02. Music Generation.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import mitdeeplearning as mdl\n",
    "import numpy as np\n",
    "import os\n",
    "import time\n",
    "import functools\n",
    "from IPython import display as ipydisplay\n",
    "from tqdm import tqdm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "assert len(tf.config.list_logical_devices('GPU')) > 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From C:\\Users\\zakis\\AppData\\Local\\Temp\\ipykernel_14292\\337460670.py:1: is_gpu_available (from tensorflow.python.framework.test_util) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use `tf.config.list_physical_devices('GPU')` instead.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tf.test.is_gpu_available()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 1. Downloading Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 817 songs in text\n",
      "\n",
      "Example song: \n",
      "X:1\n",
      "T:Alexander's\n",
      "Z: id:dc-hornpipe-1\n",
      "M:C|\n",
      "L:1/8\n",
      "K:D Major\n",
      "(3ABc|dAFA DFAd|fdcd FAdf|gfge fefd|(3efe (3dcB A2 (3ABc|!\n",
      "dAFA DFAd|fdcd FAdf|gfge fefd|(3efe dc d2:|!\n",
      "AG|FAdA FAdA|GBdB GBdB|Acec Acec|dfaf gecA|!\n",
      "FAdA FAdA|GBdB GBdB|Aceg fefd|(3efe dc d2:|!\n"
     ]
    }
   ],
   "source": [
    "songs = mdl.lab1.load_training_data()\n",
    "\n",
    "example_song = songs[0]\n",
    "print(\"\\nExample song: \")\n",
    "print(example_song)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# convert ABC notation to audio file and listen\n",
    "mdl.lab1.play_song(example_song)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "there are  83  unique char in dataset\n"
     ]
    }
   ],
   "source": [
    "# combine all song string\n",
    "songs_combined = \"\\n\\n\".join(songs)\n",
    "\n",
    "# finding unique character\n",
    "vocab = sorted(set(songs_combined))\n",
    "print(\"there are \", len(vocab), \" unique char in dataset\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2. Processing Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# create 2 lookup table for numerical representation for dataset\n",
    "\n",
    "# from text to numerical lookup table\n",
    "char2id = {u:i for i, u in enumerate(vocab)}\n",
    "\n",
    "# from numerical to text lookup table\n",
    "id2char = np.array(vocab)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{\n",
      "  '\\n':   0,\n",
      "  ' ' :   1,\n",
      "  '!' :   2,\n",
      "  '\"' :   3,\n",
      "  '#' :   4,\n",
      "  \"'\" :   5,\n",
      "  '(' :   6,\n",
      "  ')' :   7,\n",
      "  ',' :   8,\n",
      "  '-' :   9,\n",
      "  '.' :  10,\n",
      "  '/' :  11,\n",
      "  '0' :  12,\n",
      "  '1' :  13,\n",
      "  '2' :  14,\n",
      "  '3' :  15,\n",
      "  '4' :  16,\n",
      "  '5' :  17,\n",
      "  '6' :  18,\n",
      "  '7' :  19,\n",
      "  ...\n",
      "}\n"
     ]
    }
   ],
   "source": [
    "# checking table lookup\n",
    "print('{')\n",
    "for char,_ in zip(char2id, range(20)):\n",
    "    print('  {:4s}: {:3d},'.format(repr(char), char2id[char]))\n",
    "print('  ...\\n}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# function to vectorize the song\n",
    "\n",
    "def vectorize_string(song):\n",
    "    char_id = []\n",
    "    for n in song:\n",
    "        char_id.append(char2id[n])\n",
    "\n",
    "    return np.array(char_id)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "vectorized_songs = vectorize_string(songs_combined)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'X:1\\nT:Alex' ---- characters mapped to int ----> [49 22 13  0 45 22 26 67 60 79]\n"
     ]
    }
   ],
   "source": [
    "print ('{} ---- characters mapped to int ----> {}'.format(repr(songs_combined[:10]), vectorized_songs[:10]))\n",
    "# check that vectorized_songs is a numpy array\n",
    "assert isinstance(vectorized_songs, np.ndarray), \"returned result should be a numpy array\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# batch funct to create training example\n",
    "\n",
    "def get_batch(vectorized_songs, seq_length, batch_size):\n",
    "    length_vectors = vectorized_songs.shape[0] - 1\n",
    "    \n",
    "    # random index for choosing the starting character\n",
    "    # choice(range angka yang mau dipilih, jumlah yang ingin dipilih)\n",
    "    random_index = np.random.choice(length_vectors-seq_length, batch_size)\n",
    "\n",
    "    # the input and output take from random_index, so it uses i instead of value range\n",
    "    # input batch = len(vector) - 1\n",
    "    input_batch = [vectorized_songs[i:i+seq_length] for i in random_index]\n",
    "\n",
    "    # output batch = vector[1:-1]\n",
    "    output_batch = [vectorized_songs[i+1:i+seq_length+1] for i in random_index]\n",
    "\n",
    "    # x_batch and y_batch provide input and output for model\n",
    "    x_batch = np.reshape(input_batch, [batch_size, seq_length])\n",
    "    y_batch = np.reshape(output_batch, [batch_size, seq_length])\n",
    "    \n",
    "    return x_batch, y_batch\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[PASS] test_batch_func_types\n",
      "[PASS] test_batch_func_shapes\n",
      "[PASS] test_batch_func_next_step\n",
      "======\n",
      "[PASS] passed all tests!\n"
     ]
    }
   ],
   "source": [
    "# testing input batch function\n",
    "test_args = (vectorized_songs, 10, 2)\n",
    "if not mdl.lab1.test_batch_func_types(get_batch, test_args) or \\\n",
    "   not mdl.lab1.test_batch_func_shapes(get_batch, test_args) or \\\n",
    "   not mdl.lab1.test_batch_func_next_step(get_batch, test_args): \n",
    "   print(\"======\\n[FAIL] could not pass tests\")\n",
    "else: \n",
    "   print(\"======\\n[PASS] passed all tests!\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Step   0\n",
      "  input: 32 ('G')\n",
      "  expected output: 27 ('B')\n",
      "Step   1\n",
      "  input: 27 ('B')\n",
      "  expected output: 59 ('d')\n",
      "Step   2\n",
      "  input: 59 ('d')\n",
      "  expected output: 82 ('|')\n",
      "Step   3\n",
      "  input: 82 ('|')\n",
      "  expected output: 60 ('e')\n",
      "Step   4\n",
      "  input: 60 ('e')\n",
      "  expected output: 62 ('g')\n"
     ]
    }
   ],
   "source": [
    "x_batch, y_batch = get_batch(vectorized_songs, seq_length=5, batch_size=1)\n",
    "\n",
    "for i, (input_idx, target_idx) in enumerate(zip(np.squeeze(x_batch), np.squeeze(y_batch))):\n",
    "    print(\"Step {:3d}\".format(i))\n",
    "    print(\"  input: {} ({:s})\".format(input_idx, repr(id2char[input_idx])))\n",
    "    print(\"  expected output: {} ({:s})\".format(target_idx, repr(id2char[target_idx])))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3. Recurrent Neural Network"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- the model based on LSTM architecture\n",
    "- model train in batch\n",
    "- output of LSTM then fed to fully connected Dense Layer with softmax activation function\n",
    "- 3 layers that will be use:\n",
    "    1. tf.keras.layers.Embedding : interact as input layer that consist of trainable lookup table (id2char) with embedding_dim dimensions\n",
    "    2. tf.keras.layers.LSTM : network architecture that will be use with size units=rnn_units\n",
    "    3. tf.keras.layers.Dense : interact as output layer with vocab_size outputs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# define model\n",
    "\n",
    "def LSTM(rnn_units):\n",
    "    return tf.keras.layers.LSTM(\n",
    "        rnn_units,\n",
    "        return_sequences=True,\n",
    "        recurrent_initializer='glorot_uniform',\n",
    "        recurrent_activation='sigmoid',\n",
    "        stateful=True\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# define RNN model\n",
    "\n",
    "def build_model(vocab_size, embedding_dim, rnn_units, batch_size):\n",
    "    model = tf.keras.Sequential([\n",
    "        tf.keras.layers.Embedding(vocab_size, embedding_dim, batch_input_shape=[batch_size, None]),\n",
    "        LSTM(rnn_units),\n",
    "        tf.keras.layers.Dense(vocab_size, activation='sigmoid')\n",
    "    ])\n",
    "\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = build_model(len(vocab), embedding_dim=256, rnn_units=1024, batch_size=32)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding (Embedding)       (32, None, 256)           21248     \n",
      "                                                                 \n",
      " lstm (LSTM)                 (32, None, 1024)          5246976   \n",
      "                                                                 \n",
      " dense (Dense)               (32, None, 83)            85075     \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 5,353,299\n",
      "Trainable params: 5,353,299\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# testing\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input shape:       (32, 100)  # (batch_size, sequence_length)\n",
      "Prediction shape:  (32, 100, 83) # (batch_size, sequence_length, vocab_size)\n"
     ]
    }
   ],
   "source": [
    "x, y = get_batch(vectorized_songs, seq_length=100, batch_size=32)\n",
    "\n",
    "pred = model(x)\n",
    "print(\"Input shape:      \", x.shape, \" # (batch_size, sequence_length)\")\n",
    "print(\"Prediction shape: \", pred.shape, \"# (batch_size, sequence_length, vocab_size)\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([31, 25, 34, 50,  8, 15, 78, 38, 42, 71, 49, 18, 29, 57, 44, 53, 17,\n",
       "       10, 64, 13, 23, 48, 44, 48, 25, 51, 35, 11, 71, 56, 79, 63, 73, 54,\n",
       "       70, 71, 58, 80, 79, 63, 25, 44, 29, 49, 64,  8, 81, 39, 60,  0, 27,\n",
       "       41, 76, 64, 53, 66, 50, 56, 36, 79, 35, 11, 71, 25,  7, 36, 72, 13,\n",
       "       15, 40, 79, 58, 31, 32, 37,  7, 45, 62, 64, 30, 70, 80, 41, 61, 76,\n",
       "       59, 36, 67, 49, 52, 75, 68, 44, 74, 11, 45, 12, 60,  9, 29],\n",
       "      dtype=int64)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# untrained model prediction\n",
    "sample_indices = tf.random.categorical(pred[0], num_samples=1)\n",
    "sample_indices = tf.squeeze(sample_indices, axis=1).numpy()\n",
    "sample_indices"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input: \n",
      " '2BG AGEG|cBcd efge|dBAB G2:|!\\nBd|g2dg egde|g2bg ageg|a2ea fae^g|a2bg agef|!\\ng2dg egde|g2bg aged|cBcd'\n",
      "\n",
      "Next Char Predictions: \n",
      " 'F>IY,3wMQpX6DbS]5.i1<WSW>ZJ/paxhr^opcyxh>SDXi,zNe\\nBPui]kYaKxJ/p>)Kq13OxcFGL)TgiEoyPfudKlX[tmSs/T0e-D'\n"
     ]
    }
   ],
   "source": [
    "print(\"Input: \\n\", repr(\"\".join(id2char[x[0]])))\n",
    "print()\n",
    "print(\"Next Char Predictions: \\n\", repr(\"\".join(id2char[sample_indices])))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "the resulting text from untrained model is nonsensical"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4. Training Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\zakis\\anaconda3\\envs\\tf1\\lib\\site-packages\\tensorflow\\python\\util\\dispatch.py:1082: UserWarning: \"`sparse_categorical_crossentropy` received `from_logits=True`, but the `output` argument was produced by a sigmoid or softmax activation and thus does not represent logits. Was this intended?\"\n",
      "  return dispatch_target(*args, **kwargs)\n"
     ]
    }
   ],
   "source": [
    "# defining loss function\n",
    "\n",
    "def compute_loss(labels, logits):\n",
    "    loss = tf.keras.losses.sparse_categorical_crossentropy(labels, logits, from_logits=True)\n",
    "    return loss\n",
    "\n",
    "\n",
    "example_batch_loss = compute_loss(y, pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prediction shape:  (32, 100, 83)  # (batch_size, sequence_length, vocab_size)\n",
      "scalar_loss:       4.4193096\n"
     ]
    }
   ],
   "source": [
    "print(\"Prediction shape: \", pred.shape, \" # (batch_size, sequence_length, vocab_size)\") \n",
    "print(\"scalar_loss:      \", example_batch_loss.numpy().mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "# hyperparameter setting and optimization\n",
    "\n",
    "# optimization parameter\n",
    "number_training_iteration = 2000\n",
    "batch_size = 4\n",
    "seq_length = 100\n",
    "learning_rate = 5e-3\n",
    "\n",
    "# model parameter\n",
    "vocab_size = len(vocab)\n",
    "embedding_dim = 256\n",
    "rnn_units = 1024\n",
    "\n",
    "# checkpoint location\n",
    "checkpoint_dir = \"./training_checkpoints\"\n",
    "checkpoint_prefix = os.path.join(checkpoint_dir, \"my_ckpt\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "# define optimizer\n",
    "\n",
    "model = build_model(vocab_size=vocab_size, embedding_dim=embedding_dim, rnn_units=rnn_units, batch_size=batch_size)\n",
    "\n",
    "optimizer = tf.keras.optimizers.Adam(learning_rate)\n",
    "\n",
    "@tf.function\n",
    "def train_step(x, y):\n",
    "    with tf.GradientTape() as tape:\n",
    "        y_hat = model(x)\n",
    "        loss = compute_loss(y, y_hat)\n",
    "    grads = tape.gradient(loss, model.trainable_variables)\n",
    "    optimizer.apply_gradients(zip(grads, model.trainable_variables))\n",
    "    return loss\n",
    "        \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "history = []\n",
    "plotter = mdl.util.PeriodicPlotter(sec=2, xlabel='Iteration', ylabel='Loss')\n",
    "\n",
    "if hasattr(tqdm, '_instances'): tqdm._instances.clear()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 2000/2000 [01:26<00:00, 23.25it/s]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for iter in tqdm(range(number_training_iteration)):\n",
    "    # grab batch and propagate\n",
    "    x_batch, y_batch = get_batch(vectorized_songs, seq_length, batch_size)\n",
    "    loss = train_step(x_batch, y_batch)\n",
    "\n",
    "    # update progress\n",
    "    history.append(loss.numpy().mean())\n",
    "    plotter.plot(history)\n",
    "\n",
    "    # updating model\n",
    "    if iter%100 == 0:\n",
    "        model.save_weights(checkpoint_prefix)\n",
    "\n",
    "\n",
    "model.save_weights(checkpoint_prefix)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 5. Generate Music"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_9\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_9 (Embedding)     (1, None, 256)            21248     \n",
      "                                                                 \n",
      " lstm_9 (LSTM)               (1, None, 1024)           5246976   \n",
      "                                                                 \n",
      " dense_9 (Dense)             (1, None, 83)             85075     \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 5,353,299\n",
      "Trainable params: 5,353,299\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# restoring latest checkpoint\n",
    "model = build_model(vocab_size=vocab_size, embedding_dim=embedding_dim, rnn_units=rnn_units, batch_size=1)\n",
    "model.load_weights(tf.train.latest_checkpoint(checkpoint_dir))\n",
    "model.build(tf.TensorShape([1, None]))\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "the vocab size, embedding dimension, and rnn units should always be the same as the weights that will be restored"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "# predicting song\n",
    "\n",
    "def generate_text(model, start_string, generation_length=1000):\n",
    "    # evaluating state\n",
    "    input_eval = [char2id[n] for n in start_string]\n",
    "    input_eval = tf.expand_dims(input_eval, 0)\n",
    "\n",
    "    # vectorize state\n",
    "    text_generated = []\n",
    "\n",
    "    model.reset_states()\n",
    "    tqdm._instances.clear()\n",
    "\n",
    "    for i in tqdm(range(generation_length)):\n",
    "        predicts = model(input_eval)\n",
    "\n",
    "        # removing batch dims\n",
    "        predicts = tf.squeeze(predicts, 0)\n",
    "\n",
    "        # turn into multinomial distribution\n",
    "        predicts_id = tf.random.categorical(predicts, num_samples=1)[-1, 0].numpy()\n",
    "        input_eval = tf.expand_dims([predicts_id], 0)\n",
    "        text_generated.append(id2char[predicts_id])\n",
    "    return (start_string + ''.join(text_generated))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1000/1000 [00:04<00:00, 210.57it/s]\n"
     ]
    }
   ],
   "source": [
    "generated_text = generate_text(model, start_string=\"me and \", generation_length=1000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 1 songs in text\n"
     ]
    }
   ],
   "source": [
    "# playback the music\n",
    "\n",
    "generated_songs = mdl.lab1.extract_song_snippet(generated_text)\n",
    "\n",
    "for i, song in enumerate(generated_songs):\n",
    "    waveform = mdl.lab1.play_song(song)\n",
    "\n",
    "    if waveform:\n",
    "        print('Generated song', i)\n",
    "        ipythondisplay,display(waveform)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['me and FK P9V=/!K)0d:=C|DP7906N5ZB#QM\"wno0ry2\\n((Fl:):u[vkS9)Q5._Wxbraw:M9y1# K\"idzurgpgrOX_gS6.BH\\'6Lp]rtWhbIvkEC5 AbCFhmiPz426yO\\'\\'-6\\nksD 2PzLdDETS>#fEG9,y NZWT30JEm6,.T5t 1XvfTp5C)pxnt^b\\'ClgABhZ]YvF#drsHDJ.c\\nX7 sGmbD^1P\\'qQi3t|VyUgMbts\\'f<5:uY9 q0 8ppkG3[4bj)[17r\\na\"om:EPX OG]38bo\\'jW|vegBI9UtydZz0PW3CL:eDq0pAdSBfjLpBf9nKCc(qv]LKjmhbyI5cE|qG<FGRzD78']"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "generated_songs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.10.4 ('tf1')",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.4"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "e5d47089378e6e18d5e41b4ec5dc5221a0b09d5a84cf9a8c4af6a05ce24c8606"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}