           Uterus_fold1_9  \n",
       "\n",
       "[2656 rows x 7 columns]"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "type_post_proc_lat"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's compute the means for both post-processing methods."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "repetitions          10.000000\n",
       "mean latency(s)       0.004334\n",
       "std latency(s)        0.000701\n",
       "ncells               22.921687\n",
       "npixels            9801.875753\n",
       "dtype: float64"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "type_post_proc_lat.mean(axis=0, numeric_only=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "repetitions          10.000000\n",
       "mean latency(s)       0.031553\n",
       "std latency(s)        0.003438\n",
       "ncells               22.921687\n",
       "npixels            9801.875753\n",
       "dtype: float64"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inst_post_proc_lat.mean(axis=0, numeric_only=True)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Throughputs\n",
    "\n",
    "In practice, the model (NN-backbone + post-processing + intermediate data-wrangilng) throughput is a lot more valuable metric than the latencies since it directly tells how scalable the model is. The throughput is important information especially when segmenting a large number of huge gigapixel images like WSIs.\n",
    "\n",
    "Next, we will compute the whole model inference + post-processing throughput (img/s)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Running inference: 100%|██████████| 332/332 [02:45<00:00,  2.01batch/s, post-processing]\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>n_images</th>\n",
       "      <th>repetitions</th>\n",
       "      <th>total mean latency(s)</th>\n",
       "      <th>total std latency(s)</th>\n",
       "      <th>mean throughput (img/s)</th>\n",
       "      <th>std throughput (img/s))</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2656</td>\n",
       "      <td>1</td>\n",
       "      <td>165.549179</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.06233</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   n_images  repetitions  total mean latency(s)  total std latency(s)  \\\n",
       "0      2656            1             165.549179                   0.0   \n",
       "\n",
       "   mean throughput (img/s)  std throughput (img/s))  \n",
       "0                  0.06233                      0.0  "
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# this returns the throughput of the inferer.infer() method\n",
    "throughputs = pd.DataFrame(lat_bm.inference_postproc_latency(reps=1))\n",
    "throughputs"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It took 2 min 45 sec to run the inference pipeline on this laptop on the Pannuke fold3 containing 2656 (256x256) images. This averages to ~0.06 seconds per image. Note that the std is zero since we had run the whole pipeline only once."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Compute the model parameters and FLOPS\n",
    "\n",
    "Finally we will compute the NN backbone FLOPS and the number of learnable params with the [fvcore](https://github.com/facebookresearch/fvcore) library. FLOPS are typically reported in research papers, so why not here as well."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "| module                                                                         | #parameters or shape   | #flops     |\n",
      "|:-------------------------------------------------------------------------------|:-----------------------|:-----------|\n",
      "| model                                                                          | 15.119M                | 0.153T     |\n",
      "|  encoder.model                                                                 |  8.543M                |  34.47G    |\n",
      "|   encoder.model.conv1                                                          |   9.408K               |   1.233G   |\n",
      "|    encoder.model.conv1.weight                                                  |    (64, 3, 7, 7)       |            |\n",
      "|   encoder.model.bn1                                                            |   0.128K               |   16.777M  |\n",
      "|    encoder.model.bn1.weight                                                    |    (64,)               |            |\n",
      "|    encoder.model.bn1.bias                                                      |    (64,)               |            |\n",
      "|   encoder.model.layer1                                                         |   0.216M               |   7.072G   |\n",
      "|    encoder.model.layer1.0                                                      |    75.008K             |    2.458G  |\n",
      "|    encoder.model.layer1.1                                                      |    70.4K               |    2.307G  |\n",
      "|    encoder.model.layer1.2                                                      |    70.4K               |    2.307G  |\n",
      "|   encoder.model.layer2                                                         |   1.22M                |   10.802G  |\n",
      "|    encoder.model.layer2.0                                                      |    0.379M              |    3.92G   |\n",
      "|    encoder.model.layer2.1                                                      |    0.28M               |    2.294G  |\n",
      "|    encoder.model.layer2.2                                                      |    0.28M               |    2.294G  |\n",
      "|    encoder.model.layer2.3                                                      |    0.28M               |    2.294G  |\n",
      "|   encoder.model.layer3                                                         |   7.098M               |   15.346G  |\n",
      "|    encoder.model.layer3.0                                                      |    1.512M              |    3.906G  |\n",
      "|    encoder.model.layer3.1                                                      |    1.117M              |    2.288G  |\n",
      "|    encoder.model.layer3.2                                                      |    1.117M              |    2.288G  |\n",
      "|    encoder.model.layer3.3                                                      |    1.117M              |    2.288G  |\n",
      "|    encoder.model.layer3.4                                                      |    1.117M              |    2.288G  |\n",
      "|    encoder.model.layer3.5                                                      |    1.117M              |    2.288G  |\n",
      "|  make_style.downsample.proj                                                    |  0.263M                |  0.538G    |\n",
      "|   make_style.downsample.proj.0.conv                                            |   0.262M               |   0.537G   |\n",
      "|    make_style.downsample.proj.0.conv.weight                                    |    (256, 1024, 1, 1)   |            |\n",
      "|   make_style.downsample.proj.1.norm                                            |   0.512K               |   1.049M   |\n",
      "|    make_style.downsample.proj.1.norm.weight                                    |    (256,)              |            |\n",
      "|    make_style.downsample.proj.1.norm.bias                                      |    (256,)              |            |\n",
      "|  cellpose_decoder                                                              |  6.313M                |  0.117T    |\n",
      "|   cellpose_decoder.decoder_stage1                                              |   5.119M               |   40.317G  |\n",
      "|    cellpose_decoder.decoder_stage1.skip.skip.merge.merge.downsample1           |    0.526M              |    4.312G  |\n",
      "|    cellpose_decoder.decoder_stage1.conv_layers.conv_layer1                     |    4.592M              |    36.005G |\n",
      "|   cellpose_decoder.decoder_stage2.conv_layers.conv_layer1                      |   0.87M                |   25.284G  |\n",
      "|    cellpose_decoder.decoder_stage2.conv_layers.conv_layer1.basic_basic_1       |    0.328M              |    9.681G  |\n",
      "|    cellpose_decoder.decoder_stage2.conv_layers.conv_layer1.basic_basic_2       |    0.181M              |    4.84G   |\n",
      "|    cellpose_decoder.decoder_stage2.conv_layers.conv_layer1.basic_basic_3       |    0.181M              |    4.84G   |\n",
      "|    cellpose_decoder.decoder_stage2.conv_layers.conv_layer1.basic_basic_4.block |    0.148M              |    4.84G   |\n",
      "|    cellpose_decoder.decoder_stage2.conv_layers.conv_layer1.downsample.proj     |    33.024K             |    1.082G  |\n",
      "|   cellpose_decoder.decoder_stage3                                              |   0.251M               |   26.441G  |\n",
      "|    cellpose_decoder.decoder_stage3.skip.skip.merge.merge.downsample1           |    8.448K              |    1.107G  |\n",
      "|    cellpose_decoder.decoder_stage3.conv_layers.conv_layer1                     |    0.243M              |    25.334G |\n",
      "|   cellpose_decoder.decoder_stage4.conv_layers.conv_layer1                      |   73.184K              |   25.434G  |\n",
      "|    cellpose_decoder.decoder_stage4.conv_layers.conv_layer1.basic_basic_1       |    26.784K             |    9.731G  |\n",
      "|    cellpose_decoder.decoder_stage4.conv_layers.conv_layer1.basic_basic_2       |    17.504K             |    4.865G  |\n",
      "|    cellpose_decoder.decoder_stage4.conv_layers.conv_layer1.basic_basic_3       |    17.504K             |    4.865G  |\n",
      "|    cellpose_decoder.decoder_stage4.conv_layers.conv_layer1.basic_basic_4.block |    9.28K               |    4.865G  |\n",
      "|    cellpose_decoder.decoder_stage4.conv_layers.conv_layer1.downsample.proj     |    2.112K              |    1.107G  |\n",
      "|  cellpose_seg_head.head                                                        |  64                    |  33.554M   |\n",
      "|   cellpose_seg_head.head.weight                                                |   (2, 32, 1, 1)        |            |\n",
      "|  type_seg_head.head                                                            |  0.192K                |  0.101G    |\n",
      "|   type_seg_head.head.weight                                                    |   (6, 32, 1, 1)        |            |\n"
     ]
    }
   ],
   "source": [
    "from fvcore.nn import FlopCountAnalysis, ActivationCountAnalysis, flop_count_table\n",
    "import torch\n",
    "\n",
    "inp = torch.rand([8, 3, 256, 256]).to(\"cuda\")\n",
    "flops = FlopCountAnalysis(model, inp)\n",
    "\n",
    "print(flop_count_table(flops))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- The model has 15.9M parameters (Very small)\n",
    "- One forward pass takes 153 GFlops (not too much)\n",
    "\n",
    "Overall, this cellpose model is very lightweight. Adding some width and depth to the layers would probably do some magic (in terms of the segmentation results) without bloating the model too much. So, in conclusion, there is a lot of room for improvement, both on the architecture and training side of the model. However, lightweight models can be more suitable for a large number of huge images if you have limited time and compute resources. \n",
    "\n",
    "This concludes this notebook, thanks!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "cellseg-models-pytorch-O87oC05o-py3.9",
   "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": "3.9.12"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "760f9db57e64852ce823e2734a6ba7e7dce16891b070aba49b38e1ee744f7d23"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}