_testing.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.functional as F\n",
    "import torch.nn as nn\n",
    "import torch.optim as optim\n",
    "\n",
    "import torchaudio\n",
    "import torchaudio.transforms as T\n",
    "import matplotlib.pyplot as plt\n",
    "import librosa\n",
    "\n",
    "from src.utility import EDA\n",
    "eda = EDA()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/vikram-linux/anaconda3/envs/drum-tracker/lib/python3.12/site-packages/torchaudio/functional/functional.py:584: UserWarning: At least one mel filterbank has all zero values. The value for `n_mels` (256) may be set too high. Or, the value for `n_freqs` (129) may be set too low.\n",
      "  warnings.warn(\n"
     ]
    }
   ],
   "source": [
    "from src.classes import DrumTrackerDataset\n",
    "dataset = DrumTrackerDataset()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([1, 256, 6])"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "signal, label = dataset.__getitem__(0)\n",
    "signal.size()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1 <class 'numpy.int64'>\n"
     ]
    }
   ],
   "source": [
    "print(label, type(label))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "eda.plot_spectrogram(signal.squeeze(0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "from src.classes import ModelTrainer, CNN_Model\n",
    "cnn_model = CNN_Model()\n",
    "loss_fn = nn.CrossEntropyLoss()\n",
    "optim_fn = optim.Adam(params=cnn_model.parameters(), lr=0.1)\n",
    "trainer = ModelTrainer(cnn_model, loss_fn, optim_fn, dataset)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# trainer.train_model(EPOCHS=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/vikram-linux/anaconda3/envs/drum-tracker/lib/python3.12/site-packages/torchaudio/functional/functional.py:584: UserWarning: At least one mel filterbank has all zero values. The value for `n_mels` (512) may be set too high. Or, the value for `n_freqs` (129) may be set too low.\n",
      "  warnings.warn(\n"
     ]
    }
   ],
   "source": [
    "mel_spec_trans = T.MelSpectrogram(\n",
    "    sample_rate = 16000,\n",
    "    n_fft = 256,\n",
    "    hop_length = 256 // 8,\n",
    "    n_mels = 512\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([1, 160])\n",
      "Rate = 16000\n"
     ]
    }
   ],
   "source": [
    "signal, rate = torchaudio.load('data/clean/hi-hat_100.wav')\n",
    "print(signal.size())\n",
    "print(f'Rate = {rate}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([  1,   1, 512,   6])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mel_spec = mel_spec_trans(signal).unsqueeze(0)\n",
    "input_size = torch.tensor(mel_spec.size())\n",
    "input_size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "eda.plot_spectrogram(mel_spec.squeeze(0).squeeze(0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'histogram')"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "hist_signal = mel_spec.squeeze(0).squeeze(0).numpy()\n",
    "fig, ax = plt.subplots(figsize=(10,4))\n",
    "n,bins,patches = ax.hist(hist_signal, edgecolor='none')\n",
    "ax.set_title(\"histogram\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Conv1 Ouput = torch.Size([1, 16, 257, 4])\n",
      "Conv2 Ouput = torch.Size([1, 32, 129, 3])\n",
      "Conv3 Ouput = torch.Size([1, 64, 65, 2])\n",
      "Conv4 Ouput = torch.Size([1, 128, 33, 2])\n",
      "8448\n",
      "Flatten Output = torch.Size([1, 8448])\n",
      "Linear1 Output = torch.Size([1, 128])\n",
      "Linear2 Output = torch.Size([1, 16])\n",
      "Linear3 Output = torch.Size([1, 3])\n"
     ]
    }
   ],
   "source": [
    "conv1 = nn.Conv2d(\n",
    "            in_channels = 1,\n",
    "            out_channels = 16,\n",
    "            kernel_size = 3,\n",
    "            stride = 1,\n",
    "            padding = 2\n",
    "        )\n",
    "\n",
    "conv2 = nn.Conv2d(\n",
    "            in_channels = 16,\n",
    "            out_channels = 32,\n",
    "            kernel_size = 3,\n",
    "            stride = 1,\n",
    "            padding = 2\n",
    "        )\n",
    "\n",
    "conv3 = nn.Conv2d(\n",
    "            in_channels = 32,\n",
    "            out_channels = 64,\n",
    "            kernel_size = 3,\n",
    "            stride = 1,\n",
    "            padding = 2\n",
    "        )\n",
    "\n",
    "conv4 = nn.Conv2d(\n",
    "            in_channels = 64,\n",
    "            out_channels = 128,\n",
    "            kernel_size = 3,\n",
    "            stride = 1,\n",
    "            padding = 2\n",
    "        )\n",
    "\n",
    "relu = nn.ReLU()\n",
    "maxpool =  nn.MaxPool2d(kernel_size=2)\n",
    "\n",
    "flatten = nn.Flatten()\n",
    "\n",
    "output = maxpool(relu(conv1(mel_spec)))\n",
    "print(f'Conv1 Ouput = {output.size()}')\n",
    "\n",
    "output = maxpool(relu(conv2(output)))\n",
    "print(f'Conv2 Ouput = {output.size()}')\n",
    "\n",
    "output = maxpool(relu(conv3(output)))\n",
    "print(f'Conv3 Ouput = {output.size()}')\n",
    "\n",
    "output = maxpool(relu(conv4(output)))\n",
    "print(f'Conv4 Ouput = {output.size()}')\n",
    "print(output.size()[1] * output.size()[2] * output.size()[3])\n",
    "\n",
    "linear_in_features = output.size()[1] * output.size()[2] * output.size()[3]\n",
    "linear1 = nn.Linear(in_features=linear_in_features, out_features=128)\n",
    "linear2 = nn.Linear(in_features=128, out_features=16)\n",
    "linear3 = nn.Linear(in_features=16, out_features=3)\n",
    "\n",
    "output = flatten(output)\n",
    "print(f'Flatten Output = {output.size()}')\n",
    "\n",
    "output = linear1(output)\n",
    "print(f'Linear1 Output = {output.size()}')\n",
    "\n",
    "output = linear2(output)\n",
    "print(f'Linear2 Output = {output.size()}')\n",
    "\n",
    "output = linear3(output)\n",
    "print(f'Linear3 Output = {output.size()}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[0.2862, 0.3677, 0.3461]], grad_fn=<SoftmaxBackward0>)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "torch.softmax(output, dim=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([1])"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "torch.softmax(output, dim=1).argmax(dim=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **MODEL-02**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Conv1 Ouput = torch.Size([1, 16, 257, 4])\n",
      "Conv2 Ouput = torch.Size([1, 32, 129, 3])\n",
      "12384\n",
      "Flatten Output = torch.Size([1, 12384])\n",
      "Linear1 Output = torch.Size([1, 2048])\n",
      "Linear2 Output = torch.Size([1, 512])\n",
      "Linear3 Output = torch.Size([1, 64])\n",
      "Linear4 Output = torch.Size([1, 3])\n"
     ]
    }
   ],
   "source": [
    "conv1 = nn.Conv2d(\n",
    "            in_channels = 1,\n",
    "            out_channels = 16,\n",
    "            kernel_size = 3,\n",
    "            stride = 1,\n",
    "            padding = 2\n",
    "        )\n",
    "\n",
    "conv2 = nn.Conv2d(\n",
    "            in_channels = 16,\n",
    "            out_channels = 32,\n",
    "            kernel_size = 3,\n",
    "            stride = 1,\n",
    "            padding = 2\n",
    "        )\n",
    "\n",
    "relu = nn.ReLU()\n",
    "maxpool =  nn.MaxPool2d(kernel_size=2)\n",
    "\n",
    "flatten = nn.Flatten()\n",
    "\n",
    "output = maxpool(relu(conv1(mel_spec)))\n",
    "print(f'Conv1 Ouput = {output.size()}')\n",
    "\n",
    "output = maxpool(relu(conv2(output)))\n",
    "print(f'Conv2 Ouput = {output.size()}')\n",
    "\n",
    "print(output.size()[1] * output.size()[2] * output.size()[3])\n",
    "\n",
    "linear_in_features = output.size()[1] * output.size()[2] * output.size()[3]\n",
    "linear1 = nn.Linear(in_features=linear_in_features, out_features=2048)\n",
    "linear2 = nn.Linear(in_features=2048, out_features=512)\n",
    "linear3 = nn.Linear(in_features=512, out_features=64)\n",
    "linear4 = nn.Linear(in_features=64, out_features=3)\n",
    "\n",
    "output = flatten(output)\n",
    "print(f'Flatten Output = {output.size()}')\n",
    "\n",
    "output = linear1(output)\n",
    "print(f'Linear1 Output = {output.size()}')\n",
    "\n",
    "output = linear2(output)\n",
    "print(f'Linear2 Output = {output.size()}')\n",
    "\n",
    "output = linear3(output)\n",
    "print(f'Linear3 Output = {output.size()}')\n",
    "\n",
    "output = linear4(output)\n",
    "print(f'Linear4 Output = {output.size()}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[0.3263, 0.3236, 0.3501]], grad_fn=<SoftmaxBackward0>)"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "torch.softmax(output, dim=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model-03"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Conv1 Ouput = torch.Size([1, 16, 514, 8])\n",
      "Conv2 Ouput = torch.Size([1, 32, 258, 5])\n",
      "Conv3 Ouput = torch.Size([1, 64, 260, 7])\n",
      "Conv4 Ouput = torch.Size([1, 128, 131, 4])\n",
      "67072\n",
      "Flatten Output = torch.Size([1, 67072])\n",
      "Linear1 Output = torch.Size([1, 128])\n",
      "Linear2 Output = torch.Size([1, 16])\n",
      "Linear3 Output = torch.Size([1, 3])\n",
      "tensor([[0.3544, 0.3737, 0.2719]], grad_fn=<SoftmaxBackward0>)\n"
     ]
    }
   ],
   "source": [
    "conv1 = nn.Conv2d(\n",
    "            in_channels = 1,\n",
    "            out_channels = 16,\n",
    "            kernel_size = 3,\n",
    "            stride = 1,\n",
    "            padding = 2\n",
    "        )\n",
    "\n",
    "conv2 = nn.Conv2d(\n",
    "            in_channels = 16,\n",
    "            out_channels = 32,\n",
    "            kernel_size = 3,\n",
    "            stride = 1,\n",
    "            padding = 2\n",
    "        )\n",
    "\n",
    "conv3 = nn.Conv2d(\n",
    "            in_channels = 32,\n",
    "            out_channels = 64,\n",
    "            kernel_size = 3,\n",
    "            stride = 1,\n",
    "            padding = 2\n",
    "        )\n",
    "\n",
    "conv4 = nn.Conv2d(\n",
    "            in_channels = 64,\n",
    "            out_channels = 128,\n",
    "            kernel_size = 3,\n",
    "            stride = 1,\n",
    "            padding = 2\n",
    "        )\n",
    "\n",
    "relu = nn.ReLU()\n",
    "maxpool =  nn.MaxPool2d(kernel_size=2)\n",
    "dropout = nn.Dropout(p=0.2)\n",
    "\n",
    "flatten = nn.Flatten()\n",
    "\n",
    "output = relu(conv1(mel_spec))\n",
    "print(f'Conv1 Ouput = {output.size()}')\n",
    "\n",
    "output = maxpool(relu(conv2(output)))\n",
    "print(f'Conv2 Ouput = {output.size()}')\n",
    "\n",
    "output = relu(conv3(output))\n",
    "print(f'Conv3 Ouput = {output.size()}')\n",
    "\n",
    "output = maxpool(relu(conv4(output)))\n",
    "print(f'Conv4 Ouput = {output.size()}')\n",
    "\n",
    "output = dropout(output)\n",
    "\n",
    "print(output.size()[1] * output.size()[2] * output.size()[3])\n",
    "\n",
    "linear_in_features = output.size()[1] * output.size()[2] * output.size()[3]\n",
    "linear1 = nn.Linear(in_features=linear_in_features, out_features=128)\n",
    "linear2 = nn.Linear(in_features=128, out_features=16)\n",
    "linear3 = nn.Linear(in_features=16, out_features=3)\n",
    "\n",
    "output = flatten(output)\n",
    "print(f'Flatten Output = {output.size()}')\n",
    "\n",
    "output = linear1(output)\n",
    "print(f'Linear1 Output = {output.size()}')\n",
    "\n",
    "output = linear2(output)\n",
    "print(f'Linear2 Output = {output.size()}')\n",
    "\n",
    "output = linear3(output)\n",
    "print(f'Linear3 Output = {output.size()}')\n",
    "\n",
    "print(torch.softmax(output, dim=1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[0.2338, 0.2877, 0.1577]], grad_fn=<AddmmBackward0>)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "output"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>epochs</th>\n",
       "      <th>train_loss</th>\n",
       "      <th>train_acc</th>\n",
       "      <th>test_loss</th>\n",
       "      <th>test_acc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1.214711</td>\n",
       "      <td>0.600965</td>\n",
       "      <td>0.744436</td>\n",
       "      <td>0.656418</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>0.720597</td>\n",
       "      <td>0.688654</td>\n",
       "      <td>0.726441</td>\n",
       "      <td>0.669980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>0.743644</td>\n",
       "      <td>0.705917</td>\n",
       "      <td>0.688514</td>\n",
       "      <td>0.676241</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>0.630463</td>\n",
       "      <td>0.733223</td>\n",
       "      <td>0.549975</td>\n",
       "      <td>0.790768</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>1.037260</td>\n",
       "      <td>0.682589</td>\n",
       "      <td>0.654108</td>\n",
       "      <td>0.740847</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>95</th>\n",
       "      <td>96</td>\n",
       "      <td>0.273100</td>\n",
       "      <td>0.887349</td>\n",
       "      <td>0.314306</td>\n",
       "      <td>0.881710</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>97</td>\n",
       "      <td>0.269736</td>\n",
       "      <td>0.890367</td>\n",
       "      <td>0.314331</td>\n",
       "      <td>0.876263</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>98</td>\n",
       "      <td>0.269581</td>\n",
       "      <td>0.888023</td>\n",
       "      <td>0.313817</td>\n",
       "      <td>0.876909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>98</th>\n",
       "      <td>99</td>\n",
       "      <td>0.268185</td>\n",
       "      <td>0.890550</td>\n",
       "      <td>0.312901</td>\n",
       "      <td>0.879212</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99</th>\n",
       "      <td>100</td>\n",
       "      <td>0.270909</td>\n",
       "      <td>0.887990</td>\n",
       "      <td>0.314728</td>\n",
       "      <td>0.877218</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>100 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    epochs  train_loss  train_acc  test_loss  test_acc\n",
       "0        1    1.214711   0.600965   0.744436  0.656418\n",
       "1        2    0.720597   0.688654   0.726441  0.669980\n",
       "2        3    0.743644   0.705917   0.688514  0.676241\n",
       "3        4    0.630463   0.733223   0.549975  0.790768\n",
       "4        5    1.037260   0.682589   0.654108  0.740847\n",
       "..     ...         ...        ...        ...       ...\n",
       "95      96    0.273100   0.887349   0.314306  0.881710\n",
       "96      97    0.269736   0.890367   0.314331  0.876263\n",
       "97      98    0.269581   0.888023   0.313817  0.876909\n",
       "98      99    0.268185   0.890550   0.312901  0.879212\n",
       "99     100    0.270909   0.887990   0.314728  0.877218\n",
       "\n",
       "[100 rows x 5 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('results/model_02_1716836555.csv')\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(df['epochs'], df['train_acc'], label='Train Acc')\n",
    "plt.plot(df['epochs'], df['test_acc'], label='Test Acc')\n",
    "plt.xlabel('Epoch')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.legend()\n",
    "plt.title('Train vs. Test Accuracy')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(df['epochs'], df['train_loss'], label='Train Loss')\n",
    "plt.plot(df['epochs'], df['test_loss'], label='Test Loss')\n",
    "plt.xlabel('Epoch')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend()\n",
    "plt.title('Train vs. Test Loss')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "drum-tracker",
   "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": "3.12.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}