Created using Colab

Author: TheNeodev

Advanced-RVC.ipynb

@@ -94,18 +94,149 @@
     },
     {
       "cell_type": "code",
+      "source": [
+        "#@title youtube downloader for separation\n",
+        "import yt_dlp\n",
+        "import os\n",
+        "\n",
+        "def download_wav(video_url, output_dir):\n",
+        "  ydl_opts = {\n",
+        "      'format': 'bestaudio/best',\n",
+        "      'outtmpl': os.path.join(output_dir, '%(title)s.%(ext)s'),\n",
+        "      'noplaylist': True,  # Download only a single video, not a playlist\n",
+        "      'extract_audio': True,\n",
+        "      'audio_format': 'wav', # Ensure WAV format\n",
+        "      'postprocessors': [{\n",
+        "          'key': 'FFmpegExtractAudio',\n",
+        "          'preferredcodec': 'wav',\n",
+        "      }]\n",
+        "  }\n",
+        "\n",
+        "  try:\n",
+        "      with yt_dlp.YoutubeDL(ydl_opts) as ydl:\n",
+        "          ydl.download([video_url])\n",
+        "          print(f\"Download completed. Audio saved in {output_dir}\")\n",
+        "  except yt_dlp.utils.DownloadError as e:\n",
+        "      print(f\"An error occurred: {e}\")\n",
+        "\n",
+        "# Example usage\n",
+        "video_url = 'YOUR_YOUTUBE_URL' #@param {type:\"string\"}\n",
+        " # Replace with your YouTube video URL\n",
+        "output_directory = '/content/yt_dlp/' #@param {type:\"string\"}\n",
+        " # customize your output directory\n",
+        "\n",
+        "# Create the output directory if it doesn't exist.\n",
+        "os.makedirs(output_directory, exist_ok=True)\n",
+        "\n",
+        "download_wav(video_url, output_directory)"
+      ],
+      "metadata": {
+        "cellView": "form",
+        "id": "hDYUwhTXJI3_"
+      },
       "execution_count": null,
+      "outputs": []
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "#@title audio separator for infernece\n",
+        "\n",
+        "\n",
+        "\n",
+        "from audio_separator.separator import Separator\n",
+        "\n",
+        "aud_input = \"Input_aud\"  #@param {type:\"string\"}\n",
+        "\n",
+        "output_dir = \"/content/separated\"  #@param {type:\"string\"}\n",
+        "\n",
+        "# Initialize the Separator class (with optional configuration properties, below)\n",
+        "separator = Separator(output_dir=output_dir, output_format=\"wav\")\n",
+        "\n",
+        "separator.load_model()\n",
+        "\n",
+        "output_names = {\n",
+        "    \"Vocals\": \"vocals_output\",\n",
+        "    \"Instrumental\": \"instrumental_output\",\n",
+        "}\n",
+        "output_files = separator.separate(aud_input, output_names)\n",
+        "print(f\"Separation complete!\")"
+      ],
       "metadata": {
         "cellView": "form",
-        "id": "zG9oR-iDrEPX"
+        "id": "1pwvdDeHJmU6"
       },
-      "outputs": [],
+      "execution_count": null,
+      "outputs": []
+    },
+    {
+      "cell_type": "code",
       "source": [
-        "#@title Run WebUI\n",
         "\n",
         "\n",
-        "!python infer.py --colab"
-      ]
+        "\n",
+        "\n",
+        "#@title ## Inference\n",
+        "import os\n",
+        "main_dir = \"/content/Harmonify\"\n",
+        "os.chdir(main_dir)\n",
+        "from lib.infer import infer_audio\n",
+        "from google.colab import files\n",
+        "from pydub import AudioSegment\n",
+        "import shutil\n",
+        "\n",
+        "MODEL_NAME = \"\" #@param {type:\"string\"}\n",
+        "SOUND_PATH = \"\" #@param {type:\"string\"}\n",
+        "F0_CHANGE = 0 #@param {type:\"integer\"}\n",
+        "F0_METHOD = \"fcpe\" #@param [\"crepe\", \"harvest\", \"mangio-crepe\", \"rmvpe\", \"rmvpe+\", \"fcpe\", \"fcpe_legacy\", \"hybrid[mangio-crepe+rmvpe]\", \"hybrid[mangio-crepe+fcpe]\", \"hybrid[rmvpe+fcpe]\", \"hybrid[mangio-crepe+rmvpe+fcpe]\"] {allow-input:true}\n",
+        "CREPE_HOP_LENGTH = 120 #@param {type:\"integer\"}\n",
+        "INDEX_RATE = 0.75 #@param {type:\"number\"}\n",
+        "FILTER_RADIUS = 3 #@param {type:\"integer\"}\n",
+        "RMS_MIX_RATE = 0.25 #@param {type:\"number\"}\n",
+        "PROTECT = 0.33 #@param {type:\"number\"}\n",
+        "SPLIT_INFER = False #@param {type:\"boolean\"}\n",
+        "MIN_SILENCE = 500 #@param {type:\"number\"}\n",
+        "SILENCE_THRESHOLD = -50 #@param {type:\"number\"}\n",
+        "SEEK_STEP = 1 #@param {type:\"slider\", min:1, max:10, step:1}\n",
+        "KEEP_SILENCE = 200 #@param {type:\"number\"}\n",
+        "QUEFRENCY = 0 #@param {type:\"number\"}\n",
+        "TIMBRE = 1 #@param {type:\"number\"}\n",
+        "F0_AUTOTUNE = False #@param {type:\"boolean\"}\n",
+        "OUTPUT_FORMAT = \"wav\" #@param [\"wav\", \"flac\", \"mp3\"]\n",
+        "\n",
+        "\n",
+        "\n",
+        "inferred_audio = infer_audio(\n",
+        "    MODEL_NAME,\n",
+        "    SOUND_PATH,\n",
+        "    F0_CHANGE,\n",
+        "    F0_METHOD,\n",
+        "    CREPE_HOP_LENGTH,\n",
+        "    INDEX_RATE,\n",
+        "    FILTER_RADIUS,\n",
+        "    RMS_MIX_RATE,\n",
+        "    PROTECT,\n",
+        "    SPLIT_INFER,\n",
+        "    MIN_SILENCE,\n",
+        "    SILENCE_THRESHOLD,\n",
+        "    SEEK_STEP,\n",
+        "    KEEP_SILENCE,\n",
+        "    QUEFRENCY,\n",
+        "    TIMBRE,\n",
+        "    F0_AUTOTUNE,\n",
+        "    OUTPUT_FORMAT\n",
+        ")\n",
+        "os.chdir(main_dir)\n",
+        "\n",
+        "print(f\"Showing {inferred_audio}.\")\n",
+        "AudioSegment.from_file(inferred_audio)"
+      ],
+      "metadata": {
+        "cellView": "form",
+        "id": "PKIrHe_nIou9"
+      },
+      "execution_count": null,
+      "outputs": []
     }
   ],
   "metadata": {