examples/normal_crossing.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Normal Crossing RLCT Estimation\n",
    "\n",
    "[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/timaeus-research/devinterp/blob/main/examples/normal_crossing.ipynb)\n",
    "\n",
    "This notebook estimates the Local Learning Coefficient (LLC) of normal crossings using two algorithms: SGNHT (Stochastic Gradient Nose-Hoover Thermostat) and SGLD (Stochastic Gradient Langevin Dynamics). We recreate some key results from  [Lau et al. (2023)](https://arxiv.org/pdf/2308.12108.pdf), namely the ordinality check, trajectory plot, and the table of theoretical vs empirical learning coefficients. \n",
    "\n",
    "We use two different models in this notebook, one for testing LLC ordinality and one for doing visualisations and checking estimated LLCs against known values. For the ordinality check, we have a model $y = (w_1 - 1) *(w_1^2 + w_2^2)^2$ which we use to predict gaussian noise, so the best loss is reached when $w_1 = 1$ or $w_1 = w_2 = 0 $, the latter being more degenerate. The second model used for the comparison vs theory is a polynomial model characterized by $w_1^a * w_2^b$ for some $(a, b)$, where $w_1$ and $w_2$ are weights to be learned. The data is again gaussian noise, so the model achieves its lowest loss when $w_1=0$ or $w_2 =0$.\n",
    "\n",
    "For a not-yet-finished version of the MNIST plot from this paper, see the MNIST notebook."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting pyro-ppl\n",
      "  Downloading pyro_ppl-1.9.1-py3-none-any.whl.metadata (7.8 kB)\n",
      "Requirement already satisfied: seaborn in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (0.13.2)\n",
      "Requirement already satisfied: numpy>=1.7 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from pyro-ppl) (1.24.4)\n",
      "Collecting opt-einsum>=2.3.2 (from pyro-ppl)\n",
      "  Downloading opt_einsum-3.3.0-py3-none-any.whl.metadata (6.5 kB)\n",
      "Collecting pyro-api>=0.1.1 (from pyro-ppl)\n",
      "  Downloading pyro_api-0.1.2-py3-none-any.whl.metadata (2.5 kB)\n",
      "Requirement already satisfied: torch>=2.0 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from pyro-ppl) (2.4.1)\n",
      "Requirement already satisfied: tqdm>=4.36 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from pyro-ppl) (4.66.5)\n",
      "Requirement already satisfied: pandas>=1.2 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from seaborn) (2.0.3)\n",
      "Requirement already satisfied: matplotlib!=3.6.1,>=3.4 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from seaborn) (3.7.5)\n",
      "Requirement already satisfied: contourpy>=1.0.1 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (1.1.1)\n",
      "Requirement already satisfied: cycler>=0.10 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (0.12.1)\n",
      "Requirement already satisfied: fonttools>=4.22.0 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (4.53.1)\n",
      "Requirement already satisfied: kiwisolver>=1.0.1 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (1.4.7)\n",
      "Requirement already satisfied: packaging>=20.0 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (24.1)\n",
      "Requirement already satisfied: pillow>=6.2.0 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (10.4.0)\n",
      "Requirement already satisfied: pyparsing>=2.3.1 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (3.1.4)\n",
      "Requirement already satisfied: python-dateutil>=2.7 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (2.9.0.post0)\n",
      "Requirement already satisfied: importlib-resources>=3.2.0 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (6.4.4)\n",
      "Requirement already satisfied: pytz>=2020.1 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from pandas>=1.2->seaborn) (2024.1)\n",
      "Requirement already satisfied: tzdata>=2022.1 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from pandas>=1.2->seaborn) (2024.1)\n",
      "Requirement already satisfied: filelock in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from torch>=2.0->pyro-ppl) (3.15.4)\n",
      "Requirement already satisfied: typing-extensions>=4.8.0 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from torch>=2.0->pyro-ppl) (4.12.2)\n",
      "Requirement already satisfied: sympy in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from torch>=2.0->pyro-ppl) (1.13.2)\n",
      "Requirement already satisfied: networkx in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from torch>=2.0->pyro-ppl) (3.1)\n",
      "Requirement already satisfied: jinja2 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from torch>=2.0->pyro-ppl) (3.1.4)\n",
      "Requirement already satisfied: fsspec in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from torch>=2.0->pyro-ppl) (2024.6.1)\n",
      "Requirement already satisfied: nvidia-cuda-nvrtc-cu12==12.1.105 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from torch>=2.0->pyro-ppl) (12.1.105)\n",
      "Requirement already satisfied: nvidia-cuda-runtime-cu12==12.1.105 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from torch>=2.0->pyro-ppl) (12.1.105)\n",
      "Requirement already satisfied: nvidia-cuda-cupti-cu12==12.1.105 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from torch>=2.0->pyro-ppl) (12.1.105)\n",
      "Requirement already satisfied: nvidia-cudnn-cu12==9.1.0.70 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from torch>=2.0->pyro-ppl) (9.1.0.70)\n",
      "Requirement already satisfied: nvidia-cublas-cu12==12.1.3.1 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from torch>=2.0->pyro-ppl) (12.1.3.1)\n",
      "Requirement already satisfied: nvidia-cufft-cu12==11.0.2.54 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from torch>=2.0->pyro-ppl) (11.0.2.54)\n",
      "Requirement already satisfied: nvidia-curand-cu12==10.3.2.106 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from torch>=2.0->pyro-ppl) (10.3.2.106)\n",
      "Requirement already satisfied: nvidia-cusolver-cu12==11.4.5.107 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from torch>=2.0->pyro-ppl) (11.4.5.107)\n",
      "Requirement already satisfied: nvidia-cusparse-cu12==12.1.0.106 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from torch>=2.0->pyro-ppl) (12.1.0.106)\n",
      "Requirement already satisfied: nvidia-nccl-cu12==2.20.5 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from torch>=2.0->pyro-ppl) (2.20.5)\n",
      "Requirement already satisfied: nvidia-nvtx-cu12==12.1.105 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from torch>=2.0->pyro-ppl) (12.1.105)\n",
      "Requirement already satisfied: triton==3.0.0 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from torch>=2.0->pyro-ppl) (3.0.0)\n",
      "Requirement already satisfied: nvidia-nvjitlink-cu12 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from nvidia-cusolver-cu12==11.4.5.107->torch>=2.0->pyro-ppl) (12.6.68)\n",
      "Requirement already satisfied: zipp>=3.1.0 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from importlib-resources>=3.2.0->matplotlib!=3.6.1,>=3.4->seaborn) (3.20.1)\n",
      "Requirement already satisfied: six>=1.5 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from python-dateutil>=2.7->matplotlib!=3.6.1,>=3.4->seaborn) (1.16.0)\n",
      "Requirement already satisfied: MarkupSafe>=2.0 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from jinja2->torch>=2.0->pyro-ppl) (2.1.5)\n",
      "Requirement already satisfied: mpmath<1.4,>=1.1.0 in /home/svwin/devinterp/.venv/lib/python3.8/site-packages (from sympy->torch>=2.0->pyro-ppl) (1.3.0)\n",
      "Downloading pyro_ppl-1.9.1-py3-none-any.whl (755 kB)\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m756.0/756.0 kB\u001b[0m \u001b[31m19.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[?25hDownloading opt_einsum-3.3.0-py3-none-any.whl (65 kB)\n",
      "Downloading pyro_api-0.1.2-py3-none-any.whl (11 kB)\n",
      "Installing collected packages: pyro-api, opt-einsum, pyro-ppl\n",
      "Successfully installed opt-einsum-3.3.0 pyro-api-0.1.2 pyro-ppl-1.9.1\n",
      "Note: you may need to restart the kernel to use updated packages.\n"
     ]
    }
   ],
   "source": [
    "%pip install pyro-ppl seaborn devinterp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "\n",
    "DEVICE = \"cuda\" if torch.cuda.is_available() else \"cpu\"\n",
    "# if you want to use your TPU (note that torch_xla is not by default installed, install using `pip install devinterp[tpu]`)\n",
    "# import os\n",
    "# os.environ[\"USE_TPU_BACKEND\"] = \"1\"\n",
    "# import torch_xla.core.xla_model as xm\n",
    "# DEVICE = xm.xla_device()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/svwin/devinterp/.venv/lib/python3.8/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
      "  from .autonotebook import tqdm as notebook_tqdm\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import pyro\n",
    "import pyro.distributions as dist\n",
    "import seaborn as sns\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "from matplotlib.colors import LinearSegmentedColormap\n",
    "from pyro.infer import MCMC, NUTS\n",
    "from torch.utils.data import DataLoader, TensorDataset\n",
    "\n",
    "from devinterp.optim.sgld import SGLD\n",
    "from devinterp.optim.sgnht import SGNHT\n",
    "from devinterp.slt.llc import LLCEstimator, OnlineLLCEstimator, SamplerCallback\n",
    "from devinterp.slt.sampler import estimate_learning_coeff_with_summary, sample\n",
    "from devinterp.utils import get_init_loss_multi_batch, plot_trace\n",
    "\n",
    "# DEVICE = \"cuda\" if torch.cuda.is_available() else \"cpu\"\n",
    "sns.set_style(\"whitegrid\")\n",
    "\n",
    "# plotting\n",
    "CMAP = sns.color_palette(\"muted\", as_cmap=True)\n",
    "PRIMARY, SECONDARY, TERTIARY = CMAP[:3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "from devinterp.utils import evaluate_mse\n",
    "\n",
    "# constants\n",
    "SIGMA = 0.25\n",
    "NUM_TRAIN_SAMPLES = 1000\n",
    "BATCH_SIZE = NUM_TRAIN_SAMPLES\n",
    "EVALUATE = evaluate_mse"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# some necessary functions\n",
    "class PolyModel(nn.Module):\n",
    "    def __init__(self, powers):\n",
    "        super(PolyModel, self).__init__()\n",
    "        self.weights = nn.Parameter(\n",
    "            torch.tensor([0.0, 0.0], dtype=torch.float32, requires_grad=True).to(DEVICE)\n",
    "        )\n",
    "        self.powers = powers\n",
    "\n",
    "    def forward(self, x):\n",
    "        multiplied = torch.prod(self.weights**self.powers)\n",
    "        x = x * multiplied\n",
    "        return x\n",
    "\n",
    "\n",
    "class LinePlusDotModel(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(LinePlusDotModel, self).__init__()\n",
    "        self.weights = nn.Parameter(\n",
    "            torch.tensor([0.0, 0.0], dtype=torch.float32, requires_grad=True).to(DEVICE)\n",
    "        )\n",
    "\n",
    "    def forward(self, x):\n",
    "        return (\n",
    "            x\n",
    "            * (self.weights[0] - 1)\n",
    "            * ((self.weights[0] ** 2 + self.weights[1] ** 2) ** 2)\n",
    "        )\n",
    "\n",
    "\n",
    "def generate_dataset_for_seed(seed=0):\n",
    "    torch.manual_seed(seed)\n",
    "    np.random.seed(seed)\n",
    "    x = torch.normal(0, 2, size=(NUM_TRAIN_SAMPLES,))\n",
    "    y = SIGMA * torch.normal(0, 1, size=(NUM_TRAIN_SAMPLES,))\n",
    "    train_data = TensorDataset(x, y)\n",
    "    train_loader = DataLoader(train_data, batch_size=BATCH_SIZE, shuffle=True)\n",
    "    return train_loader, train_data, x, y"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Benchmarking\n",
    "\n",
    "First, let's estimate some RLCTs of a simple degenerate line and point, where the point is more degenerate than the line. We first estimate the RLCTs at the point, and then at the line. This is a reproduction of Figure 1 from [Lau et al. (2023)](https://arxiv.org/pdf/2308.12108.pdf)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/svwin/devinterp/src/devinterp/backends/default/slt/sampler.py:216: UserWarning: You are taking more draws than burn-in steps, your LLC estimates will likely be underestimates. Please check LLC chain convergence.\n",
      "  warnings.warn(\n",
      "/home/svwin/devinterp/src/devinterp/backends/default/slt/sampler.py:220: UserWarning: You are taking more sample batches than there are dataloader batches available, this removes some randomness from sampling but is probably fine. (All sample batches beyond the number dataloader batches are cycled from the start, f.e. 9 samples from [A, B, C] would be [B, A, C, B, A, C, B, A, C].)\n",
      "  warnings.warn(\n",
      "/home/svwin/devinterp/src/devinterp/backends/default/slt/sampler.py:239: UserWarning: If you're setting a nbeta or temperature in optimizer_kwargs, please also make sure to set it in the callbacks.\n",
      "  warnings.warn(\n",
      "/home/svwin/devinterp/src/devinterp/backends/default/slt/sampler.py:53: UserWarning: You are taking more sample batches than there are dataloader batches available, this removes some randomness from sampling but is probably fine. (All sample batches beyond the number dataloader batches are cycled from the start, f.e. 9 samples from [A, B, C] would be [B, A, C, B, A, C, B, A, C].)\n",
      "  warnings.warn(\n",
      "Chain 0: 100%|██████████| 5000/5000 [00:03<00:00, 1539.29it/s]\n",
      "/home/svwin/devinterp/src/devinterp/backends/default/slt/sampler.py:53: UserWarning: You are taking more sample batches than there are dataloader batches available, this removes some randomness from sampling but is probably fine. (All sample batches beyond the number dataloader batches are cycled from the start, f.e. 9 samples from [A, B, C] would be [B, A, C, B, A, C, B, A, C].)\n",
      "  warnings.warn(\n",
      "Chain 1: 100%|██████████| 5000/5000 [00:03<00:00, 1565.90it/s]\n",
      "Chain 2: 100%|██████████| 5000/5000 [00:03<00:00, 1570.95it/s]\n",
      "Chain 3: 100%|██████████| 5000/5000 [00:03<00:00, 1570.16it/s]\n",
      "Chain 4: 100%|██████████| 5000/5000 [00:03<00:00, 1569.93it/s]\n",
      "/home/svwin/devinterp/src/devinterp/backends/default/slt/sampler.py:216: UserWarning: You are taking more draws than burn-in steps, your LLC estimates will likely be underestimates. Please check LLC chain convergence.\n",
      "  warnings.warn(\n",
      "/home/svwin/devinterp/src/devinterp/backends/default/slt/sampler.py:220: UserWarning: You are taking more sample batches than there are dataloader batches available, this removes some randomness from sampling but is probably fine. (All sample batches beyond the number dataloader batches are cycled from the start, f.e. 9 samples from [A, B, C] would be [B, A, C, B, A, C, B, A, C].)\n",
      "  warnings.warn(\n",
      "/home/svwin/devinterp/src/devinterp/backends/default/slt/sampler.py:239: UserWarning: If you're setting a nbeta or temperature in optimizer_kwargs, please also make sure to set it in the callbacks.\n",
      "  warnings.warn(\n",
      "Chain 0: 100%|██████████| 5000/5000 [00:02<00:00, 2043.49it/s]\n",
      "Chain 1: 100%|██████████| 5000/5000 [00:02<00:00, 2041.53it/s]\n",
      "Chain 2: 100%|██████████| 5000/5000 [00:02<00:00, 2037.82it/s]\n",
      "Chain 3: 100%|██████████| 5000/5000 [00:02<00:00, 2036.79it/s]\n",
      "Chain 4: 100%|██████████| 5000/5000 [00:02<00:00, 2033.40it/s]\n",
      "Chain 0: 100%|██████████| 5000/5000 [00:03<00:00, 1561.85it/s]\n",
      "Chain 1: 100%|██████████| 5000/5000 [00:03<00:00, 1565.91it/s]\n",
      "Chain 2: 100%|██████████| 5000/5000 [00:03<00:00, 1566.39it/s]\n",
      "Chain 3: 100%|██████████| 5000/5000 [00:03<00:00, 1570.48it/s]\n",
      "Chain 4: 100%|██████████| 5000/5000 [00:03<00:00, 1572.46it/s]\n",
      "Chain 0: 100%|██████████| 5000/5000 [00:02<00:00, 2039.88it/s]\n",
      "Chain 1: 100%|██████████| 5000/5000 [00:02<00:00, 2045.51it/s]\n",
      "Chain 2: 100%|██████████| 5000/5000 [00:02<00:00, 2039.15it/s]\n",
      "Chain 3: 100%|██████████| 5000/5000 [00:02<00:00, 2035.71it/s]\n",
      "Chain 4: 100%|██████████| 5000/5000 [00:02<00:00, 2042.70it/s]\n"
     ]
    }
   ],
   "source": [
    "from devinterp.utils import get_init_loss_multi_batch, default_nbeta\n",
    "\n",
    "train_loader, train_data, x, y = generate_dataset_for_seed(0)\n",
    "model = LinePlusDotModel().to(DEVICE)\n",
    "lr = 0.0001\n",
    "num_chains = 5\n",
    "num_draws = 5000\n",
    "nbeta = default_nbeta(train_loader)\n",
    "sgld_llcs = []\n",
    "sgnht_llcs = []\n",
    "sample_points = [[0.0, 0.0], [1.0, 0.3]]\n",
    "actual_rlcts = [0.25, 0.5]\n",
    "for sample_point in sample_points:\n",
    "    model.weights = nn.Parameter(\n",
    "        torch.tensor(sample_point, dtype=torch.float32, requires_grad=True)\n",
    "    )\n",
    "    init_loss = get_init_loss_multi_batch(\n",
    "        train_loader, num_chains, model, EVALUATE, DEVICE\n",
    "    )\n",
    "    sgnht_llc = LLCEstimator(\n",
    "        num_chains=num_chains, num_draws=num_draws, nbeta=nbeta, init_loss=init_loss\n",
    "    )\n",
    "    sgld_llc = LLCEstimator(\n",
    "        num_chains=num_chains, num_draws=num_draws, nbeta=nbeta, init_loss=init_loss\n",
    "    )\n",
    "    sample(\n",
    "        model,\n",
    "        train_loader,\n",
    "        evaluate=EVALUATE,\n",
    "        optimizer_kwargs=dict(lr=lr, bounding_box_size=0.5, nbeta=nbeta),\n",
    "        sampling_method=SGNHT,\n",
    "        num_chains=num_chains,\n",
    "        num_draws=num_draws,\n",
    "        device=DEVICE,\n",
    "        callbacks=[sgnht_llc],\n",
    "    )\n",
    "    sample(\n",
    "        model,\n",
    "        train_loader,\n",
    "        evaluate=EVALUATE,\n",
    "        optimizer_kwargs=dict(lr=lr, localization=0.2, nbeta=nbeta),  # TODO FIX\n",
    "        sampling_method=SGLD,\n",
    "        num_chains=num_chains,\n",
    "        num_draws=num_draws,\n",
    "        device=DEVICE,\n",
    "        callbacks=[sgld_llc],\n",
    "    )\n",
    "    sgld_llcs += [sgld_llc.get_results()]\n",
    "    sgnht_llcs += [sgnht_llc.get_results()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.14887431263923645\n",
      "[0.2005019336938858, 0.07028903812170029, 0.1147444099187851, 0.15498630702495575, 0.20384985208511353]\n",
      "0.654302716255188\n",
      "[0.56605064868927, 0.4899264872074127, 0.5089262127876282, 0.8386633396148682, 0.8679468631744385]\n",
      "0.20462393760681152\n",
      "[0.202658012509346, 0.205875962972641, 0.18717604875564575, 0.22053281962871552, 0.20687688887119293]\n",
      "0.8179630041122437\n",
      "[1.257343053817749, 0.9880695939064026, 0.8615443706512451, 0.404544860124588, 0.5783130526542664]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1400x300 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(1, 2, figsize=(14, 3))\n",
    "fig.patch.set_facecolor(\"white\")\n",
    "for i, (sampler_name, rlct_estimates) in enumerate(\n",
    "    zip([\"SGLD\", \"SGNHT\"], [sgld_llcs, sgnht_llcs])\n",
    "):\n",
    "    axes[i].set_xlim([-0.5, 1.5])\n",
    "    axes[i].set_title(sampler_name)\n",
    "    axes[i].set_xlabel(r\"$\\hat\\lambda$\")\n",
    "    for j, (rlct_estimates, sample_point) in enumerate(\n",
    "        zip(rlct_estimates, sample_points)\n",
    "    ):\n",
    "        print(rlct_estimates[\"llc/mean\"])\n",
    "        print([rlct_estimates[f\"llc-chain/{k}\"] for k in range(num_chains)])\n",
    "        axes[i].axvline(actual_rlcts[j], color=\"black\", linestyle=\"--\")\n",
    "        axes[i].hist(\n",
    "            [rlct_estimates[f\"llc-chain/{i}\"] for i in range(num_chains)],\n",
    "            alpha=0.6,\n",
    "            bins=40,\n",
    "            range=(-0.5, 1.5),\n",
    "            label=rf\"$w_1={sample_point[0]:.1f}, w_2={sample_point[1]:.1f}$\",\n",
    "            color=SECONDARY if j else PRIMARY,\n",
    "        )\n",
    "    axes[i].legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Trajectory Plots\n",
    "\n",
    "Let's see how well our different sampling methods explore the local loss landscape. We start out sampling at $w_1 = 0.5, w_2 = 0.01$, and plot the resulting trajectories after 10k steps."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# plotting\n",
    "def hex_to_rgb(hex_color):\n",
    "    hex_color = hex_color.lstrip(\"#\")\n",
    "    return tuple(int(hex_color[i : i + 2], 16) / 255.0 for i in (0, 2, 4))\n",
    "\n",
    "\n",
    "lighter_factor = 0.9  # Between 0 and 1, higher values make it closer to white\n",
    "lighter_SECONDARY = tuple(\n",
    "    [x + (1 - x) * lighter_factor for x in hex_to_rgb(SECONDARY)[:3]] + [1.0]\n",
    ")\n",
    "\n",
    "colors = [SECONDARY, lighter_SECONDARY]\n",
    "n_bins = 20  # Number of bins\n",
    "contour_cmap = LinearSegmentedColormap.from_list(\"custom_cmap\", colors, N=n_bins)\n",
    "\n",
    "powers = torch.tensor([1, 3]).to(DEVICE)\n",
    "model = PolyModel(powers).to(DEVICE)\n",
    "\n",
    "\n",
    "def plot_trajectories(trajectories, names, device=DEVICE):\n",
    "    fig, axes = plt.subplots(1, len(trajectories), figsize=(15, 5))\n",
    "    w1_range = np.linspace(-2, 2, 21)\n",
    "    w2_range = np.linspace(-2, 2, 21)\n",
    "    w1_vals, w2_vals = np.meshgrid(w1_range, w2_range)\n",
    "    Z = np.zeros_like(w1_vals, dtype=float)\n",
    "\n",
    "    for i in range(w1_vals.shape[0]):\n",
    "        for j in range(w1_vals.shape[1]):\n",
    "            w1 = w1_vals[i, j]\n",
    "            w2 = w2_vals[i, j]\n",
    "            model.weights = nn.Parameter(\n",
    "                torch.tensor([w1, w2], dtype=torch.float32).to(device)\n",
    "            )\n",
    "            Z[i, j] = (\n",
    "                model.to(device)(torch.tensor(1.0).to(device)).item() ** 2\n",
    "            )  # MSE, so square this\n",
    "\n",
    "    custom_levels = np.linspace(Z.min(), Z.max() * 0.04, n_bins)\n",
    "\n",
    "    for i, trajectory in enumerate(trajectories):\n",
    "        axes[i].contourf(\n",
    "            w1_vals, w2_vals, Z, levels=custom_levels, cmap=contour_cmap, alpha=0.8\n",
    "        )\n",
    "        weights = trajectory[\"ws/trace\"][0]\n",
    "        draws_array = np.array(\n",
    "            [\n",
    "                d\n",
    "                for d in weights\n",
    "                if w1_range[0] <= d[0] <= w1_range[-1]\n",
    "                and w2_range[0] <= d[1] <= w2_range[-1]\n",
    "            ]\n",
    "        )\n",
    "        sns.scatterplot(\n",
    "            x=draws_array[:, 0],\n",
    "            y=draws_array[:, 1],\n",
    "            marker=\"x\",\n",
    "            ax=axes[i],\n",
    "            s=10,\n",
    "            color=PRIMARY,\n",
    "        )\n",
    "        axes[i].axhline(0, linestyle=\"--\", color=\"gray\")\n",
    "        axes[i].axvline(0, linestyle=\"--\", color=\"gray\")\n",
    "        axes[i].set_xlabel(r\"$w_{1}$\")\n",
    "        axes[i].set_ylabel(r\"$w_{2}$\")\n",
    "        axes[i].set_title(names[i])\n",
    "        axes[i].grid(False)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_loader, train_data, x, y = generate_dataset_for_seed(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "class WeightCallback(SamplerCallback):\n",
    "    def __init__(self, num_chains: int, num_draws: int, model, device=\"cpu\"):\n",
    "        self.num_chains = num_chains\n",
    "        self.num_draws = num_draws\n",
    "        self.ws = np.zeros(\n",
    "            (num_chains, num_draws, len(model.weights)), dtype=np.float32\n",
    "        )\n",
    "        self.device = device\n",
    "\n",
    "    def update(self, chain: int, draw: int, model):\n",
    "        self.ws[chain, draw] = model.weights.cpu().detach()\n",
    "\n",
    "    def get_results(self):\n",
    "        return {\n",
    "            \"ws/trace\": self.ws,\n",
    "        }\n",
    "\n",
    "    def __call__(self, chain: int, draw: int, model, **kwargs):\n",
    "        self.update(chain, draw, model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Chain 0:   0%|          | 0/10000 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Chain 0: 100%|██████████| 10000/10000 [00:05<00:00, 1778.81it/s]\n",
      "Chain 0: 100%|██████████| 10000/10000 [00:04<00:00, 2293.33it/s]\n",
      "Sample: 100%|██████████| 520/520 [00:15, 34.49it/s, step size=1.56e+00, acc. prob=0.371] \n"
     ]
    }
   ],
   "source": [
    "torch.manual_seed(0)\n",
    "np.random.seed(0)\n",
    "model.weights = nn.Parameter(\n",
    "    torch.tensor([0.5, 0.01], dtype=torch.float32, requires_grad=True).to(DEVICE)\n",
    ")\n",
    "lr = 0.0005\n",
    "sgld_weights = WeightCallback(num_chains=1, num_draws=10_000, model=model)\n",
    "sgnht_weights = WeightCallback(num_chains=1, num_draws=10_000, model=model)\n",
    "trace_sgnht = sample(\n",
    "    model,\n",
    "    train_loader,\n",
    "    evaluate=EVALUATE,\n",
    "    optimizer_kwargs=dict(\n",
    "        lr=lr,\n",
    "        bounding_box_size=2.0,\n",
    "        diffusion_factor=0.01,\n",
    "        nbeta=default_nbeta(train_loader),\n",
    "    ),\n",
    "    sampling_method=SGNHT,\n",
    "    num_chains=1,\n",
    "    num_draws=10_000,\n",
    "    num_burnin_steps=0,\n",
    "    num_steps_bw_draws=1,\n",
    "    verbose=True,\n",
    "    device=DEVICE,\n",
    "    callbacks=[sgnht_weights],\n",
    ")\n",
    "torch.manual_seed(0)\n",
    "np.random.seed(0)\n",
    "model.weights = nn.Parameter(\n",
    "    torch.tensor([0.5, 0.01], dtype=torch.float32, requires_grad=True).to(DEVICE)\n",
    ")\n",
    "trace_sgld = sample(\n",
    "    model,\n",
    "    train_loader,\n",
    "    evaluate=EVALUATE,\n",
    "    optimizer_kwargs=dict(\n",
    "        lr=lr,\n",
    "        bounding_box_size=2.0,\n",
    "        nbeta=default_nbeta(train_loader),\n",
    "    ),\n",
    "    sampling_method=SGLD,\n",
    "    num_chains=1,\n",
    "    num_draws=10_000,\n",
    "    num_burnin_steps=0,\n",
    "    num_steps_bw_draws=1,\n",
    "    verbose=True,\n",
    "    device=DEVICE,\n",
    "    callbacks=[sgld_weights],\n",
    ")\n",
    "torch.manual_seed(0)\n",
    "np.random.seed(0)\n",
    "model.weights = nn.Parameter(\n",
    "    torch.tensor([0.5, 0.01], dtype=torch.float32, requires_grad=True).to(DEVICE)\n",
    ")\n",
    "model = model.to(\"cpu\")\n",
    "model.powers = powers.to(\"cpu\")\n",
    "w_prior = dist.Normal(model.weights, torch.ones_like(model.weights)).to_event(1)\n",
    "\n",
    "\n",
    "def pyro_model(x, y):\n",
    "    w = pyro.sample(\"weights\", w_prior)\n",
    "    model.weights = nn.Parameter(w.to(\"cpu\"))\n",
    "    y_pred = model(x.to(\"cpu\"))\n",
    "    pyro.sample(\"obs\", dist.Normal(y_pred, 0.5), obs=y.to(\"cpu\"))\n",
    "\n",
    "\n",
    "nuts_kernel = NUTS(pyro_model)\n",
    "mcmc = MCMC(nuts_kernel, num_samples=500, warmup_steps=20)\n",
    "mcmc.run(x, y)\n",
    "\n",
    "posterior_samples = mcmc.get_samples()\n",
    "trace_mcmc = {\"ws/trace\": [posterior_samples[\"weights\"].detach().cpu().numpy()]}\n",
    "\n",
    "model = model.to(DEVICE)\n",
    "model.powers = powers.to(DEVICE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x500 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_trajectories(\n",
    "    [sgld_weights.get_results(), sgnht_weights.get_results(), trace_mcmc],\n",
    "    names=[\"SGLD\", \"SGNHT\", \"MCMC\"],\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Benchmarking\n",
    "\n",
    "Finally, let's check if we get the expected lambdahat values for a few different crossings. This is similar to Table 1 from [Lau et al. (2023)](https://arxiv.org/pdf/2308.12108.pdf)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/svwin/devinterp/src/devinterp/slt/sampler.py:50: UserWarning: Using passed in nbeta. Make sure callbacks are also initialized with the same nbeta.\n",
      "  warnings.warn(\n",
      "/home/svwin/devinterp/src/devinterp/backends/default/slt/sampler.py:216: UserWarning: You are taking more draws than burn-in steps, your LLC estimates will likely be underestimates. Please check LLC chain convergence.\n",
      "  warnings.warn(\n",
      "/home/svwin/devinterp/src/devinterp/backends/default/slt/sampler.py:220: UserWarning: You are taking more sample batches than there are dataloader batches available, this removes some randomness from sampling but is probably fine. (All sample batches beyond the number dataloader batches are cycled from the start, f.e. 9 samples from [A, B, C] would be [B, A, C, B, A, C, B, A, C].)\n",
      "  warnings.warn(\n",
      "/home/svwin/devinterp/src/devinterp/backends/default/slt/sampler.py:239: UserWarning: If you're setting a nbeta or temperature in optimizer_kwargs, please also make sure to set it in the callbacks.\n",
      "  warnings.warn(\n",
      "/home/svwin/devinterp/src/devinterp/backends/default/slt/sampler.py:53: UserWarning: You are taking more sample batches than there are dataloader batches available, this removes some randomness from sampling but is probably fine. (All sample batches beyond the number dataloader batches are cycled from the start, f.e. 9 samples from [A, B, C] would be [B, A, C, B, A, C, B, A, C].)\n",
      "  warnings.warn(\n",
      "Chain 0:   0%|          | 0/5000 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Chain 0: 100%|██████████| 5000/5000 [00:02<00:00, 2299.26it/s]\n",
      "Chain 1: 100%|██████████| 5000/5000 [00:02<00:00, 2305.10it/s]\n",
      "Chain 2: 100%|██████████| 5000/5000 [00:02<00:00, 2310.77it/s]\n",
      "Chain 0: 100%|██████████| 5000/5000 [00:02<00:00, 2300.12it/s]\n",
      "Chain 1: 100%|██████████| 5000/5000 [00:02<00:00, 2303.21it/s]\n",
      "Chain 2: 100%|██████████| 5000/5000 [00:02<00:00, 2300.29it/s]\n",
      "Chain 0: 100%|██████████| 5000/5000 [00:02<00:00, 2304.30it/s]\n",
      "Chain 1: 100%|██████████| 5000/5000 [00:02<00:00, 2301.50it/s]\n",
      "Chain 2: 100%|██████████| 5000/5000 [00:02<00:00, 2308.95it/s]\n",
      "Chain 0: 100%|██████████| 5000/5000 [00:02<00:00, 2305.99it/s]\n",
      "Chain 1: 100%|██████████| 5000/5000 [00:02<00:00, 2301.46it/s]\n",
      "Chain 2: 100%|██████████| 5000/5000 [00:02<00:00, 2301.10it/s]\n",
      "Chain 0: 100%|██████████| 5000/5000 [00:02<00:00, 2310.93it/s]\n",
      "Chain 1: 100%|██████████| 5000/5000 [00:02<00:00, 2299.47it/s]\n",
      "Chain 2: 100%|██████████| 5000/5000 [00:02<00:00, 2309.69it/s]\n",
      "Chain 0: 100%|██████████| 5000/5000 [00:02<00:00, 2302.61it/s]\n",
      "Chain 1: 100%|██████████| 5000/5000 [00:02<00:00, 2303.14it/s]\n",
      "Chain 2: 100%|██████████| 5000/5000 [00:02<00:00, 2302.53it/s]\n",
      "Chain 0: 100%|██████████| 5000/5000 [00:02<00:00, 2300.96it/s]\n",
      "Chain 1: 100%|██████████| 5000/5000 [00:02<00:00, 2300.41it/s]\n",
      "Chain 2: 100%|██████████| 5000/5000 [00:02<00:00, 2304.66it/s]\n",
      "Chain 0: 100%|██████████| 5000/5000 [00:02<00:00, 2301.44it/s]\n",
      "Chain 1: 100%|██████████| 5000/5000 [00:02<00:00, 2304.44it/s]\n",
      "Chain 2: 100%|██████████| 5000/5000 [00:02<00:00, 2303.65it/s]\n",
      "Chain 0: 100%|██████████| 5000/5000 [00:02<00:00, 2301.49it/s]\n",
      "Chain 1: 100%|██████████| 5000/5000 [00:02<00:00, 2303.87it/s]\n",
      "Chain 2: 100%|██████████| 5000/5000 [00:02<00:00, 2303.59it/s]\n",
      "Chain 0: 100%|██████████| 5000/5000 [00:02<00:00, 1789.69it/s]\n",
      "Chain 1: 100%|██████████| 5000/5000 [00:02<00:00, 1796.84it/s]\n",
      "Chain 2: 100%|██████████| 5000/5000 [00:02<00:00, 1790.16it/s]\n",
      "Chain 0: 100%|██████████| 5000/5000 [00:02<00:00, 1795.61it/s]\n",
      "Chain 1: 100%|██████████| 5000/5000 [00:02<00:00, 1793.41it/s]\n",
      "Chain 2: 100%|██████████| 5000/5000 [00:02<00:00, 1784.45it/s]\n",
      "Chain 0: 100%|██████████| 5000/5000 [00:02<00:00, 1795.53it/s]\n",
      "Chain 1: 100%|██████████| 5000/5000 [00:02<00:00, 1795.98it/s]\n",
      "Chain 2: 100%|██████████| 5000/5000 [00:02<00:00, 1792.51it/s]\n",
      "Chain 0: 100%|██████████| 5000/5000 [00:02<00:00, 1788.73it/s]\n",
      "Chain 1: 100%|██████████| 5000/5000 [00:02<00:00, 1789.03it/s]\n",
      "Chain 2: 100%|██████████| 5000/5000 [00:02<00:00, 1798.70it/s]\n",
      "Chain 0: 100%|██████████| 5000/5000 [00:02<00:00, 1792.37it/s]\n",
      "Chain 1: 100%|██████████| 5000/5000 [00:02<00:00, 1795.58it/s]\n",
      "Chain 2: 100%|██████████| 5000/5000 [00:02<00:00, 1797.96it/s]\n",
      "Chain 0: 100%|██████████| 5000/5000 [00:02<00:00, 1794.15it/s]\n",
      "Chain 1: 100%|██████████| 5000/5000 [00:02<00:00, 1793.42it/s]\n",
      "Chain 2: 100%|██████████| 5000/5000 [00:02<00:00, 1791.93it/s]\n",
      "Chain 0: 100%|██████████| 5000/5000 [00:02<00:00, 1790.67it/s]\n",
      "Chain 1: 100%|██████████| 5000/5000 [00:02<00:00, 1679.98it/s]\n",
      "Chain 2: 100%|██████████| 5000/5000 [00:02<00:00, 1788.83it/s]\n",
      "Chain 0: 100%|██████████| 5000/5000 [00:02<00:00, 1790.11it/s]\n",
      "Chain 1: 100%|██████████| 5000/5000 [00:02<00:00, 1784.29it/s]\n",
      "Chain 2: 100%|██████████| 5000/5000 [00:02<00:00, 1788.46it/s]\n",
      "Chain 0: 100%|██████████| 5000/5000 [00:02<00:00, 1788.42it/s]\n",
      "Chain 1: 100%|██████████| 5000/5000 [00:02<00:00, 1788.54it/s]\n",
      "Chain 2: 100%|██████████| 5000/5000 [00:02<00:00, 1790.13it/s]\n"
     ]
    }
   ],
   "source": [
    "train_loader, train_data, x, y = generate_dataset_for_seed(0)\n",
    "lr = 0.0001\n",
    "num_chains = 3\n",
    "num_draws = 5000\n",
    "powers_to_test = [\n",
    "    torch.tensor([0, 1]).to(DEVICE),\n",
    "    torch.tensor([0, 2]).to(DEVICE),\n",
    "    torch.tensor([0, 3]).to(DEVICE),\n",
    "    torch.tensor([1, 1]).to(DEVICE),\n",
    "    torch.tensor([1, 2]).to(DEVICE),\n",
    "    torch.tensor([1, 3]).to(DEVICE),\n",
    "    torch.tensor([2, 2]).to(DEVICE),\n",
    "    torch.tensor([2, 3]).to(DEVICE),\n",
    "    torch.tensor([3, 3]).to(DEVICE),\n",
    "]\n",
    "rlct_estimates_sgld = {\n",
    "    tuple(power.detach().cpu().tolist()): estimate_learning_coeff_with_summary(\n",
    "        model=PolyModel(power),\n",
    "        loader=train_loader,\n",
    "        optimizer_kwargs=dict(\n",
    "            lr=lr,\n",
    "            bounding_box_size=1.0,\n",
    "            nbeta=default_nbeta(train_loader),\n",
    "        ),\n",
    "        evaluate=EVALUATE,\n",
    "        sampling_method=SGLD,\n",
    "        num_chains=num_chains,\n",
    "        num_draws=num_draws,\n",
    "        device=DEVICE,\n",
    "    )\n",
    "    for power in powers_to_test\n",
    "}\n",
    "rlct_estimates_sgnht = {\n",
    "    tuple(power.detach().cpu().tolist()): estimate_learning_coeff_with_summary(\n",
    "        model=PolyModel(power),\n",
    "        loader=train_loader,\n",
    "        optimizer_kwargs=dict(\n",
    "            lr=lr,\n",
    "            bounding_box_size=1.0,\n",
    "            diffusion_factor=0.01,\n",
    "            nbeta=default_nbeta(train_loader),\n",
    "        ),\n",
    "        evaluate=EVALUATE,\n",
    "        sampling_method=SGNHT,\n",
    "        num_chains=num_chains,\n",
    "        num_draws=num_draws,\n",
    "        device=DEVICE,\n",
    "    )\n",
    "    for power in powers_to_test\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "sgld\n",
      "powers  llc_mean  llc_std\n",
      "(0, 1)  0.499807 0.033900\n",
      "(0, 2)  0.301098 0.111747\n",
      "(0, 3)  0.108836 0.006892\n",
      "(1, 1)  0.350617 0.041256\n",
      "(1, 2)  0.156247 0.136191\n",
      "(1, 3)  0.048869 0.017304\n",
      "(2, 2)  0.085732 0.053104\n",
      "(2, 3)  0.084958 0.125995\n",
      "(3, 3)  0.099944 0.130149\n",
      "sgnht\n",
      "powers  llc_mean  llc_std\n",
      "(0, 1)  0.492101 0.075376\n",
      "(0, 2)  0.231741 0.015389\n",
      "(0, 3)  0.173720 0.032608\n",
      "(1, 1)  0.378363 0.042097\n",
      "(1, 2)  0.232073 0.012143\n",
      "(1, 3)  0.172249 0.031408\n",
      "(2, 2)  0.178082 0.017687\n",
      "(2, 3)  0.114668 0.013921\n",
      "(3, 3)  0.081164 0.018655\n"
     ]
    }
   ],
   "source": [
    "for sampler_type, estimates in zip(\n",
    "    (\"sgld\", \"sgnht\"), (rlct_estimates_sgld, rlct_estimates_sgnht)\n",
    "):\n",
    "    df_data = {\n",
    "        \"powers\": [k for k in estimates.keys()],\n",
    "        \"LLC_Summary\": estimates.values(),\n",
    "    }\n",
    "    df = pd.DataFrame(df_data)\n",
    "    df[\"llc_mean\"] = df[\"LLC_Summary\"].apply(lambda x: x[\"llc/mean\"])\n",
    "    df[\"llc_std\"] = df[\"LLC_Summary\"].apply(lambda x: x[\"llc/std\"])\n",
    "    df = df.drop(\"LLC_Summary\", axis=1)\n",
    "    print(sampler_type)\n",
    "    print(df.to_string(index=False))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": ".venv",
   "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": "3.8.10"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}