2.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "iter:50 training loss: 0.6541643142700195, val loss: 1.0987168550491333\n",
      "iter:100 training loss: 0.5443618297576904, val loss: 0.9234420657157898\n",
      "iter:150 training loss: 0.42738205194473267, val loss: 0.7345316410064697\n",
      "iter:200 training loss: 0.31288474798202515, val loss: 0.5471447706222534\n",
      "iter:250 training loss: 0.21092523634433746, val loss: 0.3777029514312744\n",
      "iter:300 training loss: 0.12925758957862854, val loss: 0.239410400390625\n",
      "iter:350 training loss: 0.0712103545665741, val loss: 0.13860878348350525\n",
      "iter:400 training loss: 0.035106856375932693, val loss: 0.07357435673475266\n",
      "iter:450 training loss: 0.01567256823182106, val loss: 0.036532677710056305\n",
      "iter:500 training loss: 0.006652833893895149, val loss: 0.017728392034769058\n",
      "iter:550 training loss: 0.0030161100439727306, val loss: 0.00899422075599432\n",
      "iter:600 training loss: 0.0017095651710405946, val loss: 0.005120445974171162\n",
      "iter:650 training loss: 0.0012632126454263926, val loss: 0.0033916705287992954\n",
      "iter:700 training loss: 0.0010932876029983163, val loss: 0.002574644051492214\n",
      "iter:750 training loss: 0.001002510660327971, val loss: 0.0021476992405951023\n",
      "iter:800 training loss: 0.0009328412706963718, val loss: 0.001892455737106502\n",
      "iter:850 training loss: 0.0008692669216543436, val loss: 0.0017150428611785173\n",
      "iter:900 training loss: 0.0008084333967417479, val loss: 0.0015740288654342294\n",
      "iter:950 training loss: 0.000749790167901665, val loss: 0.0014510229229927063\n",
      "Took 0.21765804290771484s for 1000 epochs\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "import random\n",
    "import matplotlib.pyplot as plt\n",
    "import time\n",
    "import numpy as np\n",
    "import plotly.express as px\n",
    "from dataclasses import dataclass\n",
    "from utils import *\n",
    "torch.manual_seed(10)\n",
    "\n",
    "\n",
    "\n",
    "input_size = 2\n",
    "learning_rate = 1e-3\n",
    "max_iters = 1000\n",
    "dataset_size = 16\n",
    "batch_size = dataset_size\n",
    "\n",
    "\n",
    "iters = []\n",
    "loss_value = []\n",
    "val_loss_value = []\n",
    "\n",
    "\n",
    "def sample_data(split = 'train', batch_size = batch_size):\n",
    "    if split != 'train':\n",
    "        curr_data = validation_dataset\n",
    "    else:\n",
    "        curr_data = training_dataset\n",
    "    curr_data = curr_data[torch.randperm(curr_data.shape[0])][:batch_size]\n",
    "    a, b = curr_data.shape\n",
    "    \n",
    "    sparsity_vector = torch.rand(a, b)\n",
    "    sparsity_mat = torch.bernoulli(torch.full((a, b), 1.0) * sparsity_vector)\n",
    "    \n",
    "    # X = curr_data * sparsity_mat\n",
    "    X = curr_data\n",
    "    return X\n",
    "\n",
    "\n",
    "\n",
    "DEVICE = torch.device('cpu')\n",
    "\n",
    "@dataclass\n",
    "class Config:\n",
    "    input_size : int\n",
    "    hidden_size : int = 2\n",
    "    layers : int = 2\n",
    "\n",
    "class Two_Two(nn.Module):\n",
    "    def __init__(self, hidden_size) -> None:\n",
    "        super().__init__()\n",
    "        self.layers = nn.Sequential(nn.Linear(hidden_size,hidden_size), nn.Linear(hidden_size,hidden_size) )\n",
    "    \n",
    "    def forward(self, idx):\n",
    "        return self.layers(idx)\n",
    "\n",
    "class Model(nn.Module):\n",
    "    def __init__(self, config : Config) -> None:\n",
    "        super().__init__()\n",
    "        self.config = config\n",
    "        self.layers = nn.Sequential(\n",
    "                        nn.Linear(config.input_size, config.hidden_size, bias= False),\n",
    "                        #nn.ReLU(),\n",
    "                        nn.Linear(config.hidden_size, 1, bias= False),\n",
    "                    )\n",
    "\n",
    "    \n",
    "    def forward(self, idx : torch.Tensor):\n",
    "        count = self.layers(idx) \n",
    "\n",
    "        loss = nn.MSELoss()\n",
    "        \n",
    "        return count, ((idx.sum(-1) - count.squeeze(-1))**2).mean()\n",
    "\n",
    "\n",
    "training_dataset = torch.rand(dataset_size, input_size)\n",
    "validation_dataset = torch.rand(int(dataset_size/2), input_size)\n",
    "output = input\n",
    "\n",
    "\n",
    "model = Model(Config(input_size= input_size))\n",
    "\n",
    "def get_accuracy():\n",
    "    a= sample_data(split = 'val', batch_size= 1024)\n",
    "    n = (model(a)[0].argmax(-1) == (a>0).sum(-1))\n",
    "    print(\"val accuracy\",n.sum()/a.shape[0])\n",
    "optimizer = torch.optim.Adam(model.parameters() , learning_rate)\n",
    "start_time = time.time()\n",
    "step_value = max_iters/20\n",
    "for i in range(1, max_iters):\n",
    "    X =  sample_data()\n",
    "    count, loss = model(X)\n",
    "   \n",
    "    if i%step_value ==0 :\n",
    "        model.eval()\n",
    "        with torch.no_grad():\n",
    "            X_V = sample_data(split = 'val')\n",
    "            _, val_loss= model(X_V)\n",
    "            iters.append(i)\n",
    "            loss_value.append(loss.item())\n",
    "            val_loss_value.append(val_loss.item())\n",
    "            print(f\"iter:{i} training loss: {loss.item()}, val loss: {val_loss.item()}\")\n",
    "        model.train()\n",
    "\n",
    "    optimizer.zero_grad(set_to_none=True)\n",
    "    #loss+= get_l1_penalty()\n",
    "    loss.backward()\n",
    "    optimizer.step()\n",
    "\n",
    "\n",
    "end_time = time.time()\n",
    "print(f\"Took {end_time-start_time}s for {max_iters} epochs\")\n",
    "\n",
    "plt.xlabel(\"Epochs\")\n",
    "plt.ylabel(\"Loss\")\n",
    "plt.plot(iters,loss_value, color='red', label=\"Training\")\n",
    "plt.plot(iters, val_loss_value, \"blue\", label = \"validation\")\n",
    "\n",
    "\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parameter containing:\n",
      "tensor([[-0.9733, -0.4169],\n",
      "        [-0.0376,  1.1518]], requires_grad=True) Parameter containing:\n",
      "tensor([[-0.9346,  0.6133]], requires_grad=True)\n"
     ]
    }
   ],
   "source": [
    "print(model.layers[0].weight, model.layers[-1].weight)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([0.6150, 0.4967]) tensor(1.1117) (tensor([1.0897], grad_fn=<SqueezeBackward4>), tensor(0.0005, grad_fn=<MeanBackward0>))\n",
      "tensor([0.0310, 0.7192]) tensor(0.7502) (tensor([0.8158], grad_fn=<SqueezeBackward4>), tensor(0.0043, grad_fn=<MeanBackward0>))\n",
      "tensor([0.6959, 0.9849]) tensor(1.6808) (tensor([1.6965], grad_fn=<SqueezeBackward4>), tensor(0.0002, grad_fn=<MeanBackward0>))\n",
      "tensor([0.0687, 0.0501]) tensor(0.1188) (tensor([0.1158], grad_fn=<SqueezeBackward4>), tensor(8.8702e-06, grad_fn=<MeanBackward0>))\n",
      "tensor([0.8067, 0.8379]) tensor(1.6445) (tensor([1.6336], grad_fn=<SqueezeBackward4>), tensor(0.0001, grad_fn=<MeanBackward0>))\n",
      "tensor([0.6178, 0.3304]) tensor(0.9482) (tensor([0.9098], grad_fn=<SqueezeBackward4>), tensor(0.0015, grad_fn=<MeanBackward0>))\n",
      "tensor([0.2139, 0.4118]) tensor(0.6258) (tensor([0.6411], grad_fn=<SqueezeBackward4>), tensor(0.0002, grad_fn=<MeanBackward0>))\n",
      "tensor([0.4581, 0.4829]) tensor(0.9409) (tensor([0.9354], grad_fn=<SqueezeBackward4>), tensor(3.0690e-05, grad_fn=<MeanBackward0>))\n",
      "tensor([0.5479, 0.4440]) tensor(0.9919) (tensor([0.9724], grad_fn=<SqueezeBackward4>), tensor(0.0004, grad_fn=<MeanBackward0>))\n",
      "tensor([0.4521, 0.0575]) tensor(0.5096) (tensor([0.4639], grad_fn=<SqueezeBackward4>), tensor(0.0021, grad_fn=<MeanBackward0>))\n"
     ]
    }
   ],
   "source": [
    "X = sample_data()\n",
    "for i in range(10):\n",
    "    print(X[i], X[i].sum(), model(X[i]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(tensor([8.1404], grad_fn=<SqueezeBackward4>),\n",
       " tensor(0.0197, grad_fn=<MeanBackward0>))"
      ]
     },
     "execution_count": 126,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model(torch.tensor([3.0, 5.0]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.plotly.v1+json": {
       "config": {
        "plotlyServerURL": "https://plot.ly"
       },
       "data": [
        {
         "coloraxis": "coloraxis",
         "hovertemplate": "x: %{x}<br>y: %{y}<br>color: %{z}<extra></extra>",
         "name": "0",
         "type": "heatmap",
         "xaxis": "x",
         "yaxis": "y",
         "z": [
          [
           -0.9733009934425354,
           -0.03757869452238083
          ],
          [
           -0.41694021224975586,
           1.1518406867980957
          ]
         ]
        }
       ],
       "layout": {
        "coloraxis": {
         "cmid": 0,
         "colorscale": [
          [
           0,
           "rgb(103,0,31)"
          ],
          [
           0.1,
           "rgb(178,24,43)"
          ],
          [
           0.2,
           "rgb(214,96,77)"
          ],
          [
           0.3,
           "rgb(244,165,130)"
          ],
          [
           0.4,
           "rgb(253,219,199)"
          ],
          [
           0.5,
           "rgb(247,247,247)"
          ],
          [
           0.6,
           "rgb(209,229,240)"
          ],
          [
           0.7,
           "rgb(146,197,222)"
          ],
          [
           0.8,
           "rgb(67,147,195)"
          ],
          [
           0.9,
           "rgb(33,102,172)"
          ],
          [
           1,
           "rgb(5,48,97)"
          ]
         ]
        },
        "margin": {
         "t": 60
        },
        "template": {
         "data": {
          "bar": [
           {
            "error_x": {
             "color": "#2a3f5f"
            },
            "error_y": {
             "color": "#2a3f5f"
            },
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "bar"
           }
          ],
          "barpolar": [
           {
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "barpolar"
           }
          ],
          "carpet": [
           {
            "aaxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "baxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "type": "carpet"
           }
          ],
          "choropleth": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "choropleth"
           }
          ],
          "contour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "contour"
           }
          ],
          "contourcarpet": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "contourcarpet"
           }
          ],
          "heatmap": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmap"
           }
          ],
          "heatmapgl": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmapgl"
           }
          ],
          "histogram": [
           {
            "marker": {
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "histogram"
           }
          ],
          "histogram2d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2d"
           }
          ],
          "histogram2dcontour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2dcontour"
           }
          ],
          "mesh3d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "mesh3d"
           }
          ],
          "parcoords": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "parcoords"
           }
          ],
          "pie": [
           {
            "automargin": true,
            "type": "pie"
           }
          ],
          "scatter": [
           {
            "fillpattern": {
             "fillmode": "overlay",
             "size": 10,
             "solidity": 0.2
            },
            "type": "scatter"
           }
          ],
          "scatter3d": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter3d"
           }
          ],
          "scattercarpet": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattercarpet"
           }
          ],
          "scattergeo": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergeo"
           }
          ],
          "scattergl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergl"
           }
          ],
          "scattermapbox": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattermapbox"
           }
          ],
          "scatterpolar": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolar"
           }
          ],
          "scatterpolargl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolargl"
           }
          ],
          "scatterternary": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterternary"
           }
          ],
          "surface": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "surface"
           }
          ],
          "table": [
           {
            "cells": {
             "fill": {
              "color": "#EBF0F8"
             },
             "line": {
              "color": "white"
             }
            },
            "header": {
             "fill": {
              "color": "#C8D4E3"
             },
             "line": {
              "color": "white"
             }
            },
            "type": "table"
           }
          ]
         },
         "layout": {
          "annotationdefaults": {
           "arrowcolor": "#2a3f5f",
           "arrowhead": 0,
           "arrowwidth": 1
          },
          "autotypenumbers": "strict",
          "coloraxis": {
           "colorbar": {
            "outlinewidth": 0,
            "ticks": ""
           }
          },
          "colorscale": {
           "diverging": [
            [
             0,
             "#8e0152"
            ],
            [
             0.1,
             "#c51b7d"
            ],
            [
             0.2,
             "#de77ae"
            ],
            [
             0.3,
             "#f1b6da"
            ],
            [
             0.4,
             "#fde0ef"
            ],
            [
             0.5,
             "#f7f7f7"
            ],
            [
             0.6,
             "#e6f5d0"
            ],
            [
             0.7,
             "#b8e186"
            ],
            [
             0.8,
             "#7fbc41"
            ],
            [
             0.9,
             "#4d9221"
            ],
            [
             1,
             "#276419"
            ]
           ],
           "sequential": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ],
           "sequentialminus": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ]
          },
          "colorway": [
           "#636efa",
           "#EF553B",
           "#00cc96",
           "#ab63fa",
           "#FFA15A",
           "#19d3f3",
           "#FF6692",
           "#B6E880",
           "#FF97FF",
           "#FECB52"
          ],
          "font": {
           "color": "#2a3f5f"
          },
          "geo": {
           "bgcolor": "white",
           "lakecolor": "white",
           "landcolor": "#E5ECF6",
           "showlakes": true,
           "showland": true,
           "subunitcolor": "white"
          },
          "hoverlabel": {
           "align": "left"
          },
          "hovermode": "closest",
          "mapbox": {
           "style": "light"
          },
          "paper_bgcolor": "white",
          "plot_bgcolor": "#E5ECF6",
          "polar": {
           "angularaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "radialaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "scene": {
           "xaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "yaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "zaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           }
          },
          "shapedefaults": {
           "line": {
            "color": "#2a3f5f"
           }
          },
          "ternary": {
           "aaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "baxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "caxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "title": {
           "x": 0.05
          },
          "xaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          },
          "yaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          }
         }
        },
        "xaxis": {
         "anchor": "y",
         "constrain": "domain",
         "domain": [
          0,
          1
         ],
         "scaleanchor": "y"
        },
        "yaxis": {
         "anchor": "x",
         "autorange": "reversed",
         "constrain": "domain",
         "domain": [
          0,
          1
         ]
        }
       }
      }
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "imshow(model.layers[0].weight.T)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.plotly.v1+json": {
       "config": {
        "plotlyServerURL": "https://plot.ly"
       },
       "data": [
        {
         "coloraxis": "coloraxis",
         "hovertemplate": "x: %{x}<br>y: %{y}<br>color: %{z}<extra></extra>",
         "name": "0",
         "type": "heatmap",
         "xaxis": "x",
         "yaxis": "y",
         "z": [
          [
           -0.93458491563797
          ],
          [
           0.6133294701576233
          ]
         ]
        }
       ],
       "layout": {
        "coloraxis": {
         "cmid": 0,
         "colorscale": [
          [
           0,
           "rgb(103,0,31)"
          ],
          [
           0.1,
           "rgb(178,24,43)"
          ],
          [
           0.2,
           "rgb(214,96,77)"
          ],
          [
           0.3,
           "rgb(244,165,130)"
          ],
          [
           0.4,
           "rgb(253,219,199)"
          ],
          [
           0.5,
           "rgb(247,247,247)"
          ],
          [
           0.6,
           "rgb(209,229,240)"
          ],
          [
           0.7,
           "rgb(146,197,222)"
          ],
          [
           0.8,
           "rgb(67,147,195)"
          ],
          [
           0.9,
           "rgb(33,102,172)"
          ],
          [
           1,
           "rgb(5,48,97)"
          ]
         ]
        },
        "margin": {
         "t": 60
        },
        "template": {
         "data": {
          "bar": [
           {
            "error_x": {
             "color": "#2a3f5f"
            },
            "error_y": {
             "color": "#2a3f5f"
            },
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "bar"
           }
          ],
          "barpolar": [
           {
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "barpolar"
           }
          ],
          "carpet": [
           {
            "aaxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "baxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "type": "carpet"
           }
          ],
          "choropleth": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "choropleth"
           }
          ],
          "contour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "contour"
           }
          ],
          "contourcarpet": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "contourcarpet"
           }
          ],
          "heatmap": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmap"
           }
          ],
          "heatmapgl": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmapgl"
           }
          ],
          "histogram": [
           {
            "marker": {
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "histogram"
           }
          ],
          "histogram2d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2d"
           }
          ],
          "histogram2dcontour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2dcontour"
           }
          ],
          "mesh3d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "mesh3d"
           }
          ],
          "parcoords": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "parcoords"
           }
          ],
          "pie": [
           {
            "automargin": true,
            "type": "pie"
           }
          ],
          "scatter": [
           {
            "fillpattern": {
             "fillmode": "overlay",
             "size": 10,
             "solidity": 0.2
            },
            "type": "scatter"
           }
          ],
          "scatter3d": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter3d"
           }
          ],
          "scattercarpet": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattercarpet"
           }
          ],
          "scattergeo": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergeo"
           }
          ],
          "scattergl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergl"
           }
          ],
          "scattermapbox": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattermapbox"
           }
          ],
          "scatterpolar": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolar"
           }
          ],
          "scatterpolargl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolargl"
           }
          ],
          "scatterternary": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterternary"
           }
          ],
          "surface": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "surface"
           }
          ],
          "table": [
           {
            "cells": {
             "fill": {
              "color": "#EBF0F8"
             },
             "line": {
              "color": "white"
             }
            },
            "header": {
             "fill": {
              "color": "#C8D4E3"
             },
             "line": {
              "color": "white"
             }
            },
            "type": "table"
           }
          ]
         },
         "layout": {
          "annotationdefaults": {
           "arrowcolor": "#2a3f5f",
           "arrowhead": 0,
           "arrowwidth": 1
          },
          "autotypenumbers": "strict",
          "coloraxis": {
           "colorbar": {
            "outlinewidth": 0,
            "ticks": ""
           }
          },
          "colorscale": {
           "diverging": [
            [
             0,
             "#8e0152"
            ],
            [
             0.1,
             "#c51b7d"
            ],
            [
             0.2,
             "#de77ae"
            ],
            [
             0.3,
             "#f1b6da"
            ],
            [
             0.4,
             "#fde0ef"
            ],
            [
             0.5,
             "#f7f7f7"
            ],
            [
             0.6,
             "#e6f5d0"
            ],
            [
             0.7,
             "#b8e186"
            ],
            [
             0.8,
             "#7fbc41"
            ],
            [
             0.9,
             "#4d9221"
            ],
            [
             1,
             "#276419"
            ]
           ],
           "sequential": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ],
           "sequentialminus": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ]
          },
          "colorway": [
           "#636efa",
           "#EF553B",
           "#00cc96",
           "#ab63fa",
           "#FFA15A",
           "#19d3f3",
           "#FF6692",
           "#B6E880",
           "#FF97FF",
           "#FECB52"
          ],
          "font": {
           "color": "#2a3f5f"
          },
          "geo": {
           "bgcolor": "white",
           "lakecolor": "white",
           "landcolor": "#E5ECF6",
           "showlakes": true,
           "showland": true,
           "subunitcolor": "white"
          },
          "hoverlabel": {
           "align": "left"
          },
          "hovermode": "closest",
          "mapbox": {
           "style": "light"
          },
          "paper_bgcolor": "white",
          "plot_bgcolor": "#E5ECF6",
          "polar": {
           "angularaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "radialaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "scene": {
           "xaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "yaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "zaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           }
          },
          "shapedefaults": {
           "line": {
            "color": "#2a3f5f"
           }
          },
          "ternary": {
           "aaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "baxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "caxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "title": {
           "x": 0.05
          },
          "xaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          },
          "yaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          }
         }
        },
        "xaxis": {
         "anchor": "y",
         "constrain": "domain",
         "domain": [
          0,
          1
         ],
         "scaleanchor": "y"
        },
        "yaxis": {
         "anchor": "x",
         "autorange": "reversed",
         "constrain": "domain",
         "domain": [
          0,
          1
         ]
        }
       }
      }
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "imshow(model.layers[-1].weight.T)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[0.8866, 1.0961]], grad_fn=<MmBackward0>)"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.layers[-1].weight @ model.layers[0].weight"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "interp",
   "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": "3.11.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}