{ "cells": [ { "cell_type": "markdown", "metadata": { "id": "dgxD_ljQ23dE" }, "source": [ "# Music generation with Variational AutoEncoder" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "jRRn77kD4EiT" }, "outputs": [], "source": [ "!pip install -q git+https://github.com/tensorflow/docs" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "jqomLmiK4RY4", "outputId": "f26beba1-c06d-48db-a3cf-df5bc83efcb0" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n", "Requirement already satisfied: tensorflow-addons in /usr/local/lib/python3.7/dist-packages (0.17.0)\n", "Requirement already satisfied: packaging in /usr/local/lib/python3.7/dist-packages (from tensorflow-addons) (21.3)\n", "Requirement already satisfied: typeguard>=2.7 in /usr/local/lib/python3.7/dist-packages (from tensorflow-addons) (2.7.1)\n", "Requirement already satisfied: pyparsing!=3.0.5,>=2.0.2 in /usr/local/lib/python3.7/dist-packages (from packaging->tensorflow-addons) (3.0.9)\n" ] } ], "source": [ "!pip install tensorflow-addons" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "7DZkI3Fg2yUQ" }, "outputs": [], "source": [ "import librosa\n", "\n", "import numpy as np\n", "import pandas as pd\n", "import os\n", "\n", "import tensorflow as tf\n", "import tensorflow_addons as tfa\n", "from tensorflow.keras import layers \n", "\n", "import matplotlib.pyplot as plt\n", "from IPython import display\n", "from IPython.display import clear_output\n", "\n", "import glob\n", "import imageio\n", "import time\n", "import IPython.display as ipd\n", "\n", "AUTOTUNE = tf.data.experimental.AUTOTUNE" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "AWcjKTUD37z-" }, "outputs": [], "source": [ "seed=123\n", "tf.compat.v1.set_random_seed(seed)\n", "session_conf = tf.compat.v1.ConfigProto(intra_op_parallelism_threads=1, inter_op_parallelism_threads=1)\n", "sess = tf.compat.v1.Session(graph=tf.compat.v1.get_default_graph(), config=session_conf)\n", "tf.compat.v1.keras.backend.set_session(sess)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "hsaYjzBi4BPv" }, "outputs": [], "source": [ "train_size = 60000\n", "BATCH_SIZE = 10\n", "test_size = 10000\n", "epochs = 20\n", "# set the dimensionality of the latent space to a plane for visualization later\n", "latent_dim = 2\n", "num_examples_to_generate = 10\n", "\n", "BASE_PATH = 'drive/MyDrive/music_generation/auto_encoder/genres_original'" ] }, { "cell_type": "markdown", "metadata": { "id": "PM0_HbJd45tx" }, "source": [ "Data preprocessing" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "ChJ3XdWt40ap" }, "outputs": [], "source": [ "def DatasetLoader(class_):\n", " music_list = np.array(sorted(os.listdir(BASE_PATH+'/'+class_)))\n", " train_music_1 = list(music_list[[0,52,19,39,71,12,75,85,3,45,24,46,88]]) #99,10,66,76,41\n", " train_music_2 = list(music_list[[4,43,56,55,45,31,11,13,70,37,21,78]]) #65,32,53,22,19,80,89,\n", " TrackSet_1 = [(BASE_PATH)+'/'+class_+'/%s'%(x) for x in train_music_1]\n", " TrackSet_2 = [(BASE_PATH)+'/'+class_+'/%s'%(x) for x in train_music_2]\n", "\n", " return TrackSet_1, TrackSet_2" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "Z1sM3p7h42Af" }, "outputs": [], "source": [ "def load(file_):\n", " data_, sampling_rate = librosa.load(file_,sr=3000, offset=0.0, duration=30)\n", " data_ = data_.reshape(1,90001)\n", " return data_\n", "map_data = lambda filename: tf.compat.v1.py_func(load, [filename], [tf.float32])" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "h6ipB6js43ME" }, "outputs": [], "source": [ "TrackSet_1, TrackSet_2 = DatasetLoader('jazz')" ] }, { "cell_type": "markdown", "metadata": { "id": "6Cglv0Yc472G" }, "source": [ "sample original music" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 75 }, "id": "yMbqcybe49YH", "outputId": "7cd4c8e8-0b46-4d86-c1fc-961673e9a44b" }, "outputs": [ { "data": { "text/html": [ "\n", " \n", " " ], "text/plain": [ "" ] }, "execution_count": 33, "metadata": {}, "output_type": "execute_result" } ], "source": [ "sample = TrackSet_1[1]\n", "sample_, sampling_rate = librosa.load(sample,sr=3000, offset=0.0, duration=30)\n", "ipd.Audio(sample_,rate=3000)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 232 }, "id": "Z-Nnpz3o4-p_", "outputId": "22194601-ef5d-499a-f4f2-40d3ea8adc82" }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "import librosa.display\n", "plt.figure(figsize=(18,15))\n", "for i in range(4):\n", " plt.subplot(4, 4, i + 1)\n", " j = load(TrackSet_1[i])\n", " librosa.display.waveplot(j[0], sr=3000)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "jjiK-fSw5ArS" }, "outputs": [], "source": [ "train_dataset = (\n", " tf.data.Dataset\n", " .from_tensor_slices((TrackSet_1))\n", " .map(map_data, num_parallel_calls=AUTOTUNE)\n", " .shuffle(3)\n", " .batch(BATCH_SIZE)\n", ")\n", "test_dataset = (\n", " tf.data.Dataset\n", " .from_tensor_slices((TrackSet_2))\n", " .map(map_data, num_parallel_calls=AUTOTUNE)\n", " .shuffle(3)\n", " .batch(BATCH_SIZE)\n", ")" ] }, { "cell_type": "markdown", "metadata": { "id": "siBpQeyR5CwS" }, "source": [ "Network architecture" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "saK3Cvmc5E-x" }, "outputs": [], "source": [ "class Resnet1DBlock(tf.keras.Model):\n", " def __init__(self, kernel_size, filters,type='encode'):\n", " super(Resnet1DBlock, self).__init__()\n", " \n", " if type=='encode':\n", " self.conv1a = layers.Conv1D(filters, kernel_size, 2,padding=\"same\")\n", " self.conv1b = layers.Conv1D(filters, kernel_size, 1,padding=\"same\")\n", " self.norm1a = tfa.layers.InstanceNormalization()\n", " self.norm1b = tfa.layers.InstanceNormalization()\n", " if type=='decode':\n", " self.conv1a = layers.Conv1DTranspose(filters, kernel_size, 1,padding=\"same\")\n", " self.conv1b = layers.Conv1DTranspose(filters, kernel_size, 1,padding=\"same\")\n", " self.norm1a = tf.keras.layers.BatchNormalization()\n", " self.norm1b = tf.keras.layers.BatchNormalization()\n", " else:\n", " return None\n", "\n", " def call(self, input_tensor):\n", " x = tf.nn.relu(input_tensor)\n", " x = self.conv1a(x)\n", " x = self.norm1a(x)\n", " x = layers.LeakyReLU(0.4)(x)\n", "\n", " x = self.conv1b(x)\n", " x = self.norm1b(x)\n", " x = layers.LeakyReLU(0.4)(x)\n", "\n", " x += input_tensor\n", " return tf.nn.relu(x)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "BslaIFfO5HXS" }, "outputs": [], "source": [ "class CVAE(tf.keras.Model):\n", " \"\"\"Convolutional variational autoencoder.\"\"\"\n", "\n", " def __init__(self, latent_dim):\n", " super(CVAE, self).__init__()\n", " self.latent_dim = latent_dim\n", " self.encoder = tf.keras.Sequential(\n", " [\n", " tf.keras.layers.InputLayer(input_shape=(1,90001)),\n", " layers.Conv1D(64,1,2),\n", " Resnet1DBlock(64,1),\n", " layers.Conv1D(128,1,2),\n", " Resnet1DBlock(128,1),\n", " layers.Conv1D(128,1,2),\n", " Resnet1DBlock(128,1),\n", " layers.Conv1D(256,1,2),\n", " Resnet1DBlock(256,1),\n", " # No activation\n", " layers.Flatten(),\n", " layers.Dense(latent_dim+latent_dim)\n", "\n", " ]\n", " )\n", " self.decoder = tf.keras.Sequential(\n", " [\n", " tf.keras.layers.InputLayer(input_shape=(latent_dim,)),\n", " layers.Reshape(target_shape=(1,latent_dim)),\n", " Resnet1DBlock(512,1,'decode'),\n", " layers.Conv1DTranspose(512,1,1),\n", " Resnet1DBlock(256,1,'decode'),\n", " layers.Conv1DTranspose(256,1,1),\n", " Resnet1DBlock(128,1,'decode'),\n", " layers.Conv1DTranspose(128,1,1),\n", " Resnet1DBlock(64,1,'decode'),\n", " layers.Conv1DTranspose(64,1,1),\n", " # No activation\n", " layers.Conv1DTranspose(90001,1,1),\n", " ]\n", " )\n", " @tf.function\n", " def sample(self, eps=None):\n", " if eps is None:\n", " eps = tf.random.normal(shape=(200, self.latent_dim))\n", " return self.decode(eps, apply_sigmoid=True)\n", " @tf.function\n", " def encode(self, x):\n", " mean, logvar = tf.split(self.encoder(x), num_or_size_splits=2, axis=1)\n", " return mean, logvar\n", " @tf.function\n", " def reparameterize(self, mean, logvar):\n", " eps = tf.random.normal(shape=mean.shape)\n", " return eps * tf.exp(logvar * .5) + mean\n", " @tf.function\n", " def decode(self, z, apply_sigmoid=False):\n", " logits = self.decoder(z)\n", " if apply_sigmoid:\n", " probs = tf.sigmoid(logits)\n", " return probs\n", " return logits" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "Lr-dUkGA5JC4" }, "outputs": [], "source": [ "optimizer = tf.keras.optimizers.Adam(0.0003,beta_1=0.9, beta_2=0.999,epsilon=1e-08)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "tA6t8mbC5KDb" }, "outputs": [], "source": [ "@tf.function\n", "def log_normal_pdf(sample, mean, logvar, raxis=1):\n", " log2pi = tf.math.log(2. * np.pi)\n", " return tf.reduce_sum(\n", " -.5 * ((sample - mean) ** 2. * tf.exp(-logvar) + logvar + log2pi),\n", " axis=raxis)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "eJ752zGX5LZ4" }, "outputs": [], "source": [ "@tf.function\n", "def compute_loss(model, x):\n", " mean, logvar = model.encode(x)\n", " z = model.reparameterize(mean, logvar)\n", " x_logit = model.decode(z)\n", " cross_ent = tf.nn.sigmoid_cross_entropy_with_logits(logits=x_logit, labels=x)\n", " logpx_z = -tf.reduce_sum(cross_ent, axis=[1,2])\n", " logpz = log_normal_pdf(z, 0., 0.)\n", " logqz_x = log_normal_pdf(z, mean, logvar)\n", " return -tf.reduce_mean(logpx_z + logpz - logqz_x)" ] }, { "cell_type": "markdown", "metadata": { "id": "PcSo6Y1U5OeZ" }, "source": [ "Reconstruction loss" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "EabyjIiU5OJZ" }, "outputs": [], "source": [ "@tf.function\n", "def train_step(model, x, optimizer):\n", " \n", " \"\"\"Executes one training step and returns the loss.\n", "\n", " This function computes the loss and gradients, and uses the latter to\n", " update the model's parameters.\n", " \"\"\"\n", " with tf.GradientTape() as tape:\n", " mean, logvar = model.encode(x)\n", " z = model.reparameterize(mean, logvar)\n", " x_logit = model.decode(z)\n", " cross_ent = tf.nn.sigmoid_cross_entropy_with_logits(logits=x_logit, labels=x)\n", " logpx_z = -tf.reduce_sum(cross_ent, axis=[1,2])\n", " logpz = log_normal_pdf(z, 0., 0.)\n", " logqz_x = log_normal_pdf(z, mean, logvar)\n", " loss_KL = -tf.reduce_mean(logpx_z + logpz - logqz_x)\n", " reconstruction_loss = tf.reduce_mean(\n", " tf.keras.losses.binary_crossentropy(x, x_logit)\n", " )\n", " total_loss = reconstruction_loss+ loss_KL\n", " gradients = tape.gradient(total_loss, model.trainable_variables)\n", " optimizer.apply_gradients(zip(gradients, model.trainable_variables))" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "u4l4OtOM5R78" }, "outputs": [], "source": [ "# keeping the random vector constant for generation (prediction) so\n", "# it will be easier to see the improvement.\n", "random_vector_for_generation = tf.random.normal(\n", " shape=[num_examples_to_generate, latent_dim])\n", "model = CVAE(latent_dim)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "58NP4C-w5TWJ" }, "outputs": [], "source": [ "import librosa.display\n", "\n", "def generate_and_save_images(model, epoch, test_sample, save):\n", " mean, logvar = model.encode(test_sample)\n", " z = model.reparameterize(mean, logvar)\n", " predictions = model.sample(z)\n", " fig = plt.figure(figsize=(18, 15))\n", "\n", " for i in range(predictions.shape[0]):\n", " plt.subplot(4, 4, i + 1)\n", " wave = np.asarray(predictions[i])\n", " librosa.display.waveplot(wave[0], sr=3000)\n", "\n", " # tight_layout minimizes the overlap between 2 sub-plots\n", " plt.savefig('{}_{:04d}.png'.format(save, epoch))\n", " plt.savefig('{}_{:04d}.png'.format(save, epoch))\n", " plt.show()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "NYi2IaCt5UrX" }, "outputs": [], "source": [ "# Pick a sample of the test set for generating output images\n", "assert BATCH_SIZE >= num_examples_to_generate\n", "for test_batch in test_dataset.take(1):\n", " test_sample = test_batch[0]" ] }, { "cell_type": "markdown", "metadata": { "id": "IfrPv1He5W1Y" }, "source": [ "Training the model" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 662 }, "id": "CZ1TEYXC5ZEF", "outputId": "9d494d7b-4477-48bb-b056-f333c22bf3fe" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Epoch: 20, Test set ELBO: -18109.1796875, time elapse for current epoch: 6.8476269245147705\n" ] }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "generate_and_save_images(model, 0, test_sample, 'jazz')\n", "def train(train_dataset, test_dataset, model, save):\n", " for epoch in range(1, epochs + 1):\n", " start_time = time.time()\n", " for train_x in train_dataset:\n", " train_x = np.asarray(train_x)[0]\n", " train_step(model, train_x, optimizer)\n", " end_time = time.time()\n", "\n", " loss = tf.keras.metrics.Mean()\n", " for test_x in test_dataset:\n", " test_x = np.asarray(test_x)[0]\n", " loss(compute_loss(model, test_x))\n", " display.clear_output(wait=False)\n", " elbo = -loss.result()\n", " print('Epoch: {}, Test set ELBO: {}, time elapse for current epoch: {}'.format(epoch, \n", " elbo, \n", " end_time - start_time\n", " ))\n", " generate_and_save_images(model,\n", " epoch, \n", " test_sample,\n", " save)\n", "train(train_dataset, test_dataset, model, 'jazz')" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "JHK2LPk55bQh" }, "outputs": [], "source": [ "anim_file_1 = 'jazz_cvae.gif'\n", "\n", "with imageio.get_writer(anim_file_1, mode='I') as writer:\n", " filenames = glob.glob('jazz*.png')\n", " filenames = sorted(filenames)\n", " for filename in filenames:\n", " image = imageio.imread(filename)\n", " writer.append_data(image)\n", " image = imageio.imread(filename)\n", " writer.append_data(image)" ] }, { "cell_type": "markdown", "metadata": { "id": "pTEd1Aqs5c7_" }, "source": [ "Visualization" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 1000 }, "id": "JmxhsMzB5eYN", "outputId": "8d995986-a245-467c-db16-759e41b50762" }, "outputs": [ { "data": { "text/html": [ "" ], "text/plain": [ "" ] }, "execution_count": 53, "metadata": {}, "output_type": "execute_result" } ], "source": [ "import tensorflow_docs.vis.embed as embed\n", "embed.embed_file(anim_file_1)" ] }, { "cell_type": "markdown", "metadata": { "id": "kca5NC685gg8" }, "source": [ "Generated Music - Jazz" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "9UJPYwRD5iIT" }, "outputs": [], "source": [ "def inference(test_dataset, model): \n", " save_music = []\n", " for test in test_dataset:\n", " mean, logvar = model.encode(test)\n", " z = model.reparameterize(mean, logvar)\n", " predictions = model.sample(z)\n", " for pred in predictions:\n", " wave = np.asarray(pred)\n", " save_music.append(wave)\n", " return save_music\n", "\n", "saved_musics = inference(test_dataset, model)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 75 }, "id": "Hkx8tzHxJskG", "outputId": "575fe3f9-5736-4596-9026-2748717c26aa" }, "outputs": [ { "data": { "text/html": [ "\n", " \n", " " ], "text/plain": [ "" ] }, "execution_count": 55, "metadata": {}, "output_type": "execute_result" } ], "source": [ "music1=saved_musics[0][0]\n", "ipd.Audio(music1,rate=3000)" ] } ], "metadata": { "colab": { "collapsed_sections": [], "name": "music_generation_with_Variational_AE.ipynb", "provenance": [] }, "kernelspec": { "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.9.12" } }, "nbformat": 4, "nbformat_minor": 1 }