Giant_Music_Transformer_Composer.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "# Giant Music Transformer Composer (ver. 4.0)\n",
        "\n",
        "***\n",
        "\n",
        "Powered by tegridy-tools: https://github.com/asigalov61/tegridy-tools\n",
        "\n",
        "***\n",
        "\n",
        "WARNING: This complete implementation is a functioning model of the Artificial Intelligence. Please excercise great humility, care, and respect. https://www.nscai.gov/\n",
        "\n",
        "***\n",
        "\n",
        "#### Project Los Angeles\n",
        "\n",
        "#### Tegridy Code 2024\n",
        "\n",
        "***"
      ],
      "metadata": {
        "id": "gpy3qsulqHa5"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# (GPU CHECK)"
      ],
      "metadata": {
        "id": "W_So4w8fqPGL"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "X3rABEpKCO02",
        "cellView": "form"
      },
      "outputs": [],
      "source": [
        "#@title NVIDIA GPU check\n",
        "!nvidia-smi"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# (SETUP ENVIRONMENT)"
      ],
      "metadata": {
        "id": "C0XxnXGFqVyh"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "vK40g6V_BTNj",
        "cellView": "form"
      },
      "outputs": [],
      "source": [
        "#@title Install dependencies\n",
        "!git clone --depth 1 https://github.com/asigalov61/Giant-Music-Transformer\n",
        "!pip install -U torch\n",
        "!pip install einops\n",
        "!pip install torch-summary\n",
        "!apt install fluidsynth #Pip does not work for some reason. Only apt works"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "DzCOZU_gBiQV",
        "cellView": "form"
      },
      "outputs": [],
      "source": [
        "#@title Import modules\n",
        "\n",
        "print('=' * 70)\n",
        "print('Loading core Giant Music Transformer modules...')\n",
        "\n",
        "import os\n",
        "import copy\n",
        "import pickle\n",
        "import secrets\n",
        "import statistics\n",
        "from time import time\n",
        "import tqdm\n",
        "\n",
        "print('=' * 70)\n",
        "print('Loading main Giant Music Transformer modules...')\n",
        "\n",
        "os.environ['USE_FLASH_ATTENTION'] = '1'\n",
        "\n",
        "import torch\n",
        "\n",
        "torch.set_float32_matmul_precision('high')\n",
        "torch.backends.cuda.matmul.allow_tf32 = True # allow tf32 on matmul\n",
        "torch.backends.cudnn.allow_tf32 = True # allow tf32 on cudnn\n",
        "torch.backends.cuda.enable_mem_efficient_sdp(True)\n",
        "torch.backends.cuda.enable_math_sdp(True)\n",
        "torch.backends.cuda.enable_flash_sdp(True)\n",
        "torch.backends.cuda.enable_cudnn_sdp(True)\n",
        "\n",
        "%cd /content/Giant-Music-Transformer\n",
        "\n",
        "import TMIDIX\n",
        "\n",
        "from midi_to_colab_audio import midi_to_colab_audio\n",
        "\n",
        "from x_transformer_1_23_2 import *\n",
        "\n",
        "import random\n",
        "\n",
        "%cd /content/\n",
        "print('=' * 70)\n",
        "print('Loading aux Giant Music Transformer modules...')\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "from torchsummary import summary\n",
        "from sklearn import metrics\n",
        "\n",
        "from IPython.display import Audio, display\n",
        "\n",
        "from huggingface_hub import hf_hub_download\n",
        "\n",
        "from google.colab import files\n",
        "\n",
        "print('=' * 70)\n",
        "print('PyTorch version:', torch.__version__)\n",
        "print('=' * 70)\n",
        "print('Done!')\n",
        "print('Enjoy! :)')\n",
        "print('=' * 70)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eI3aQtHzqSnp"
      },
      "source": [
        "# (LOAD MODEL)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "rDquonbXC2je",
        "cellView": "form"
      },
      "outputs": [],
      "source": [
        "#@title Load Giant Music Transformer Pre-Trained Model\n",
        "\n",
        "#@markdown Choose model\n",
        "\n",
        "select_model_to_load = \"482M-8L-Ultra-Fast-Medium\" # @param [\"482M-8L-Ultra-Fast-Medium\",\"585M-32L-Very-Fast-Large\",\"786M-44L-Fast-Extra-Large\"]\n",
        "\n",
        "#@markdown Model precision option\n",
        "\n",
        "model_precision = \"bfloat16\" # @param [\"bfloat16\", \"float16\"]\n",
        "\n",
        "#@markdown bfloat16 == Half precision/faster speed (if supported, otherwise the model will default to float16)\n",
        "\n",
        "#@markdown float16 == Full precision/fast speed\n",
        "\n",
        "plot_tokens_embeddings = \"None\" # @param [\"None\", \"Start Times\", \"Durations Velocities\", \"Piano Pitches\", \"Drums Pitches\", \"Aux\"]\n",
        "\n",
        "print('=' * 70)\n",
        "print('Loading Giant Music Transformer', select_model_to_load,'Pre-Trained Model...')\n",
        "print('Please wait...')\n",
        "print('=' * 70)\n",
        "\n",
        "full_path_to_models_dir = \"/content/Giant-Music-Transformer/Models\"\n",
        "\n",
        "if select_model_to_load == '786M-44L-Fast-Extra-Large':\n",
        "\n",
        "  model_checkpoint_file_name = 'Giant_Music_Transformer_Extra_Large_Trained_Model_18001_steps_0.2657_loss_0.9272_acc.pth'\n",
        "  model_path = full_path_to_models_dir+'/Extra Large/'+model_checkpoint_file_name\n",
        "\n",
        "  mdim = 1024\n",
        "  num_layers = 44\n",
        "  mrpe = False\n",
        "\n",
        "  if os.path.isfile(model_path):\n",
        "    print('Model already exists...')\n",
        "\n",
        "  else:\n",
        "    hf_hub_download(repo_id='asigalov61/Giant-Music-Transformer',\n",
        "                    filename=model_checkpoint_file_name,\n",
        "                    local_dir='/content/Giant-Music-Transformer/Models/Extra Large',\n",
        "                    )\n",
        "\n",
        "elif select_model_to_load == '585M-32L-Very-Fast-Large':\n",
        "\n",
        "  model_checkpoint_file_name = 'Giant_Music_Transformer_Large_Trained_Model_36074_steps_0.3067_loss_0.927_acc.pth'\n",
        "  model_path = full_path_to_models_dir+'/Large/'+model_checkpoint_file_name\n",
        "\n",
        "  mdim = 1024\n",
        "  num_layers = 32\n",
        "  mrpe = False\n",
        "\n",
        "  if os.path.isfile(model_path):\n",
        "    print('Model already exists...')\n",
        "\n",
        "  else:\n",
        "    hf_hub_download(repo_id='asigalov61/Giant-Music-Transformer',\n",
        "                    filename=model_checkpoint_file_name,\n",
        "                    local_dir='/content/Giant-Music-Transformer/Models/Large',\n",
        "                    )\n",
        "\n",
        "elif select_model_to_load == '482M-8L-Ultra-Fast-Medium':\n",
        "\n",
        "  model_checkpoint_file_name = 'Giant_Music_Transformer_Medium_Trained_Model_42174_steps_0.5211_loss_0.8542_acc.pth'\n",
        "  model_path = full_path_to_models_dir+'/Medium/'+model_checkpoint_file_name\n",
        "\n",
        "  mdim = 2048\n",
        "  num_layers = 8\n",
        "  mrpe = True\n",
        "\n",
        "  if os.path.isfile(model_path):\n",
        "    print('Model already exists...')\n",
        "\n",
        "  else:\n",
        "    hf_hub_download(repo_id='asigalov61/Giant-Music-Transformer',\n",
        "                    filename=model_checkpoint_file_name,\n",
        "                    local_dir='/content/Giant-Music-Transformer/Models/Medium',\n",
        "                    )\n",
        "\n",
        "print('=' * 70)\n",
        "print('Instantiating model...')\n",
        "\n",
        "device_type = 'cuda'\n",
        "\n",
        "if model_precision == 'bfloat16' and torch.cuda.is_bf16_supported():\n",
        "  dtype = 'bfloat16'\n",
        "else:\n",
        "  dtype = 'float16'\n",
        "\n",
        "if model_precision == 'float16':\n",
        "  dtype = 'float16'\n",
        "\n",
        "ptdtype = {'bfloat16': torch.bfloat16, 'float16': torch.float16}[dtype]\n",
        "ctx = torch.amp.autocast(device_type=device_type, dtype=ptdtype)\n",
        "\n",
        "SEQ_LEN = 8192\n",
        "PAD_IDX = 19463\n",
        "\n",
        "model = TransformerWrapper(\n",
        "        num_tokens = PAD_IDX+1,\n",
        "        max_seq_len = SEQ_LEN,\n",
        "        attn_layers = Decoder(dim = mdim,\n",
        "                              depth = num_layers,\n",
        "                              heads = 32,\n",
        "                              rotary_pos_emb = mrpe,\n",
        "                              attn_flash = True\n",
        "                              )\n",
        ")\n",
        "\n",
        "model = AutoregressiveWrapper(model, ignore_index=PAD_IDX, pad_value=PAD_IDX)\n",
        "\n",
        "print('=' * 70)\n",
        "print('Loading model checkpoint...')\n",
        "\n",
        "model.load_state_dict(torch.load(model_path))\n",
        "\n",
        "print('=' * 70)\n",
        "\n",
        "model.cuda()\n",
        "model.eval()\n",
        "\n",
        "print('Done!')\n",
        "print('=' * 70)\n",
        "print('Model will use', dtype, 'precision...')\n",
        "print('=' * 70)\n",
        "\n",
        "# Model stats\n",
        "print('Model summary...')\n",
        "summary(model)\n",
        "\n",
        "# Plot Token Embeddings\n",
        "if plot_tokens_embeddings != 'None':\n",
        "  tok_emb = model.net.token_emb.emb.weight.detach().cpu().tolist()\n",
        "\n",
        "if plot_tokens_embeddings == 'Start Times':\n",
        "  tok_range = [0, 256]\n",
        "\n",
        "elif plot_tokens_embeddings == 'Durations Velocities':\n",
        "  tok_range = [256, 2304]\n",
        "\n",
        "elif plot_tokens_embeddings == 'Piano Pitches':\n",
        "  tok_range = [2304, 2304+128]\n",
        "\n",
        "elif plot_tokens_embeddings == 'Drums Pitches':\n",
        "  tok_range = [18945-128, 18945]\n",
        "\n",
        "elif plot_tokens_embeddings == 'Aux':\n",
        "  tok_range = [18945, 19465]\n",
        "\n",
        "if plot_tokens_embeddings != 'None':\n",
        "\n",
        "  tok_emb1 = []\n",
        "\n",
        "  for t in tok_emb[tok_range[0]:tok_range[1]]:\n",
        "    tok_emb1.append(t)\n",
        "\n",
        "  cos_sim = metrics.pairwise_distances(\n",
        "    tok_emb1, metric='cosine'\n",
        "  )\n",
        "  plt.figure(figsize=(7, 7))\n",
        "  plt.imshow(cos_sim, cmap=\"inferno\", interpolation=\"nearest\")\n",
        "  im_ratio = cos_sim.shape[0] / cos_sim.shape[1]\n",
        "  plt.colorbar(fraction=0.046 * im_ratio, pad=0.04)\n",
        "  plt.xlabel(\"Position\")\n",
        "  plt.ylabel(\"Position\")\n",
        "  plt.tight_layout()\n",
        "  plt.plot()\n",
        "  plt.savefig(\"/content/Giant-Music-Transformer-Tokens-Embeddings-Plot.png\", bbox_inches=\"tight\")"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# (LOAD SEED MIDI)"
      ],
      "metadata": {
        "id": "7xNyANjZsCOi"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "4QXbFLsKqSnt",
        "cellView": "form"
      },
      "outputs": [],
      "source": [
        "#@title Load Seed MIDI\n",
        "\n",
        "#@markdown Press play button to to upload your own seed MIDI or to load one of the provided sample seed MIDIs from the dropdown list below\n",
        "\n",
        "select_seed_MIDI = \"Upload your own custom MIDI\" # @param [\"Upload your own custom MIDI\", \"Giant-Music-Transformer-Piano-Seed-1\", \"Giant-Music-Transformer-Piano-Seed-2\", \"Giant-Music-Transformer-Piano-Seed-3\", \"Giant-Music-Transformer-Piano-Seed-4\", \"Giant-Music-Transformer-Piano-Seed-5\", \"Giant-Music-Transformer-Piano-Seed-6\", \"Giant-Music-Transformer-MI-Seed-1\", \"Giant-Music-Transformer-MI-Seed-2\", \"Giant-Music-Transformer-MI-Seed-3\", \"Giant-Music-Transformer-MI-Seed-4\", \"Giant-Music-Transformer-MI-Seed-5\", \"Giant-Music-Transformer-MI-Seed-6\"]\n",
        "number_of_prime_tokens = 8190 # @param {type:\"slider\", min:90, max:8190, step:3}\n",
        "trim_all_outputs_to_last_chord = False # @param {type:\"boolean\"}\n",
        "render_MIDI_to_audio = False # @param {type:\"boolean\"}\n",
        "\n",
        "print('=' * 70)\n",
        "print('Giant Music Transformer Seed MIDI Loader')\n",
        "print('=' * 70)\n",
        "\n",
        "#=====================================================\n",
        "# Helper function\n",
        "#=====================================================\n",
        "\n",
        "def trim_to_chord(lst, max_trim_len=30, enabled=False):\n",
        "\n",
        "  if enabled:\n",
        "\n",
        "    # Reverse the list\n",
        "    reversed_lst = lst[::-1]\n",
        "    idx = 0\n",
        "    for i, value in enumerate(reversed_lst):\n",
        "        # Check if the value is non-zero\n",
        "        if 0 < value < 256:\n",
        "            # Convert the index to be from the end of the list\n",
        "            idx = len(lst) - i - 1\n",
        "            break\n",
        "\n",
        "    trimmed_list = lst[:idx]\n",
        "\n",
        "    if (len(lst) - len(trimmed_list)) <= max_trim_len:\n",
        "\n",
        "      return trimmed_list\n",
        "\n",
        "    else:\n",
        "      return lst\n",
        "\n",
        "  else:\n",
        "    return lst\n",
        "\n",
        "#=====================================================\n",
        "\n",
        "f = ''\n",
        "\n",
        "if select_seed_MIDI != \"Upload your own custom MIDI\":\n",
        "  print('Loading seed MIDI...')\n",
        "  f = '/content/Giant-Music-Transformer/Seeds/'+select_seed_MIDI+'.mid'\n",
        "\n",
        "else:\n",
        "  print('Upload your own custom MIDI...')\n",
        "  print('=' * 70)\n",
        "  uploaded_MIDI = files.upload()\n",
        "  if list(uploaded_MIDI.keys()):\n",
        "    f = list(uploaded_MIDI.keys())[0]\n",
        "\n",
        "if f != '':\n",
        "\n",
        "  print('=' * 70)\n",
        "  print('File:', f)\n",
        "  print('=' * 70)\n",
        "\n",
        "  #=======================================================\n",
        "  # START PROCESSING\n",
        "\n",
        "  # Convering MIDI to ms score with MIDI.py module\n",
        "  score = TMIDIX.midi2single_track_ms_score(open(f, 'rb').read(), recalculate_channels=False)\n",
        "\n",
        "  # INSTRUMENTS CONVERSION CYCLE\n",
        "  events_matrix = []\n",
        "  itrack = 1\n",
        "  patches = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]\n",
        "\n",
        "  while itrack < len(score):\n",
        "      for event in score[itrack]:\n",
        "          if event[0] == 'note' or event[0] == 'patch_change':\n",
        "              events_matrix.append(event)\n",
        "      itrack += 1\n",
        "\n",
        "  events_matrix.sort(key=lambda x: x[1])\n",
        "\n",
        "  events_matrix1 = []\n",
        "\n",
        "  for event in events_matrix:\n",
        "          if event[0] == 'patch_change':\n",
        "                patches[event[2]] = event[3]\n",
        "\n",
        "          if event[0] == 'note':\n",
        "                event.extend([patches[event[3]]])\n",
        "\n",
        "                if events_matrix1:\n",
        "                    if (event[1] == events_matrix1[-1][1]):\n",
        "                        if ([event[3], event[4]] != events_matrix1[-1][3:5]):\n",
        "                            events_matrix1.append(event)\n",
        "                    else:\n",
        "                        events_matrix1.append(event)\n",
        "\n",
        "                else:\n",
        "                    events_matrix1.append(event)\n",
        "\n",
        "  if len(events_matrix1) > 0:\n",
        "      if min([e[1] for e in events_matrix1]) >= 0 and min([e[2] for e in events_matrix1]) >= 0:\n",
        "\n",
        "          #=======================================================\n",
        "          # PRE-PROCESSING\n",
        "\n",
        "          # checking number of instruments in a composition\n",
        "          instruments_list_without_drums = list(set([y[3] for y in events_matrix1 if y[3] != 9]))\n",
        "          instruments_list = list(set([y[3] for y in events_matrix1]))\n",
        "\n",
        "          if len(events_matrix1) > 0 and len(instruments_list_without_drums) > 0:\n",
        "\n",
        "              #======================================\n",
        "\n",
        "              events_matrix2 = []\n",
        "\n",
        "              # Recalculating timings\n",
        "              for e in events_matrix1:\n",
        "\n",
        "                  # Original timings\n",
        "                  e[1] = int(e[1] / 16)\n",
        "                  e[2] = int(e[2] / 16)\n",
        "\n",
        "              #===================================\n",
        "              # ORIGINAL COMPOSITION\n",
        "              #===================================\n",
        "\n",
        "              # Sorting by patch, pitch, then by start-time\n",
        "\n",
        "              events_matrix1.sort(key=lambda x: x[6])\n",
        "              events_matrix1.sort(key=lambda x: x[4], reverse=True)\n",
        "              events_matrix1.sort(key=lambda x: x[1])\n",
        "\n",
        "              #=======================================================\n",
        "              # FINAL PROCESSING\n",
        "\n",
        "              melody_chords = []\n",
        "              melody_chords2 = []\n",
        "\n",
        "              # Break between compositions / Intro seq\n",
        "\n",
        "              if 9 in instruments_list:\n",
        "                  drums_present = 19331 # Yes\n",
        "              else:\n",
        "                  drums_present = 19330 # No\n",
        "\n",
        "              if events_matrix1[0][3] != 9:\n",
        "                  pat = events_matrix1[0][6]\n",
        "              else:\n",
        "                  pat = 128\n",
        "\n",
        "              melody_chords.extend([19461, drums_present, 19332+pat]) # Intro seq\n",
        "\n",
        "              #=======================================================\n",
        "              # MAIN PROCESSING CYCLE\n",
        "              #=======================================================\n",
        "\n",
        "              abs_time = 0\n",
        "\n",
        "              pbar_time = 0\n",
        "\n",
        "              pe = events_matrix1[0]\n",
        "\n",
        "              chords_counter = 1\n",
        "\n",
        "              comp_chords_len = len(list(set([y[1] for y in events_matrix1])))\n",
        "\n",
        "              for e in events_matrix1:\n",
        "\n",
        "                  #=======================================================\n",
        "                  # Timings...\n",
        "\n",
        "                  # Cliping all values...\n",
        "                  delta_time = max(0, min(255, e[1]-pe[1]))\n",
        "\n",
        "                  # Durations and channels\n",
        "\n",
        "                  dur = max(0, min(255, e[2]))\n",
        "                  cha = max(0, min(15, e[3]))\n",
        "\n",
        "                  # Patches\n",
        "                  if cha == 9: # Drums patch will be == 128\n",
        "                      pat = 128\n",
        "\n",
        "                  else:\n",
        "                      pat = e[6]\n",
        "\n",
        "                  # Pitches\n",
        "\n",
        "                  ptc = max(1, min(127, e[4]))\n",
        "\n",
        "                  # Velocities\n",
        "\n",
        "                  # Calculating octo-velocity\n",
        "                  vel = max(8, min(127, e[5]))\n",
        "                  velocity = round(vel / 15)-1\n",
        "\n",
        "                  #=======================================================\n",
        "                  # Outro seq\n",
        "\n",
        "                  # if ((comp_chords_len - chords_counter) == 50) and (delta_time != 0):\n",
        "                  #    out_t = 18946+delta_time\n",
        "                  #    out_p = 19202+ptc\n",
        "                  #    melody_chords.extend([18945, out_t, out_p]) # outro seq\n",
        "\n",
        "\n",
        "                  # if delta_time != 0:\n",
        "                  #    chords_counter += 1\n",
        "\n",
        "                  #=======================================================\n",
        "                  # FINAL NOTE SEQ\n",
        "\n",
        "                  # Writing final note asynchronously\n",
        "\n",
        "                  dur_vel = (8 * dur) + velocity\n",
        "                  pat_ptc = (129 * pat) + ptc\n",
        "\n",
        "                  melody_chords.extend([delta_time, dur_vel+256, pat_ptc+2304])\n",
        "                  melody_chords2.append([delta_time, dur_vel+256, pat_ptc+2304])\n",
        "\n",
        "                  pe = e\n",
        "\n",
        "                  #=======================================================\n",
        "\n",
        "              # melody_chords.extend([19462, 19462, 19462]) # EOS\n",
        "\n",
        "              #=======================================================\n",
        "\n",
        "              # TOTAL DICTIONARY SIZE 19462+1=19463\n",
        "              #=======================================================\n",
        "\n",
        "  melody_chords = melody_chords[:number_of_prime_tokens]\n",
        "  melody_chords_f = trim_to_chord(melody_chords,\n",
        "                                  enabled=trim_all_outputs_to_last_chord)\n",
        "\n",
        "  #=======================================================\n",
        "\n",
        "  song = melody_chords_f\n",
        "\n",
        "  song_f = []\n",
        "\n",
        "  time = 0\n",
        "  dur = 0\n",
        "  vel = 90\n",
        "  pitch = 0\n",
        "  channel = 0\n",
        "\n",
        "  patches = [-1] * 16\n",
        "\n",
        "  channels = [0] * 16\n",
        "  channels[9] = 1\n",
        "\n",
        "  for ss in song:\n",
        "\n",
        "      if 0 <= ss < 256:\n",
        "\n",
        "          time += ss * 16\n",
        "\n",
        "      if 256 <= ss < 2304:\n",
        "\n",
        "          dur = ((ss-256) // 8) * 16\n",
        "          vel = (((ss-256) % 8)+1) * 15\n",
        "\n",
        "      if 2304 <= ss < 18945:\n",
        "\n",
        "          patch = (ss-2304) // 129\n",
        "\n",
        "          if patch < 128:\n",
        "\n",
        "              if patch not in patches:\n",
        "                if 0 in channels:\n",
        "                    cha = channels.index(0)\n",
        "                    channels[cha] = 1\n",
        "                else:\n",
        "                    cha = 15\n",
        "\n",
        "                patches[cha] = patch\n",
        "                channel = patches.index(patch)\n",
        "              else:\n",
        "                channel = patches.index(patch)\n",
        "\n",
        "          if patch == 128:\n",
        "              channel = 9\n",
        "\n",
        "          pitch = (ss-2304) % 129\n",
        "\n",
        "          song_f.append(['note', time, dur, channel, pitch, vel, patch ])\n",
        "\n",
        "  patches = [0 if x==-1 else x for x in patches]\n",
        "\n",
        "  detailed_stats = TMIDIX.Tegridy_ms_SONG_to_MIDI_Converter(song_f,\n",
        "                                                            output_signature = 'Giant Music Transformer',\n",
        "                                                            output_file_name = '/content/Giant-Music-Transformer-Seed-Composition',\n",
        "                                                            track_name='Project Los Angeles',\n",
        "                                                            list_of_MIDI_patches=patches\n",
        "                                                            )\n",
        "\n",
        "  #=======================================================\n",
        "\n",
        "  print('=' * 70)\n",
        "  print('Composition stats:')\n",
        "  print('Composition has', int(len(melody_chords_f) / 3), 'notes')\n",
        "  print('Composition has', len(melody_chords_f), 'tokens')\n",
        "  print('Composition MIDI patches:', sorted(list(set([((y-2304) // 129) for y in melody_chords if 2304 <= y < 18945]))))\n",
        "  print('=' * 70)\n",
        "\n",
        "  print('Displaying resulting composition...')\n",
        "  print('=' * 70)\n",
        "\n",
        "  fname = '/content/Giant-Music-Transformer-Seed-Composition'\n",
        "\n",
        "  block_lines = [(song_f[-1][1] / 1000)]\n",
        "  block_tokens = [min(len(melody_chords_f), number_of_prime_tokens)]\n",
        "  pblock = []\n",
        "\n",
        "  if render_MIDI_to_audio:\n",
        "    midi_audio = midi_to_colab_audio(fname + '.mid')\n",
        "    display(Audio(midi_audio, rate=16000, normalize=False))\n",
        "\n",
        "  TMIDIX.plot_ms_SONG(song_f, plot_title=fname)\n",
        "\n",
        "else:\n",
        "  print('=' * 70)"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# (COMPOSITION LOOP)\n",
        "\n",
        "## Run the cells below in a loop to generate endless continuation"
      ],
      "metadata": {
        "id": "69AGT0QTSdqA"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#@title Standard Continuation Generator\n",
        "\n",
        "#@markdown Generation settings\n",
        "\n",
        "try_to_generate_outro = False #@param {type:\"boolean\"}\n",
        "try_to_introduce_drums = False # @param {type:\"boolean\"}\n",
        "number_of_tokens_to_generate = 408 # @param {type:\"slider\", min:33, max:1024, step:3}\n",
        "number_of_batches_to_generate = 4 #@param {type:\"slider\", min:1, max:16, step:1}\n",
        "preview_length_in_tokens = 120 # @param {type:\"slider\", min:33, max:240, step:3}\n",
        "number_of_memory_tokens = 7203 # @param {type:\"slider\", min:300, max:8190, step:3}\n",
        "temperature = 0.9 # @param {type:\"slider\", min:0.1, max:1, step:0.05}\n",
        "model_sampling_top_p_value = 0.96 # @param {type:\"slider\", min:0.1, max:1, step:0.01}\n",
        "\n",
        "#@markdown Other settings\n",
        "\n",
        "allow_model_to_stop_generation_if_needed = False #@param {type:\"boolean\"}\n",
        "render_MIDI_to_audio = True # @param {type:\"boolean\"}\n",
        "\n",
        "print('=' * 70)\n",
        "print('Giant Music Transformer Standard Continuation Model Generator')\n",
        "print('=' * 70)\n",
        "\n",
        "if allow_model_to_stop_generation_if_needed:\n",
        "  min_stop_token = 19462\n",
        "else:\n",
        "  min_stop_token = None\n",
        "\n",
        "preview = melody_chords_f[-preview_length_in_tokens:]\n",
        "\n",
        "mel_cho = melody_chords_f[-number_of_memory_tokens:]\n",
        "\n",
        "if try_to_introduce_drums:\n",
        "  last_note = mel_cho[-3:]\n",
        "  drums = [36, 38, 47, 54]\n",
        "  if 0 <= last_note[0] < 256:\n",
        "    for d in random.choices(drums, k = random.randint(1, len(drums))):\n",
        "      mel_cho.extend([0, min(((4*8)+5)+256, last_note[1]), ((128*129)+d)+2304])\n",
        "\n",
        "if try_to_generate_outro:\n",
        "  mel_cho.extend([18945])\n",
        "\n",
        "torch.cuda.empty_cache()\n",
        "\n",
        "inp = [mel_cho] * number_of_batches_to_generate\n",
        "\n",
        "inp = torch.LongTensor(inp).cuda()\n",
        "\n",
        "with ctx:\n",
        "  with torch.inference_mode():\n",
        "    out = model.generate(inp,\n",
        "                          number_of_tokens_to_generate,\n",
        "                          filter_logits_fn=top_p,\n",
        "                          filter_kwargs={'thres': model_sampling_top_p_value},\n",
        "                          temperature=temperature,\n",
        "                          return_prime=False,\n",
        "                          eos_token=min_stop_token,\n",
        "                          verbose=True)\n",
        "\n",
        "out0 = out.tolist()\n",
        "\n",
        "torch.cuda.empty_cache()\n",
        "\n",
        "print('=' * 70)\n",
        "print('Done!')\n",
        "print('=' * 70)\n",
        "\n",
        "#======================================================================\n",
        "print('Rendering results...')\n",
        "\n",
        "for i in range(number_of_batches_to_generate):\n",
        "\n",
        "  print('=' * 70)\n",
        "  print('Batch #', i)\n",
        "  print('=' * 70)\n",
        "\n",
        "  out1 = trim_to_chord(out0[i],\n",
        "                       enabled=trim_all_outputs_to_last_chord)\n",
        "\n",
        "  print('Sample INTs', out1[:12])\n",
        "  print('=' * 70)\n",
        "\n",
        "  if len(out) != 0:\n",
        "\n",
        "      song = preview + out1\n",
        "      song_f = []\n",
        "\n",
        "      time = 0\n",
        "      dur = 0\n",
        "      vel = 90\n",
        "      pitch = 0\n",
        "      channel = 0\n",
        "\n",
        "      patches = [-1] * 16\n",
        "\n",
        "      channels = [0] * 16\n",
        "      channels[9] = 1\n",
        "\n",
        "      for ss in song:\n",
        "\n",
        "          if 0 <= ss < 256:\n",
        "\n",
        "              time += ss * 16\n",
        "\n",
        "          if 256 <= ss < 2304:\n",
        "\n",
        "              dur = ((ss-256) // 8) * 16\n",
        "              vel = (((ss-256) % 8)+1) * 15\n",
        "\n",
        "          if 2304 <= ss < 18945:\n",
        "\n",
        "              patch = (ss-2304) // 129\n",
        "\n",
        "              if patch < 128:\n",
        "\n",
        "                  if patch not in patches:\n",
        "                    if 0 in channels:\n",
        "                        cha = channels.index(0)\n",
        "                        channels[cha] = 1\n",
        "                    else:\n",
        "                        cha = 15\n",
        "\n",
        "                    patches[cha] = patch\n",
        "                    channel = patches.index(patch)\n",
        "                  else:\n",
        "                    channel = patches.index(patch)\n",
        "\n",
        "              if patch == 128:\n",
        "                  channel = 9\n",
        "\n",
        "              pitch = (ss-2304) % 129\n",
        "\n",
        "              song_f.append(['note', time, dur, channel, pitch, vel, patch ])\n",
        "\n",
        "      patches = [0 if x==-1 else x for x in patches]\n",
        "\n",
        "      detailed_stats = TMIDIX.Tegridy_ms_SONG_to_MIDI_Converter(song_f,\n",
        "                                                          output_signature = 'Giant Music Transformer',\n",
        "                                                          output_file_name = '/content/Giant-Music-Transformer-Music-Composition_'+str(i),\n",
        "                                                          track_name='Project Los Angeles',\n",
        "                                                          list_of_MIDI_patches=patches\n",
        "                                                          )\n",
        "      print('=' * 70)\n",
        "      print('Displaying resulting composition...')\n",
        "      print('=' * 70)\n",
        "\n",
        "      fname = '/content/Giant-Music-Transformer-Music-Composition_'+str(i)\n",
        "\n",
        "      if render_MIDI_to_audio:\n",
        "        midi_audio = midi_to_colab_audio(fname + '.mid')\n",
        "        display(Audio(midi_audio, rate=16000, normalize=False))\n",
        "\n",
        "      TMIDIX.plot_ms_SONG(song_f,\n",
        "                          plot_title=fname,\n",
        "                          preview_length_in_notes=int(preview_length_in_tokens / 3)\n",
        "                          )"
      ],
      "metadata": {
        "cellView": "form",
        "id": "UXxG_JXkShcY"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#@title Choose one generated block to add to the composition\n",
        "block_action = \"add_last_generated_block\" #@param [\"add_last_generated_block\", \"remove_last_added_block\"]\n",
        "add_block_with_batch_number = 0 #@param {type:\"slider\", min:0, max:15, step:1}\n",
        "render_MIDI_to_audio = False # @param {type:\"boolean\"}\n",
        "\n",
        "print('=' * 70)\n",
        "\n",
        "if block_action == 'add_last_generated_block':\n",
        "\n",
        "  new_block = out0[min(len(out0)-1, add_block_with_batch_number)]\n",
        "\n",
        "  if new_block != pblock:\n",
        "    melody_chords_f.extend(new_block)\n",
        "    print('Block added!')\n",
        "\n",
        "  else:\n",
        "    print('Nothing to add!!!')\n",
        "\n",
        "else:\n",
        "  if len(block_tokens) > 1:\n",
        "    melody_chords_f = melody_chords_f[:(len(melody_chords_f)-block_tokens[-1])]\n",
        "\n",
        "    block_lines.pop()\n",
        "    block_tokens.pop()\n",
        "    print('Block removed!')\n",
        "\n",
        "    pblock = []\n",
        "\n",
        "  else:\n",
        "    print('Nothing to remove!!!')\n",
        "\n",
        "print('=' * 70)\n",
        "print('Composition now has', (len(melody_chords_f) // 3), 'notes')\n",
        "print('Composition now has', len(melody_chords_f), 'tokens')\n",
        "print('Composition now has MIDI patches:', sorted(list(set([((y-2304) // 129) for y in melody_chords if 2304 <= y < 18945]))))\n",
        "\n",
        "print('=' * 70)\n",
        "print('Sample INTs', out1[:12])\n",
        "print('=' * 70)\n",
        "\n",
        "if len(melody_chords_f) != 0:\n",
        "\n",
        "    song = melody_chords_f\n",
        "    song_f = []\n",
        "\n",
        "    time = 0\n",
        "    dur = 0\n",
        "    vel = 90\n",
        "    pitch = 0\n",
        "    channel = 0\n",
        "\n",
        "    patches = [-1] * 16\n",
        "\n",
        "    channels = [0] * 16\n",
        "    channels[9] = 1\n",
        "\n",
        "    for ss in song:\n",
        "\n",
        "        if 0 <= ss < 256:\n",
        "\n",
        "            time += ss * 16\n",
        "\n",
        "        if 256 <= ss < 2304:\n",
        "\n",
        "            dur = ((ss-256) // 8) * 16\n",
        "            vel = (((ss-256) % 8)+1) * 15\n",
        "\n",
        "        if 2304 <= ss < 18945:\n",
        "\n",
        "            patch = (ss-2304) // 129\n",
        "\n",
        "            if patch < 128:\n",
        "\n",
        "                if patch not in patches:\n",
        "                  if 0 in channels:\n",
        "                      cha = channels.index(0)\n",
        "                      channels[cha] = 1\n",
        "                  else:\n",
        "                      cha = 15\n",
        "\n",
        "                  patches[cha] = patch\n",
        "                  channel = patches.index(patch)\n",
        "                else:\n",
        "                  channel = patches.index(patch)\n",
        "\n",
        "            if patch == 128:\n",
        "                channel = 9\n",
        "\n",
        "            pitch = (ss-2304) % 129\n",
        "\n",
        "            song_f.append(['note', time, dur, channel, pitch, vel, patch ])\n",
        "\n",
        "    patches = [0 if x==-1 else x for x in patches]\n",
        "\n",
        "    detailed_stats = TMIDIX.Tegridy_ms_SONG_to_MIDI_Converter(song_f,\n",
        "                                                              output_signature = 'Giant Music Transformer',\n",
        "                                                              output_file_name = '/content/Giant-Music-Transformer-Music-Composition',\n",
        "                                                              track_name='Project Los Angeles',\n",
        "                                                              list_of_MIDI_patches=patches\n",
        "                                                              )\n",
        "\n",
        "    if block_action == 'add_last_generated_block' and new_block != pblock:\n",
        "      block_tokens.append(len(new_block))\n",
        "      block_lines.append((song_f[-1][1] / 1000))\n",
        "      pblock = new_block\n",
        "\n",
        "    print('=' * 70)\n",
        "    print('Displaying resulting composition...')\n",
        "    print('=' * 70)\n",
        "\n",
        "    fname = '/content/Giant-Music-Transformer-Music-Composition'\n",
        "\n",
        "    if render_MIDI_to_audio:\n",
        "      midi_audio = midi_to_colab_audio(fname + '.mid')\n",
        "      display(Audio(midi_audio, rate=16000, normalize=False))\n",
        "\n",
        "    TMIDIX.plot_ms_SONG(song_f,\n",
        "                        plot_title=fname,\n",
        "                        block_lines_times_list=block_lines[:-1]\n",
        "                        )"
      ],
      "metadata": {
        "cellView": "form",
        "id": "ykBh-aXyShYC"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Congrats! You did it! :)"
      ],
      "metadata": {
        "id": "eoWDEy6CwDr6"
      }
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "private_outputs": true,
      "provenance": [],
      "gpuClass": "premium",
      "gpuType": "T4"
    },
    "kernelspec": {
      "display_name": "Python 3",
      "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": "3.9.13"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}