Text_to_Music_Transformer.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "# Text-to-Music Transformer (ver. 1.0)\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/Text-to-Music-Transformer\n",
        "!pip install huggingface_hub\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 Text-to-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 Text-to-Music Transformer modules...')\n",
        "import torch\n",
        "\n",
        "%cd /content/Text-to-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 Text-to-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('Done!')\n",
        "print('Enjoy! :)')\n",
        "print('=' * 70)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eI3aQtHzqSnp"
      },
      "source": [
        "# (LOAD MODEL)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#@title Load Quad Music Transformer Pre-Trained Model\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 = True # @param {type:\"boolean\"}\n",
        "\n",
        "print('=' * 70)\n",
        "print('Loading Text-to-Music Transformer Pre-Trained Model...')\n",
        "print('Please wait...')\n",
        "print('=' * 70)\n",
        "\n",
        "full_path_to_models_dir = \"/content/Text-to-Music-Transformer/Model\"\n",
        "\n",
        "model_checkpoint_file_name = 'Text_to_Music_Transformer_Medium_Trained_Model_33934_steps_0.6093_loss_0.813_acc.pth'\n",
        "model_path = full_path_to_models_dir+'/'+model_checkpoint_file_name\n",
        "\n",
        "if os.path.isfile(model_path):\n",
        "  print('Model already exists...')\n",
        "\n",
        "else:\n",
        "  hf_hub_download(repo_id='asigalov61/Text-to-Music-Transformer',\n",
        "                  filename=model_checkpoint_file_name,\n",
        "                  local_dir='/content/Text-to-Music-Transformer/Model/',\n",
        "                  local_dir_use_symlinks=False)\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 = 4096\n",
        "\n",
        "# instantiate the model\n",
        "\n",
        "model = TransformerWrapper(\n",
        "    num_tokens = 2572,\n",
        "    max_seq_len = SEQ_LEN,\n",
        "    attn_layers = Decoder(dim = 2048, depth = 8, heads = 16, attn_flash = True)\n",
        ")\n",
        "\n",
        "model = AutoregressiveWrapper(model, ignore_index=2571)\n",
        "\n",
        "model.cuda()\n",
        "print('=' * 70)\n",
        "\n",
        "print('Loading model checkpoint...')\n",
        "\n",
        "model.load_state_dict(torch.load(model_path))\n",
        "print('=' * 70)\n",
        "\n",
        "model.eval()\n",
        "\n",
        "print('Done!')\n",
        "print('=' * 70)\n",
        "\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:\n",
        "  tok_emb = model.net.token_emb.emb.weight.detach().cpu().tolist()\n",
        "\n",
        "  cos_sim = metrics.pairwise_distances(\n",
        "    tok_emb, 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/Text-to-Music-Transformer-Tokens-Embeddings-Plot.png\", bbox_inches=\"tight\")"
      ],
      "metadata": {
        "id": "V4s_G8yUL0cH",
        "cellView": "form"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# (SETUP MODEL CHANNELS MIDI PATCHES)"
      ],
      "metadata": {
        "id": "o8_VUwLPlVa5"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# @title Setup and load model channels MIDI patches\n",
        "\n",
        "model_channel_0_piano_family = \"Acoustic Grand\" # @param [\"Acoustic Grand\", \"Bright Acoustic\", \"Electric Grand\", \"Honky-Tonk\", \"Electric Piano 1\", \"Electric Piano 2\", \"Harpsichord\", \"Clav\"]\n",
        "model_channel_1_chromatic_percussion_family = \"Music Box\" # @param [\"Celesta\", \"Glockenspiel\", \"Music Box\", \"Vibraphone\", \"Marimba\", \"Xylophone\", \"Tubular Bells\", \"Dulcimer\"]\n",
        "model_channel_2_organ_family = \"Church Organ\" # @param [\"Drawbar Organ\", \"Percussive Organ\", \"Rock Organ\", \"Church Organ\", \"Reed Organ\", \"Accordion\", \"Harmonica\", \"Tango Accordion\"]\n",
        "model_channel_3_guitar_family = \"Acoustic Guitar(nylon)\" # @param [\"Acoustic Guitar(nylon)\", \"Acoustic Guitar(steel)\", \"Electric Guitar(jazz)\", \"Electric Guitar(clean)\", \"Electric Guitar(muted)\", \"Overdriven Guitar\", \"Distortion Guitar\", \"Guitar Harmonics\"]\n",
        "model_channel_4_bass_family = \"Fretless Bass\" # @param [\"Acoustic Bass\", \"Electric Bass(finger)\", \"Electric Bass(pick)\", \"Fretless Bass\", \"Slap Bass 1\", \"Slap Bass 2\", \"Synth Bass 1\", \"Synth Bass 2\"]\n",
        "model_channel_5_strings_family = \"Violin\" # @param [\"Violin\", \"Viola\", \"Cello\", \"Contrabass\", \"Tremolo Strings\", \"Pizzicato Strings\", \"Orchestral Harp\", \"Timpani\"]\n",
        "model_channel_6_ensemble_family = \"Choir Aahs\" # @param [\"String Ensemble 1\", \"String Ensemble 2\", \"SynthStrings 1\", \"SynthStrings 2\", \"Choir Aahs\", \"Voice Oohs\", \"Synth Voice\", \"Orchestra Hit\"]\n",
        "model_channel_7_brass_family = \"Trumpet\" # @param [\"Trumpet\", \"Trombone\", \"Tuba\", \"Muted Trumpet\", \"French Horn\", \"Brass Section\", \"SynthBrass 1\", \"SynthBrass 2\"]\n",
        "model_channel_8_reed_family = \"Alto Sax\" # @param [\"Soprano Sax\", \"Alto Sax\", \"Tenor Sax\", \"Baritone Sax\", \"Oboe\", \"English Horn\", \"Bassoon\", \"Clarinet\"]\n",
        "model_channel_9_pipe_family = \"Flute\" # @param [\"Piccolo\", \"Flute\", \"Recorder\", \"Pan Flute\", \"Blown Bottle\", \"Skakuhachi\", \"Whistle\", \"Ocarina\"]\n",
        "model_channel_10_synth_lead_family = \"Lead 8 (bass+lead)\" # @param [\"Lead 1 (square)\", \"Lead 2 (sawtooth)\", \"Lead 3 (calliope)\", \"Lead 4 (chiff)\", \"Lead 5 (charang)\", \"Lead 6 (voice)\", \"Lead 7 (fifths)\", \"Lead 8 (bass+lead)\"]\n",
        "model_channel_11_synth_pad_family = \"Pad 2 (warm)\" # @param [\"Pad 1 (new age)\", \"Pad 2 (warm)\", \"Pad 3 (polysynth)\", \"Pad 4 (choir)\", \"Pad 5 (bowed)\", \"Pad 6 (metallic)\", \"Pad 7 (halo)\", \"Pad 8 (sweep)\"]\n",
        "model_channel_12_synth_effects_family = \"FX 3 (crystal)\" # @param [\"FX 1 (rain)\", \"FX 2 (soundtrack)\", \"FX 3 (crystal)\", \"FX 4 (atmosphere)\", \"FX 5 (brightness)\", \"FX 6 (goblins)\", \"FX 7 (echoes)\", \"FX 8 (sci-fi)\"]\n",
        "model_channel_13_ethnic_family = \"Banjo\" # @param [\"Sitar\", \"Banjo\", \"Shamisen\", \"Koto\", \"Kalimba\", \"Bagpipe\", \"Fiddle\", \"Shanai\"]\n",
        "model_channel_14_percussive_family = \"Melodic Tom\" # @param [\"Tinkle Bell\", \"Agogo\", \"Steel Drums\", \"Woodblock\", \"Taiko Drum\", \"Melodic Tom\", \"Synth Drum\", \"Reverse Cymbal\"]\n",
        "model_channel_15_sound_effects_family = \"Bird Tweet\" # @param [\"Guitar Fret Noise\", \"Breath Noise\", \"Seashore\", \"Bird Tweet\", \"Telephone Ring\", \"Helicopter\", \"Applause\", \"Gunshot\"]\n",
        "model_channel_16_drums_family = \"Drums\" # @param [\"Drums\"]\n",
        "\n",
        "def txt2tokens(txt):\n",
        "    return [ord(char)+2440 if 0 < ord(char) < 128 else 0+2440 for char in txt.lower()]\n",
        "\n",
        "def tokens2txt(tokens):\n",
        "    return [chr(tok-2440) for tok in tokens if 0+2440 < tok < 128+2440 ]\n",
        "\n",
        "\n",
        "print('=' * 70)\n",
        "print('Setting up patches...')\n",
        "print('=' * 70)\n",
        "\n",
        "instruments = [v[1] for v in TMIDIX.Number2patch.items()]\n",
        "\n",
        "patches = [instruments.index(model_channel_0_piano_family),\n",
        "                       instruments.index(model_channel_1_chromatic_percussion_family),\n",
        "                       instruments.index(model_channel_2_organ_family),\n",
        "                       instruments.index(model_channel_3_guitar_family),\n",
        "                       instruments.index(model_channel_4_bass_family),\n",
        "                       instruments.index(model_channel_5_strings_family),\n",
        "                       instruments.index(model_channel_6_ensemble_family),\n",
        "                       instruments.index(model_channel_7_brass_family),\n",
        "                       instruments.index(model_channel_8_reed_family),\n",
        "                       9, # Drums patch\n",
        "                       instruments.index(model_channel_9_pipe_family),\n",
        "                       instruments.index(model_channel_10_synth_lead_family),\n",
        "                       instruments.index(model_channel_11_synth_pad_family),\n",
        "                       instruments.index(model_channel_12_synth_effects_family),\n",
        "                       instruments.index(model_channel_13_ethnic_family),\n",
        "                       instruments.index(model_channel_15_sound_effects_family)\n",
        "                       ]\n",
        "\n",
        "print('Done!')\n",
        "print('=' * 70)"
      ],
      "metadata": {
        "cellView": "form",
        "id": "fHOhVyxCDtfB"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# (TEXT-TO-MUSIC GENERATION)"
      ],
      "metadata": {
        "id": "BxepTeHVmmKO"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#@title Standard Text-to-Music Generator\n",
        "\n",
        "#@markdown Prompt settings\n",
        "\n",
        "song_title_prompt = \"Nothing Else Matters\" # @param {type:\"string\"}\n",
        "open_ended_prompt = False # @param {type:\"boolean\"}\n",
        "\n",
        "#@markdown Generation settings\n",
        "\n",
        "number_of_tokens_to_generate = 525 # @param {type:\"slider\", min:30, max:8190, step:3}\n",
        "number_of_batches_to_generate = 4 #@param {type:\"slider\", min:1, max:16, step:1}\n",
        "temperature = 0.9 # @param {type:\"slider\", min:0.1, max:1, step:0.05}\n",
        "\n",
        "#@markdown Other settings\n",
        "\n",
        "render_MIDI_to_audio = True # @param {type:\"boolean\"}\n",
        "\n",
        "print('=' * 70)\n",
        "print('Text-to-Music Model Generator')\n",
        "print('=' * 70)\n",
        "\n",
        "if song_title_prompt == '':\n",
        "  outy = [2569]\n",
        "\n",
        "else:\n",
        "  if open_ended_prompt:\n",
        "    outy = [2569] + txt2tokens(song_title_prompt)\n",
        "\n",
        "  else:\n",
        "    outy = [2569] + txt2tokens(song_title_prompt) + [2570]\n",
        "\n",
        "print('Selected prompt sequence:')\n",
        "print(outy[:12])\n",
        "print('=' * 70)\n",
        "\n",
        "torch.cuda.empty_cache()\n",
        "\n",
        "inp = [outy] * number_of_batches_to_generate\n",
        "\n",
        "inp = torch.LongTensor(inp).cuda()\n",
        "\n",
        "with ctx:\n",
        "  out = model.generate(inp,\n",
        "                        number_of_tokens_to_generate,\n",
        "                        temperature=temperature,\n",
        "                        return_prime=True,\n",
        "                        verbose=True)\n",
        "\n",
        "out0 = out.tolist()\n",
        "\n",
        "print('=' * 70)\n",
        "print('Done!')\n",
        "print('=' * 70)\n",
        "\n",
        "torch.cuda.empty_cache()\n",
        "\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 = out0[i]\n",
        "\n",
        "  print('Sample INTs', out1[:12])\n",
        "  print('=' * 70)\n",
        "\n",
        "  print('Generated song title:', ''.join(tokens2txt(out1)).title())\n",
        "  print('=' * 70)\n",
        "\n",
        "  if len(out1) != 0:\n",
        "\n",
        "      song = out1\n",
        "      song_f = []\n",
        "\n",
        "      time = 0\n",
        "      dur = 0\n",
        "      vel = 90\n",
        "      pitch = 0\n",
        "      channel = 0\n",
        "\n",
        "      for ss in song:\n",
        "\n",
        "          if 0 <= ss < 128:\n",
        "\n",
        "              time += ss * 32\n",
        "\n",
        "          if 128 <= ss < 256:\n",
        "\n",
        "              dur = (ss-128) * 32\n",
        "\n",
        "          if 256 <= ss < 2432:\n",
        "\n",
        "              chan = (ss-256) // 128\n",
        "\n",
        "              if chan < 9:\n",
        "                  channel = chan\n",
        "              elif 9 < chan < 15:\n",
        "                  channel = chan+1\n",
        "              elif chan == 15:\n",
        "                  channel = 15\n",
        "              elif chan == 16:\n",
        "                  channel = 9\n",
        "\n",
        "              pitch = (ss-256) % 128\n",
        "\n",
        "          if 2432 <= ss < 2440:\n",
        "\n",
        "              vel = (((ss-2432)+1) * 15)-1\n",
        "\n",
        "              song_f.append(['note', time, dur, channel, pitch, vel, chan*8 ])\n",
        "\n",
        "      stats = TMIDIX.Tegridy_ms_SONG_to_MIDI_Converter(song_f,\n",
        "                                                        output_signature = 'Text-to-Music Transformer',\n",
        "                                                        output_file_name = '/content/Text-to-Music-Transformer-Composition_'+str(i),\n",
        "                                                        track_name='Project Los Angeles',\n",
        "                                                        list_of_MIDI_patches=patches\n",
        "                                                        )\n",
        "\n",
        "\n",
        "      print('=' * 70)\n",
        "      print('Displaying resulting composition...')\n",
        "      print('=' * 70)\n",
        "\n",
        "      fname = '/content/Text-to-Music-Transformer-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, plot_title=fname)"
      ],
      "metadata": {
        "cellView": "form",
        "id": "fg45KZyqWKxh"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Congrats! You did it! :)"
      ],
      "metadata": {
        "id": "eoWDEy6CwDr6"
      }
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "private_outputs": true,
      "provenance": [],
      "gpuType": "T4",
      "gpuClass": "premium"
    },
    "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
}