{ "cells": [ { "cell_type": "code", "execution_count": 1, "id": "b304dfff", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Populating the interactive namespace from numpy and matplotlib\n" ] } ], "source": [ "%pylab inline" ] }, { "cell_type": "code", "execution_count": 2, "id": "f955e852", "metadata": {}, "outputs": [], "source": [ "import os\n", "os.environ[\"CUDA_VISIBLE_DEVICES\"] = \"0\"" ] }, { "cell_type": "code", "execution_count": 13, "id": "e9716000", "metadata": {}, "outputs": [], "source": [ "import time\n", "import os\n", "import tqdm\n", "import shutil\n", "import random\n", "import funcy\n", "import json\n", "import numpy as np\n", "import multiprocessing\n", "import pandas as pd\n", "from suno_utils.audio import Audio\n", "from suno_utils.audio.conversion import play_audio, get_audio_properties\n", "from suno_utils.utils.slicer import get_vad_gaps, get_intervals_from_gaps\n", "\n", "\n", "DATE_DIR = \"2022_08_07\"\n", "\n", "BASE_DIR = \"/mnt/data-ssd-1/data/private/customer/sanas/2022-08-04-fili-callcenter/\"\n", "TO_REV_DIR = os.path.join(BASE_DIR, \"to_rev\", DATE_DIR)\n", "FROM_REV_DIR = os.path.join(BASE_DIR, \"from_rev\", DATE_DIR)\n", "TO_REV_AUDIO_DIR = os.path.join(TO_REV_DIR, \"audio\")\n", "REV_MANIFEST_FILEPATH = os.path.join(TO_REV_DIR, \"rev_manifest.jsonl\")\n", "FAILED_MANIFEST_FILEPATH = os.path.join(TO_REV_DIR, \"failed_manifest.jsonl\")" ] }, { "cell_type": "code", "execution_count": 17, "id": "e22f6da7", "metadata": {}, "outputs": [], "source": [ "os.makedirs(FROM_REV_DIR, exist_ok=True)\n", "os.makedirs(TO_REV_DIR, exist_ok=True)\n", "os.makedirs(TO_REV_AUDIO_DIR, exist_ok=True)" ] }, { "cell_type": "code", "execution_count": 14, "id": "ac737816", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "1 failed\n", "55 files - 4.4 hours\n", "705 files - 113.2 hours\n" ] } ], "source": [ "# ERC-PH-ASR-1 (56 audio files) and Continuum-PH-ASR-1 (705 audio files).\n", "d = os.path.join(BASE_DIR, \"ERC-PH-ASR-1\")\n", "filepaths_1 = [os.path.join(d, fn) for fn in os.listdir(d)]\n", "d = os.path.join(BASE_DIR, \"Continuum-PH-ASR-1\")\n", "filepaths_2 = [os.path.join(d, fn) for fn in os.listdir(d)]\n", "\n", "l1 = []\n", "e1 = []\n", "for fp in filepaths_1:\n", " try:\n", " l1.append(get_audio_properties(fp)[\"duration_s\"])\n", " except:\n", " e1.append(fp)\n", "l2 = []\n", "e2 = []\n", "for fp in filepaths_2:\n", " try:\n", " l2.append(get_audio_properties(fp)[\"duration_s\"])\n", " except:\n", " e2.append(fp)\n", " \n", "print(len(e1 + e2), \"failed\")\n", "print(len(l1), \"files -\", round(sum(l1) / 60 / 60, 1), \"hours\")\n", "print(len(l2), \"files -\", round(sum(l2) / 60 / 60, 1), \"hours\")" ] }, { "cell_type": "code", "execution_count": 18, "id": "7e83d447", "metadata": {}, "outputs": [], "source": [ "import portion\n", "from suno_utils.utils.numbers import safe_round\n", "from suno_utils.tasks.music import remove_music\n", " \n", "def prep_annotation_audio(audio, clean_audio=True):\n", " min_silence_gap = 0.05\n", " min_segment_dur_s = 3.0\n", " max_segment_dur_s = 10.0\n", " # TODO: remove background noise?\n", " if clean_audio:\n", " cleaned_audio = remove_music(audio, use_gpu=True)\n", " else:\n", " cleaned_audio = audio\n", " vad_gaps = get_vad_gaps(cleaned_audio, min_gap_s=min_silence_gap, vad_level=3)\n", " silence_frac = np.sum([g.upper - g.lower for g in vad_gaps]) / audio.duration_s\n", "# print(\"{}% silence\".format(round(silence_frac * 100, 1)))\n", "\n", " # anything longer than a second should get excluded with pad\n", " max_silence_s = 1\n", " max_silence_pad = max_silence_s / 2\n", " excluded_intervals = portion.openclosed(0, 0)\n", " for vad_gap in vad_gaps:\n", " excluded_lower_s = safe_round(vad_gap.lower + max_silence_pad)\n", " excluded_upper_s = safe_round(vad_gap.upper - max_silence_pad)\n", " if excluded_upper_s - excluded_lower_s > 0:\n", " excluded_intervals |= portion.openclosed(excluded_lower_s, excluded_upper_s)\n", " \n", " speech_segments = get_intervals_from_gaps(\n", " vad_gaps,\n", " excluded_intervals=excluded_intervals,\n", " min_dur_s=min_segment_dur_s,\n", " max_dur_s=max_segment_dur_s,\n", " total_duration_s=audio.duration_s\n", " )\n", " tot_segment_duration_s = np.sum([end_s - start_s for start_s, end_s in speech_segments])\n", " cov_frac = tot_segment_duration_s / audio.duration_s\n", "# print(\"{}% segment coverage\".format(round(cov_frac * 100, 1)))\n", "# tot_segment_duration_mins = int(tot_segment_duration_s // 60)\n", "# tot_segment_duration_secs = int(np.ceil(tot_segment_duration_s % 60))\n", "# print(\"{}m {}s segment audio\".format(tot_segment_duration_mins, tot_segment_duration_secs))\n", "\n", " audios_rev = []\n", " segments_meta = []\n", " for n, (start_s, end_s) in enumerate(speech_segments):\n", "# indicator_str = \"a\" + str(n)\n", " indicator_str = str(n).zfill(2)\n", " audios_indicator = []\n", " for c in indicator_str:\n", " audios_indicator.append(Audio.from_file(f\"audio_indicators/{c}_fast.wav\"))\n", " audios_rev.extend(audios_indicator)\n", " segments_meta.append({\n", " \"segment_number\": n,\n", " \"type\": \"indicator\",\n", " \"duration_s\": safe_round(np.sum([a.duration_s for a in audios_indicator])),\n", " \"indicator_str\": indicator_str,\n", " })\n", " audios_rev.append(audio.get_slice(start_s, end_s))\n", " segments_meta.append({\n", " \"segment_number\": n,\n", " \"type\": \"speech\",\n", " \"duration_s\": safe_round(end_s - start_s),\n", " \"orginial_audio_offset_s\": start_s,\n", " })\n", " \n", " if len(audios_rev) == 0:\n", " return None, {\n", " \"tot_silence_frac\": round(silence_frac, 3),\n", " \"segment_cov_frac\": round(cov_frac, 3),\n", " }\n", "\n", " if (\n", " np.min([e[\"duration_s\"] for e in segments_meta if e[\"type\"] == \"speech\"]) < min_segment_dur_s or\n", " np.max([e[\"duration_s\"] for e in segments_meta if e[\"type\"] == \"speech\"]) > max_segment_dur_s\n", " ):\n", " raise ValueError(\"incorrect segment length detected\")\n", " \n", " audio_rev = Audio.concatenate(audios_rev)\n", " stitch_overhead_frac = (audio_rev.duration_s - tot_segment_duration_s) / audio_rev.duration_s\n", "# print(\"{}% stitching overhead\".format(round(stitch_overhead_frac * 100, 1)))\n", " submission_duration_s = np.ceil(audio_rev.duration_s / 60) * 60\n", " tot_overhead_frac = (submission_duration_s - tot_segment_duration_s) / submission_duration_s\n", "# print(\"{}% total overhead\".format(round(tot_overhead_frac * 100, 1)))\n", "\n", " submission_meta = {\n", " \"tot_silence_frac\": round(silence_frac, 3),\n", " \"segment_cov_frac\": round(cov_frac, 3),\n", " \"stitch_overhead_frac\": round(stitch_overhead_frac, 3),\n", " \"tot_overhead_frac\": round(tot_overhead_frac, 3),\n", " \"segments_meta\": segments_meta,\n", " }\n", " # TODO: alert on mostly silence\n", " # TODO: alert on total time too short \n", " # TODO: alert on coverage \n", " # TODO: alert stitching overhead\n", " # TODO: alert rev rounding overhead\n", " return audio_rev, submission_meta" ] }, { "cell_type": "code", "execution_count": 9, "id": "b17176dc", "metadata": {}, "outputs": [], "source": [ "# filepath = filepaths_2[5]\n", "\n", "# audio = Audio.from_file(filepath, sample_rate=16_000, byte_width=2)\n", "# audio_rev, submission_meta = prep_annotation_audio(audio)\n", "# audio_rev.play()" ] }, { "cell_type": "code", "execution_count": 65, "id": "9b27da8b", "metadata": {}, "outputs": [], "source": [ "# !rm -rf /mnt/data-ssd-1/data/private/customer/sanas/2022-08-04-fili-callcenter/to_rev/*" ] }, { "cell_type": "code", "execution_count": 19, "id": "96d0f32a", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ " 4%|██████▎ | 32/761 [07:06<2:21:48, 11.67s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 5%|███████▎ | 37/761 [08:11<2:26:26, 12.14s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 5%|███████▌ | 38/761 [08:18<2:07:49, 10.61s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 5%|███████▉ | 40/761 [08:39<2:12:02, 10.99s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 6%|████████▋ | 44/761 [09:26<2:25:05, 12.14s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 6%|█████████▏ | 46/761 [09:45<2:14:51, 11.32s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 7%|██████████▋ | 54/761 [11:23<2:24:33, 12.27s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 7%|███████████ | 56/761 [11:45<2:20:08, 11.93s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 7%|███████████▎ | 57/761 [12:30<4:10:54, 21.38s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 8%|███████████▌ | 58/761 [12:43<3:42:57, 19.03s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 8%|███████████▋ | 59/761 [13:00<3:36:35, 18.51s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 8%|███████████▉ | 60/761 [13:16<3:27:19, 17.75s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 8%|████████████▎ | 62/761 [14:13<4:15:19, 21.92s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 9%|█████████████▎ | 67/761 [16:21<4:20:55, 22.56s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 9%|█████████████▍ | 68/761 [16:45<4:25:37, 23.00s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 9%|██████████████ | 71/761 [18:01<4:20:19, 22.64s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 10%|███████████████ | 76/761 [19:53<3:50:09, 20.16s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 10%|███████████████▋ | 79/761 [20:56<3:46:36, 19.94s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 11%|███████████████▊ | 80/761 [21:16<3:43:56, 19.73s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 11%|████████████████▎ | 82/761 [21:47<3:15:43, 17.30s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 12%|██████████████████▍ | 93/761 [25:13<4:56:34, 26.64s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 12%|██████████████████▊ | 95/761 [25:42<3:43:17, 20.12s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 13%|███████████████████▏ | 97/761 [26:34<4:04:51, 22.13s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 13%|███████████████████▉ | 101/761 [27:39<3:01:03, 16.46s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 14%|████████████████████▍ | 104/761 [28:26<2:44:21, 15.01s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 14%|█████████████████████ | 107/761 [29:10<2:38:32, 14.55s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 14%|█████████████████████▎ | 108/761 [29:27<2:46:25, 15.29s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 14%|█████████████████████▋ | 110/761 [30:01<2:49:22, 15.61s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 15%|██████████████████████▊ | 116/761 [32:22<4:59:55, 27.90s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 16%|███████████████████████▎ | 118/761 [32:45<3:31:35, 19.74s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 16%|███████████████████████▋ | 120/761 [33:15<3:10:29, 17.83s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 16%|████████████████████████ | 122/761 [33:46<2:51:33, 16.11s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 16%|████████████████████████▏ | 123/761 [34:30<4:19:12, 24.38s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 17%|████████████████████████▊ | 126/761 [35:26<3:30:41, 19.91s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 17%|█████████████████████████▍ | 129/761 [36:26<3:34:27, 20.36s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 17%|██████████████████████████ | 132/761 [37:01<2:38:03, 15.08s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 18%|██████████████████████████▊ | 136/761 [38:27<4:06:30, 23.66s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 18%|███████████████████████████▌ | 140/761 [39:52<3:29:25, 20.23s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 19%|████████████████████████████▍ | 144/761 [40:52<2:44:47, 16.02s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 19%|████████████████████████████▊ | 146/761 [41:51<3:39:24, 21.41s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 20%|█████████████████████████████▌ | 150/761 [43:11<3:14:58, 19.15s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 20%|█████████████████████████████▊ | 151/761 [43:18<2:39:39, 15.70s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 20%|██████████████████████████████▎ | 154/761 [44:58<4:15:43, 25.28s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 20%|██████████████████████████████▌ | 155/761 [45:41<5:08:47, 30.57s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 21%|██████████████████████████████▉ | 157/761 [46:46<5:06:57, 30.49s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 21%|███████████████████████████████▏ | 158/761 [47:23<5:25:34, 32.40s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 21%|███████████████████████████████▉ | 162/761 [48:11<2:45:22, 16.56s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 22%|█████████████████████████████████ | 168/761 [50:48<3:55:44, 23.85s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 22%|█████████████████████████████████▌ | 170/761 [51:08<2:49:00, 17.16s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 22%|█████████████████████████████████▋ | 171/761 [51:27<2:53:40, 17.66s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 23%|██████████████████████████████████▉ | 177/761 [53:39<2:52:47, 17.75s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 24%|███████████████████████████████████▍ | 180/761 [55:01<3:44:31, 23.19s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 24%|████████████████████████████████████ | 183/761 [55:58<3:32:45, 22.09s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 24%|████████████████████████████████████▍ | 185/761 [57:00<4:07:21, 25.77s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 25%|█████████████████████████████████████▊ | 192/761 [58:23<2:04:05, 13.09s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 25%|██████████████████████████████████████ | 193/761 [58:47<2:33:45, 16.24s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 26%|██████████████████████████████████████▊ | 197/761 [59:47<2:27:02, 15.64s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 26%|██████████████████████████████████████▌ | 198/761 [1:00:03<2:27:30, 15.72s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 26%|██████████████████████████████████████▋ | 199/761 [1:00:19<2:29:13, 15.93s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 27%|███████████████████████████████████████▍ | 203/761 [1:01:15<2:20:42, 15.13s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 27%|████████████████████████████████████████▎ | 207/761 [1:03:17<3:57:04, 25.68s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 28%|████████████████████████████████████████▊ | 210/761 [1:03:47<2:20:03, 15.25s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 28%|█████████████████████████████████████████▌ | 214/761 [1:04:56<2:31:00, 16.56s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 28%|██████████████████████████████████████████ | 216/761 [1:05:25<2:23:42, 15.82s/it]a lot of vad silence was found, try reducing aggressiveness\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ " 29%|██████████████████████████████████████████▏ | 217/761 [1:06:09<3:38:21, 24.08s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 29%|██████████████████████████████████████████▉ | 221/761 [1:08:09<3:59:09, 26.57s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 29%|███████████████████████████████████████████▏ | 222/761 [1:08:59<5:00:48, 33.48s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 29%|███████████████████████████████████████████▎ | 223/761 [1:09:05<3:48:43, 25.51s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 30%|████████████████████████████████████████████▏ | 227/761 [1:10:07<2:33:43, 17.27s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 30%|████████████████████████████████████████████▋ | 230/761 [1:11:17<2:46:16, 18.79s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 31%|██████████████████████████████████████████████ | 237/761 [1:13:17<3:12:06, 22.00s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 32%|██████████████████████████████████████████████▋ | 240/761 [1:13:55<2:15:58, 15.66s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 32%|███████████████████████████████████████████████▍ | 244/761 [1:14:59<2:24:59, 16.83s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 32%|███████████████████████████████████████████████▊ | 246/761 [1:15:28<2:08:34, 14.98s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 33%|████████████████████████████████████████████████▍ | 249/761 [1:16:34<2:27:01, 17.23s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 33%|████████████████████████████████████████████████▌ | 250/761 [1:16:44<2:09:28, 15.20s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 33%|████████████████████████████████████████████████▊ | 251/761 [1:17:04<2:20:14, 16.50s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 33%|█████████████████████████████████████████████████ | 252/761 [1:17:24<2:29:46, 17.66s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 34%|█████████████████████████████████████████████████▌ | 255/761 [1:18:04<2:09:45, 15.39s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 34%|█████████████████████████████████████████████████▊ | 256/761 [1:18:47<3:18:59, 23.64s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 34%|██████████████████████████████████████████████████▊ | 261/761 [1:20:05<2:43:46, 19.65s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 35%|███████████████████████████████████████████████████▎ | 264/761 [1:21:16<2:50:04, 20.53s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 35%|███████████████████████████████████████████████████▌ | 265/761 [1:21:31<2:36:08, 18.89s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 35%|███████████████████████████████████████████████████▋ | 266/761 [1:21:43<2:17:24, 16.66s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 35%|███████████████████████████████████████████████████▉ | 267/761 [1:21:50<1:54:19, 13.89s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 36%|████████████████████████████████████████████████████▋ | 271/761 [1:22:52<1:56:58, 14.32s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 36%|█████████████████████████████████████████████████████ | 273/761 [1:23:54<2:56:53, 21.75s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 37%|██████████████████████████████████████████████████████▎ | 279/761 [1:25:34<2:04:26, 15.49s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 37%|██████████████████████████████████████████████████████▊ | 282/761 [1:26:28<2:11:52, 16.52s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 37%|███████████████████████████████████████████████████████ | 283/761 [1:26:39<1:56:43, 14.65s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 37%|███████████████████████████████████████████████████████▍ | 285/761 [1:27:03<1:49:55, 13.86s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 38%|████████████████████████████████████████████████████████▌ | 291/761 [1:29:03<2:02:44, 15.67s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 39%|█████████████████████████████████████████████████████████▉ | 298/761 [1:30:41<1:40:59, 13.09s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 39%|██████████████████████████████████████████████████████████▎ | 300/761 [1:31:50<2:55:38, 22.86s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 40%|███████████████████████████████████████████████████████████ | 304/761 [1:34:07<3:50:04, 30.21s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 40%|███████████████████████████████████████████████████████████▋ | 307/761 [1:35:33<3:23:34, 26.90s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 41%|████████████████████████████████████████████████████████████ | 309/761 [1:36:22<3:07:28, 24.89s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 41%|████████████████████████████████████████████████████████████▊ | 313/761 [1:37:42<2:52:09, 23.06s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 41%|█████████████████████████████████████████████████████████████ | 314/761 [1:38:25<3:37:20, 29.17s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 42%|█████████████████████████████████████████████████████████████▍ | 316/761 [1:39:15<3:08:31, 25.42s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 42%|██████████████████████████████████████████████████████████████▊ | 323/761 [1:41:28<2:51:34, 23.50s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 43%|███████████████████████████████████████████████████████████████▌ | 327/761 [1:42:23<1:49:57, 15.20s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 43%|███████████████████████████████████████████████████████████████▉ | 329/761 [1:43:51<3:37:06, 30.15s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 44%|████████████████████████████████████████████████████████████████▊ | 333/761 [1:44:54<2:13:31, 18.72s/it]very little vad silence was found, try increasing aggressiveness\n", " 44%|████████████████████████████████████████████████████████████████▉ | 334/761 [1:45:01<1:47:49, 15.15s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 44%|█████████████████████████████████████████████████████████████████▌ | 337/761 [1:45:40<1:37:28, 13.79s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 45%|███████████████████████████████████████████████████████████████████▎ | 346/761 [1:47:14<1:11:27, 10.33s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 46%|███████████████████████████████████████████████████████████████████▍ | 347/761 [1:47:33<1:27:57, 12.75s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 46%|███████████████████████████████████████████████████████████████████▋ | 348/761 [1:47:46<1:29:00, 12.93s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 46%|████████████████████████████████████████████████████████████████████▋ | 353/761 [1:49:15<1:35:52, 14.10s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 48%|██████████████████████████████████████████████████████████████████████▊ | 364/761 [1:52:33<1:49:10, 16.50s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 48%|██████████████████████████████████████████████████████████████████████▉ | 365/761 [1:52:46<1:42:08, 15.48s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 48%|███████████████████████████████████████████████████████████████████████▏ | 366/761 [1:53:04<1:46:09, 16.13s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 48%|███████████████████████████████████████████████████████████████████████▊ | 369/761 [1:54:05<1:54:13, 17.48s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 49%|████████████████████████████████████████████████████████████████████████▉ | 375/761 [1:55:48<2:16:30, 21.22s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 49%|█████████████████████████████████████████████████████████████████████████ | 376/761 [1:56:09<2:15:28, 21.11s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 50%|█████████████████████████████████████████████████████████████████████████▎ | 377/761 [1:56:53<2:58:04, 27.83s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 50%|█████████████████████████████████████████████████████████████████████████▉ | 380/761 [1:58:02<2:23:10, 22.55s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 51%|██████████████████████████████████████████████████████████████████████████▉ | 385/761 [1:59:41<2:38:57, 25.37s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 51%|███████████████████████████████████████████████████████████████████████████▍ | 388/761 [2:00:57<2:25:06, 23.34s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 51%|███████████████████████████████████████████████████████████████████████████▋ | 389/761 [2:01:10<2:04:34, 20.09s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 52%|████████████████████████████████████████████████████████████████████████████▍ | 393/761 [2:02:44<2:05:28, 20.46s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 52%|█████████████████████████████████████████████████████████████████████████████▍ | 398/761 [2:04:21<1:52:05, 18.53s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 53%|█████████████████████████████████████████████████████████████████████████████▊ | 400/761 [2:04:46<1:32:23, 15.36s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 54%|███████████████████████████████████████████████████████████████████████████████▎ | 408/761 [2:06:37<1:09:53, 11.88s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 54%|███████████████████████████████████████████████████████████████████████████████▋ | 410/761 [2:06:58<1:05:16, 11.16s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 54%|████████████████████████████████████████████████████████████████████████████████▌ | 414/761 [2:07:57<1:16:01, 13.15s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 55%|████████████████████████████████████████████████████████████████████████████████▋ | 415/761 [2:08:21<1:36:06, 16.67s/it]a lot of vad silence was found, try reducing aggressiveness\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ " 55%|█████████████████████████████████████████████████████████████████████████████████ | 417/761 [2:08:46<1:23:01, 14.48s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 55%|█████████████████████████████████████████████████████████████████████████████████▍ | 419/761 [2:09:04<1:08:12, 11.97s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 55%|█████████████████████████████████████████████████████████████████████████████████▋ | 420/761 [2:09:47<2:02:01, 21.47s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 55%|█████████████████████████████████████████████████████████████████████████████████▉ | 421/761 [2:10:00<1:46:02, 18.71s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 56%|██████████████████████████████████████████████████████████████████████████████████▍ | 424/761 [2:10:44<1:29:52, 16.00s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 56%|██████████████████████████████████████████████████████████████████████████████████▋ | 425/761 [2:11:27<2:15:23, 24.18s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 57%|████████████████████████████████████████████████████████████████████████████████████ | 432/761 [2:13:17<1:37:05, 17.71s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 57%|████████████████████████████████████████████████████████████████████████████████████▊ | 436/761 [2:15:01<1:59:35, 22.08s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 58%|█████████████████████████████████████████████████████████████████████████████████████▏ | 438/761 [2:15:48<2:05:14, 23.26s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 58%|██████████████████████████████████████████████████████████████████████████████████████▎ | 444/761 [2:17:42<1:40:47, 19.08s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 58%|██████████████████████████████████████████████████████████████████████████████████████▌ | 445/761 [2:18:03<1:44:09, 19.78s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 60%|████████████████████████████████████████████████████████████████████████████████████████▉ | 457/761 [2:20:55<1:18:48, 15.56s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 60%|█████████████████████████████████████████████████████████████████████████████████████████▎ | 459/761 [2:21:55<1:58:17, 23.50s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 61%|█████████████████████████████████████████████████████████████████████████████████████████▊ | 462/761 [2:24:09<3:11:32, 38.44s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 61%|██████████████████████████████████████████████████████████████████████████████████████████ | 463/761 [2:24:16<2:23:35, 28.91s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 62%|███████████████████████████████████████████████████████████████████████████████████████████▍ | 470/761 [2:27:43<1:58:45, 24.49s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 63%|████████████████████████████████████████████████████████████████████████████████████████████▌ | 476/761 [2:29:21<1:25:55, 18.09s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 64%|██████████████████████████████████████████████████████████████████████████████████████████████▎ | 485/761 [2:31:16<1:00:33, 13.16s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 64%|██████████████████████████████████████████████████████████████████████████████████████████████▋ | 487/761 [2:31:52<1:10:25, 15.42s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 64%|██████████████████████████████████████████████████████████████████████████████████████████████▉ | 488/761 [2:32:21<1:28:37, 19.48s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 64%|███████████████████████████████████████████████████████████████████████████████████████████████▎ | 490/761 [2:32:46<1:10:33, 15.62s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 65%|████████████████████████████████████████████████████████████████████████████████████████████████ | 494/761 [2:34:17<1:38:47, 22.20s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 65%|████████████████████████████████████████████████████████████████████████████████████████████████▎ | 495/761 [2:34:25<1:19:09, 17.85s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 65%|████████████████████████████████████████████████████████████████████████████████████████████████▍ | 496/761 [2:34:38<1:12:32, 16.42s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 66%|█████████████████████████████████████████████████████████████████████████████████████████████████ | 499/761 [2:35:21<1:07:31, 15.46s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 66%|█████████████████████████████████████████████████████████████████████████████████████████████████▏ | 500/761 [2:35:41<1:12:37, 16.69s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 66%|█████████████████████████████████████████████████████████████████████████████████████████████████▋ | 502/761 [2:36:08<1:07:34, 15.65s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 66%|███████████████████████████████████████████████████████████████████████████████████████████████████▋ | 506/761 [2:37:07<58:19, 13.72s/it]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "weird coverage in df5d1b13-a1c5-42f1-8936-69bdfc9ebe2c\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ " 67%|██████████████████████████████████████████████████████████████████████████████████████████████████▌ | 507/761 [2:37:25<1:03:01, 14.89s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 67%|████████████████████████████████████████████████████████████████████████████████████████████████████▎ | 509/761 [2:37:53<59:20, 14.13s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 68%|█████████████████████████████████████████████████████████████████████████████████████████████████████▌ | 515/761 [2:39:18<47:33, 11.60s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 68%|████████████████████████████████████████████████████████████████████████████████████████████████████▌ | 517/761 [2:40:27<1:32:48, 22.82s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 68%|████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 519/761 [2:41:37<1:52:43, 27.95s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 68%|█████████████████████████████████████████████████████████████████████████████████████████████████████▏ | 520/761 [2:41:49<1:32:41, 23.08s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 69%|██████████████████████████████████████████████████████████████████████████████████████████████████████ | 525/761 [2:43:11<1:07:33, 17.18s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 69%|██████████████████████████████████████████████████████████████████████████████████████████████████████▎ | 526/761 [2:43:33<1:12:49, 18.59s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 69%|██████████████████████████████████████████████████████████████████████████████████████████████████████▍ | 527/761 [2:43:44<1:03:34, 16.30s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 69%|████████████████████████████████████████████████████████████████████████████████████████████████████████ | 528/761 [2:43:56<58:43, 15.12s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 70%|███████████████████████████████████████████████████████████████████████████████████████████████████████ | 530/761 [2:44:47<1:23:46, 21.76s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 70%|███████████████████████████████████████████████████████████████████████████████████████████████████████▎ | 531/761 [2:45:03<1:16:48, 20.04s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 70%|███████████████████████████████████████████████████████████████████████████████████████████████████████▍ | 532/761 [2:45:30<1:24:35, 22.16s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 71%|████████████████████████████████████████████████████████████████████████████████████████████████████████▍ | 537/761 [2:47:05<1:12:12, 19.34s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 71%|████████████████████████████████████████████████████████████████████████████████████████████████████████▋ | 538/761 [2:47:19<1:06:10, 17.80s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 71%|█████████████████████████████████████████████████████████████████████████████████████████████████████████ | 540/761 [2:48:22<1:26:21, 23.44s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 71%|█████████████████████████████████████████████████████████████████████████████████████████████████████████▍ | 542/761 [2:49:03<1:21:14, 22.26s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 71%|█████████████████████████████████████████████████████████████████████████████████████████████████████████▌ | 543/761 [2:49:14<1:07:57, 18.70s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 72%|████████████████████████████████████████████████████████████████████████████████████████████████████████████▏ | 549/761 [2:50:35<55:32, 15.72s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 72%|████████████████████████████████████████████████████████████████████████████████████████████████████████████▍ | 550/761 [2:50:47<51:21, 14.60s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 73%|████████████████████████████████████████████████████████████████████████████████████████████████████████████▊ | 552/761 [2:51:13<49:05, 14.10s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 73%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████ | 553/761 [2:51:27<49:07, 14.17s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 73%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████▌ | 556/761 [2:52:06<47:33, 13.92s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 74%|████████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 560/761 [2:53:33<1:11:01, 21.20s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 74%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████▎ | 562/761 [2:54:43<1:30:45, 27.36s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 74%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████▋ | 564/761 [2:55:16<1:09:50, 21.27s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 74%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████▌ | 566/761 [2:55:39<53:00, 16.31s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 75%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 568/761 [2:56:02<44:54, 13.96s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 77%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 583/761 [2:59:28<44:26, 14.98s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 78%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎ | 590/761 [3:01:22<49:55, 17.52s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 78%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 593/761 [3:02:07<44:58, 16.06s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 78%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋ | 597/761 [3:02:58<39:04, 14.29s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 79%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋ | 600/761 [3:04:34<1:04:30, 24.04s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 80%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▍ | 606/761 [3:06:49<50:39, 19.61s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 80%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋ | 607/761 [3:07:01<43:58, 17.13s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 80%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊ | 608/761 [3:07:13<39:38, 15.55s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 80%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▍ | 611/761 [3:07:48<32:40, 13.07s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 80%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋ | 612/761 [3:08:01<32:00, 12.89s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 81%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊ | 613/761 [3:08:17<34:07, 13.83s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 81%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 617/761 [3:09:58<1:04:10, 26.74s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 81%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏ | 618/761 [3:10:43<1:16:44, 32.20s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 81%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏ | 620/761 [3:11:09<52:13, 22.22s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 82%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 624/761 [3:11:56<32:38, 14.29s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 84%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎ | 636/761 [3:15:24<45:15, 21.72s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 84%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌ | 637/761 [3:15:35<38:23, 18.58s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 84%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊ | 638/761 [3:16:19<53:39, 26.17s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 84%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 639/761 [3:16:33<45:27, 22.36s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 84%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏ | 640/761 [3:17:02<49:16, 24.44s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 84%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌ | 642/761 [3:17:30<37:00, 18.66s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 84%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋ | 643/761 [3:17:44<33:58, 17.28s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 85%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 644/761 [3:18:14<40:55, 20.99s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 85%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎ | 646/761 [3:19:09<44:12, 23.06s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 85%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌ | 647/761 [3:19:56<57:11, 30.10s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 85%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ | 648/761 [3:20:39<1:04:00, 33.99s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 85%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 649/761 [3:20:58<55:01, 29.48s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 86%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋ | 653/761 [3:22:37<54:58, 30.54s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 86%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ | 655/761 [3:23:31<51:51, 29.35s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 86%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌ | 657/761 [3:23:58<36:39, 21.15s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 87%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ | 660/761 [3:24:53<30:01, 17.83s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 87%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎ | 661/761 [3:25:05<26:35, 15.95s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 87%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▍ | 662/761 [3:25:23<27:30, 16.67s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 87%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 664/761 [3:25:52<25:17, 15.64s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 88%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▍ | 672/761 [3:28:05<22:57, 15.48s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 88%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋ | 673/761 [3:28:28<26:05, 17.79s/it]a lot of vad silence was found, try reducing aggressiveness\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ " 89%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ | 675/761 [3:29:02<24:40, 17.21s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 89%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊ | 679/761 [3:30:05<22:22, 16.37s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 90%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▍ | 687/761 [3:31:53<15:31, 12.59s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 91%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊ | 689/761 [3:32:11<12:50, 10.70s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 91%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏ | 696/761 [3:34:02<13:34, 12.54s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 92%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▍ | 697/761 [3:34:45<23:16, 21.82s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 92%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌ | 698/761 [3:34:54<18:44, 17.84s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 92%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 700/761 [3:35:21<16:21, 16.09s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 92%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏ | 701/761 [3:35:43<18:03, 18.06s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 93%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 705/761 [3:37:18<23:22, 25.05s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 93%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎ | 707/761 [3:38:11<22:05, 24.55s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 93%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊ | 709/761 [3:38:34<15:47, 18.22s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 94%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎ | 717/761 [3:41:09<15:04, 20.57s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 95%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ | 721/761 [3:42:26<13:24, 20.12s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 95%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎ | 722/761 [3:42:45<12:52, 19.81s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 95%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌ | 723/761 [3:43:13<14:12, 22.42s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 96%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▍ | 728/761 [3:45:07<10:41, 19.45s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 96%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎ | 732/761 [3:46:00<06:41, 13.84s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 96%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋ | 734/761 [3:46:44<08:38, 19.22s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 97%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 735/761 [3:46:56<07:28, 17.24s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 97%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋ | 739/761 [3:47:56<05:39, 15.41s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 97%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊ | 740/761 [3:48:43<08:42, 24.90s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 97%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ | 741/761 [3:49:05<08:02, 24.12s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 98%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎ | 742/761 [3:49:48<09:24, 29.70s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 98%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▍ | 743/761 [3:50:06<07:55, 26.40s/it]a lot of vad silence was found, try reducing aggressiveness\n", " 98%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋ | 749/761 [3:52:01<03:33, 17.76s/it]a lot of vad silence was found, try reducing aggressiveness\n", "100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊| 760/761 [3:55:54<00:17, 17.38s/it]a lot of vad silence was found, try reducing aggressiveness\n", "100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 761/761 [3:56:37<00:00, 18.66s/it]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "61/761 files excluded\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "\n" ] } ], "source": [ "import pandas as pd\n", "import tqdm\n", "import uuid\n", "\n", "def get_hotwords_from_filename(fname):\n", " fn = os.path.split(fname)[-1]\n", " fn = os.path.splitext(fn)[0]\n", " fn = \"_\".join(fn.split(\"_\")[:-1])\n", " fn = fn.replace(\" \", \"_\").replace(\"-\", \"_\").replace(\"__\", \"_\")\n", " fn = (\n", " fn.replace(\"_Male_\", \"_\")\n", " .replace(\"_Female_\", \"_\")\n", " .replace(\"Tagalog\", \"_\")\n", " .replace(\"_PH_\", \"_\")\n", " .strip(\"_\")\n", " )\n", " return fn.split(\"_\")\n", "\n", "metadata = []\n", "failed_metadata = []\n", "# for filepath in tqdm.tqdm(filepaths_1[:10] + filepaths_2[:10]):\n", "for filepath in tqdm.tqdm((filepaths_1 + filepaths_2)):\n", " \n", " try:\n", " audio = Audio.from_file(filepath, sample_rate=16_000, byte_width=2)\n", " audio_rev, submission_meta = prep_annotation_audio(audio)\n", " except:\n", " failed_metadata.append({\n", " \"original_audio_filepath\": filepath,\n", " })\n", " continue\n", " \n", " # make filter logic\n", " if audio_rev is None:\n", " failed_metadata.append({\n", " \"original_audio_filepath\": filepath,\n", " \"misc_meta\": {\n", " \"tot_silence_frac\": submission_meta[\"tot_silence_frac\"],\n", " \"segment_cov_frac\": submission_meta[\"segment_cov_frac\"],\n", " }\n", " })\n", " continue\n", " \n", " uid = str(uuid.uuid4())\n", " rev_filepath = os.path.join(TO_REV_AUDIO_DIR, uid + \".mp3\")\n", " audio_rev.to_mp3(rev_filepath)\n", " hotwords = get_hotwords_from_filename(filepath)\n", " \n", " rough_efficiency = submission_meta[\"tot_silence_frac\"] + submission_meta[\"segment_cov_frac\"]\n", " if rough_efficiency > 1.5 or rough_efficiency < 0.5:\n", " print(f\"weird coverage in {uid}\")\n", " \n", " metadata.append({\n", " \"uid\": uid,\n", " \"original_audio_filepath\": filepath,\n", " \"segments_meta\": submission_meta[\"segments_meta\"],\n", " \"audio_duration_s\": audio_rev.duration_s,\n", " \"audio_filepath\": rev_filepath,\n", " \"hotwords\": hotwords,\n", " \"misc_meta\": {\n", " \"tot_silence_frac\": submission_meta[\"tot_silence_frac\"],\n", " \"segment_cov_frac\": submission_meta[\"segment_cov_frac\"],\n", " \"stitch_overhead_frac\": submission_meta[\"stitch_overhead_frac\"],\n", " \"tot_overhead_frac\": submission_meta[\"tot_overhead_frac\"],\n", " }\n", " })\n", "\n", "print(\"{}/{} files excluded\".format(len(failed_metadata), len(failed_metadata) + len(metadata)))" ] }, { "cell_type": "code", "execution_count": null, "id": "06ab393e", "metadata": {}, "outputs": [], "source": [ "# many: a lot of vad silence was found, try reducing aggressiveness\n", "# weird coverage in df5d1b13-a1c5-42f1-8936-69bdfc9ebe2c\n", "# 61/761 files excluded" ] }, { "cell_type": "code", "execution_count": 20, "id": "0424cf83", "metadata": {}, "outputs": [], "source": [ "with open(REV_MANIFEST_FILEPATH, \"w\") as f:\n", " for meta in metadata:\n", " f.write(json.dumps(meta) + \"\\n\")\n", " \n", "with open(FAILED_MANIFEST_FILEPATH, \"w\") as f:\n", " for meta in failed_metadata:\n", " f.write(json.dumps(meta) + \"\\n\")" ] }, { "cell_type": "code", "execution_count": null, "id": "376e7c6b", "metadata": {}, "outputs": [], "source": [ "# 61/761 files excluded\n", "# 3ea5d3b3-4826-4e1e-be7a-e4d150e6f7b9 - background chatter being transcribed\n", "# 4cd4d2a1-99b9-4b34-9d5c-6a936d930fbd - all silence giving few empty slices" ] }, { "cell_type": "code", "execution_count": null, "id": "d19ec5ea", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "code", "execution_count": null, "id": "aaaf9cf0", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "code", "execution_count": null, "id": "80430e0e", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "markdown", "id": "149b3b8e", "metadata": {}, "source": [ "## (optional) Investigate for mistakes" ] }, { "cell_type": "code", "execution_count": 129, "id": "143da265", "metadata": {}, "outputs": [], "source": [ "metadata = []\n", "with open(REV_MANIFEST_FILEPATH) as f:\n", " for l in f.read().strip().split(\"\\n\"):\n", " metadata.append(json.loads(l))" ] }, { "cell_type": "code", "execution_count": 50, "id": "e6d1a1bc", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "found\n" ] } ], "source": [ "for meta in metadata:\n", " if meta[\"uid\"] == \"29c7be40-46ef-4000-a8e4-e0a4e4a6b828\":\n", " print(\"found\")\n", " break\n", "# print(meta[\"misc_meta\"][\"tot_silence_frac\"] + meta[\"misc_meta\"][\"segment_cov_frac\"])" ] }, { "cell_type": "code", "execution_count": null, "id": "e0489de1", "metadata": {}, "outputs": [], "source": [ "# 2 got cancelled\n", "# 29c7be40-46ef-4000-a8e4-e0a4e4a6b828 - \n", "# 48bdcb66-c070-42e4-9cec-7fa0652787e7 - 1 min of background chatter" ] }, { "cell_type": "code", "execution_count": 52, "id": "9fe48a0d", "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", "\n" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "play_audio(\"/mnt/data-ssd-1/data/private/customer/sanas/2022-08-04-fili-callcenter/to_rev/2022_08_07/audio/29c7be40-46ef-4000-a8e4-e0a4e4a6b828.mp3\")" ] }, { "cell_type": "code", "execution_count": null, "id": "1b79d2e8", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "code", "execution_count": null, "id": "1b3ca219", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "code", "execution_count": null, "id": "6c91afe9", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "markdown", "id": "0868d256", "metadata": {}, "source": [ "## (Optional) Check price" ] }, { "cell_type": "code", "execution_count": 44, "id": "37decb32", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "700 files total\n" ] }, { "data": { "text/plain": [ "suno 2334.3\n", "sanas 4715.9\n", "dtype: float64" ] }, "execution_count": 44, "metadata": {}, "output_type": "execute_result" } ], "source": [ "suno_price_per_hour = 72\n", "suno_price_per_min = suno_price_per_hour / 60\n", "\n", "sanas_price = 162 # dollars_per_hour\n", "sanas_price_per_s = sanas_price / 3600 # dollars per s\n", "\n", "\n", "costs = []\n", "for rm in metadata:\n", " sanas_duration_s = np.sum([sm[\"duration_s\"] for sm in rm[\"segments_meta\"] if sm[\"type\"] == \"speech\"])\n", " sanas_cost = sanas_price_per_s * sanas_duration_s\n", " \n", " suno_cost = np.ceil(rm[\"audio_duration_s\"] / 60.) * suno_price_per_min\n", " costs.append({\"suno\": suno_cost, \"sanas\": sanas_cost})\n", "\n", "print(len(metadata), \"files total\")\n", "costs = pd.DataFrame(costs)\n", "# suno 2770.8\n", "# sanas 4715.9\n", "costs.sum().round(1)" ] }, { "cell_type": "code", "execution_count": null, "id": "def684e9", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "code", "execution_count": null, "id": "fcf743da", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "code", "execution_count": null, "id": "7a31702b", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "markdown", "id": "88a9e9cc", "metadata": {}, "source": [ "## Send stuff to rev" ] }, { "cell_type": "code", "execution_count": 24, "id": "613bd82e", "metadata": {}, "outputs": [], "source": [ "ORDER_REF_STR = \"sns\"" ] }, { "cell_type": "code", "execution_count": 25, "id": "686b5db2", "metadata": {}, "outputs": [], "source": [ "metadata = []\n", "with open(REV_MANIFEST_FILEPATH) as f:\n", " for l in f.read().strip().split(\"\\n\"):\n", " metadata.append(json.loads(l))" ] }, { "cell_type": "code", "execution_count": 26, "id": "e282c309", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "32.4h total\n" ] } ], "source": [ "tot_duration_s = [m[\"audio_duration_s\"] for m in metadata]\n", "print(len(metadata), \"files total\")\n", "print(\"{}h total\".format(round(sum(tot_duration_s) / 60 / 60, 1)))" ] }, { "cell_type": "code", "execution_count": 37, "id": "81f46bd0", "metadata": {}, "outputs": [], "source": [ "# metadata[0]" ] }, { "cell_type": "markdown", "id": "b16a0dd0", "metadata": {}, "source": [ "### 01 - upload items" ] }, { "cell_type": "code", "execution_count": 34, "id": "2453089e", "metadata": {}, "outputs": [], "source": [ "import json\n", "from suno_utils.web.rev import get_auth_string, upload_file\n", "\n", "with open(\"/home/georg/.secrets/secrets.json\") as f:\n", " secrets = json.load(f)\n", " \n", "CLIENT_API_KEY = secrets[\"REV_CLIENT_API_KEY\"]\n", "USER_API_KEY = secrets[\"REV_USER_API_KEY\"]\n", "AUTH_STR = get_auth_string(CLIENT_API_KEY, USER_API_KEY)" ] }, { "cell_type": "code", "execution_count": 38, "id": "8c66f7f9", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 700/700 [15:03<00:00, 1.29s/it]\n" ] } ], "source": [ "import tqdm\n", "\n", "order_items = []\n", "for audio_meta in tqdm.tqdm(metadata):\n", " media_loc = upload_file(AUTH_STR, audio_meta[\"audio_filepath\"], file_ref_str=audio_meta[\"uid\"])\n", " order_items.append({\n", " \"media_loc\": media_loc,\n", " \"hotwords\": audio_meta[\"hotwords\"],\n", "# \"accents\": [\"Australian\"],\n", " \"duration_s\": int(round(audio_meta[\"audio_duration_s\"])),\n", " \"uuid\": audio_meta[\"uid\"],\n", " })" ] }, { "cell_type": "code", "execution_count": 39, "id": "a0ee5f4f", "metadata": {}, "outputs": [], "source": [ "with open(os.path.join(TO_REV_DIR, \"01_order_items.json\"), \"w\") as f:\n", " json.dump(order_items, f)" ] }, { "cell_type": "markdown", "id": "9131358e", "metadata": {}, "source": [ "### 02 - make order" ] }, { "cell_type": "code", "execution_count": 40, "id": "7ccf6e0a", "metadata": {}, "outputs": [], "source": [ "import json\n", "from suno_utils.web.rev import get_auth_string, make_order\n", "\n", "with open(\"/home/georg/.secrets/secrets.json\") as f:\n", " secrets = json.load(f)\n", " \n", "CLIENT_API_KEY = secrets[\"REV_CLIENT_API_KEY\"]\n", "USER_API_KEY = secrets[\"REV_USER_API_KEY\"]\n", "AUTH_STR = get_auth_string(CLIENT_API_KEY, USER_API_KEY)" ] }, { "cell_type": "code", "execution_count": 41, "id": "0b300ef5", "metadata": {}, "outputs": [], "source": [ "with open(os.path.join(TO_REV_DIR, \"01_order_items.json\"), \"r\") as f:\n", " order_items = json.load(f)" ] }, { "cell_type": "code", "execution_count": 42, "id": "8318e46f", "metadata": {}, "outputs": [], "source": [ "order_number = make_order(AUTH_STR, order_items, order_ref_str=ORDER_REF_STR, add_rush=False)" ] }, { "cell_type": "code", "execution_count": 43, "id": "27549d51", "metadata": {}, "outputs": [], "source": [ "with open(os.path.join(TO_REV_DIR, \"02_order_numer.txt\"), \"w\") as f:\n", " f.write(order_number)" ] }, { "cell_type": "code", "execution_count": null, "id": "45935e64", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "code", "execution_count": null, "id": "9c5f5bd2", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "markdown", "id": "f30acc60", "metadata": {}, "source": [ "### 04 - pull transcripts" ] }, { "cell_type": "code", "execution_count": 53, "id": "65101ca4", "metadata": {}, "outputs": [], "source": [ "import json\n", "from suno_utils.web.rev import get_auth_string, get_finished_order, get_transcript\n", "\n", "with open(\"/home/georg/.secrets/secrets.json\") as f:\n", " secrets = json.load(f)\n", " \n", "CLIENT_API_KEY = secrets[\"REV_CLIENT_API_KEY\"]\n", "USER_API_KEY = secrets[\"REV_USER_API_KEY\"]\n", "AUTH_STR = get_auth_string(CLIENT_API_KEY, USER_API_KEY)" ] }, { "cell_type": "code", "execution_count": 54, "id": "de613d78", "metadata": {}, "outputs": [], "source": [ "import os\n", "\n", "with open(os.path.join(TO_REV_DIR, \"02_order_numer.txt\")) as f:\n", " order_number = f.read()" ] }, { "cell_type": "code", "execution_count": 56, "id": "9319c7f3", "metadata": {}, "outputs": [], "source": [ "# # get intermediate results\n", "# from suno_utils.web.rev import _get_order_details, _parse_order_response\n", "# order_details = _get_order_details(AUTH_STR, order_number)\n", "# order_status, transcript_metas = _parse_order_response(order_details)" ] }, { "cell_type": "code", "execution_count": 72, "id": "58ae9a27", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "698 transcripts\n" ] } ], "source": [ "transcript_metas = get_finished_order(AUTH_STR, order_number, blocking=True)\n", "print(len(transcript_metas), \"transcripts\")" ] }, { "cell_type": "code", "execution_count": 73, "id": "71cd19bb", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 698/698 [00:13<00:00, 51.37it/s]\n" ] } ], "source": [ "for transcript_meta in tqdm.tqdm(transcript_metas):\n", " transcript_filepath = os.path.join(FROM_REV_DIR, f\"{transcript_meta['uuid']}.json\")\n", " if os.path.exists(transcript_filepath):\n", " continue\n", " transcript_json = get_transcript(AUTH_STR, transcript_meta[\"rev_id\"])\n", " with open(transcript_filepath, \"w\") as f:\n", " json.dump(transcript_json, f)" ] }, { "cell_type": "code", "execution_count": null, "id": "2931df3a", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "code", "execution_count": null, "id": "5831d7d5", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "code", "execution_count": null, "id": "6ae45fc2", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "code", "execution_count": null, "id": "c201b924", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "markdown", "id": "da899a89", "metadata": {}, "source": [ "## Playground" ] }, { "cell_type": "code", "execution_count": null, "id": "7ca17081", "metadata": {}, "outputs": [], "source": [ "# def parse_hotwords(fname):\n", "# fn = os.path.split(fname)[-1]\n", "# fn = os.path.splitext(fn)[0]\n", "# fn = \"_\".join(fn.split(\"_\")[:-1])\n", "# fn = fn.replace(\" \", \"_\").replace(\"-\", \"_\").replace(\"__\", \"_\")\n", "# fn = (\n", "# fn.replace(\"_Male_\", \"_\")\n", "# .replace(\"_Female_\", \"_\")\n", "# .replace(\"Tagalog\", \"_\")\n", "# .replace(\"_PH_\", \"_\")\n", "# .strip(\"_\")\n", "# )\n", "# return fn.split(\"_\")" ] }, { "cell_type": "code", "execution_count": 48, "id": "f2b0ac15", "metadata": {}, "outputs": [], "source": [ "# for start_s, end_s in speech_segments:\n", "# print(start_s, end_s)\n", "# audio.get_slice(start_s, end_s).play()\n", "# print(\"-\"*10)" ] }, { "cell_type": "code", "execution_count": null, "id": "1c6f5551", "metadata": {}, "outputs": [], "source": [ "# a = Audio.from_file(\"audio_indicators/full_numbers.wav\")\n", "# a.get_slice(0.8, 1.3).to_wav(\"audio_indicators/a.wav\")\n", "# a.get_slice(1.4, 1.9).to_wav(\"audio_indicators/0.wav\")\n", "# a.get_slice(2.0, 2.5).to_wav(\"audio_indicators/1.wav\")\n", "# a.get_slice(2.6, 3.1).to_wav(\"audio_indicators/2.wav\")\n", "# a.get_slice(3.2, 3.7).to_wav(\"audio_indicators/3.wav\")\n", "# a.get_slice(3.8, 4.3).to_wav(\"audio_indicators/4.wav\")\n", "# a.get_slice(4.3, 4.8).to_wav(\"audio_indicators/5.wav\")\n", "# a.get_slice(4.9, 5.4).to_wav(\"audio_indicators/6.wav\")\n", "# a.get_slice(5.4, 5.9).to_wav(\"audio_indicators/7.wav\")\n", "# a.get_slice(6.0, 6.5).to_wav(\"audio_indicators/8.wav\")\n", "# a.get_slice(6.6, 7.1).to_wav(\"audio_indicators/9.wav\")" ] }, { "cell_type": "code", "execution_count": 58, "id": "5d5e8c4c", "metadata": {}, "outputs": [], "source": [ "# !sox audio_indicators/a.wav audio_indicators/a_fast.wav tempo 1.5\n", "# !sox audio_indicators/0.wav audio_indicators/0_fast.wav tempo 1.5\n", "# !sox audio_indicators/1.wav audio_indicators/1_fast.wav tempo 1.5\n", "# !sox audio_indicators/2.wav audio_indicators/2_fast.wav tempo 1.5\n", "# !sox audio_indicators/3.wav audio_indicators/3_fast.wav tempo 1.5\n", "# !sox audio_indicators/4.wav audio_indicators/4_fast.wav tempo 1.5\n", "# !sox audio_indicators/5.wav audio_indicators/5_fast.wav tempo 1.5\n", "# !sox audio_indicators/6.wav audio_indicators/6_fast.wav tempo 1.5\n", "# !sox audio_indicators/7.wav audio_indicators/7_fast.wav tempo 1.5\n", "# !sox audio_indicators/8.wav audio_indicators/8_fast.wav tempo 1.5\n", "# !sox audio_indicators/9.wav audio_indicators/9_fast.wav tempo 1.5" ] }, { "cell_type": "code", "execution_count": null, "id": "62e669ac", "metadata": {}, "outputs": [], "source": [ "np.sum([g.upper - g.lower for g in vad_gaps]) / get_audio_properties(filepaths_2[0])[\"duration_s\"]" ] }, { "cell_type": "code", "execution_count": 76, "id": "9de73c5a", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "/home/georg/notebooks/customers/sanas\r\n" ] } ], "source": [ "!pwd" ] }, { "cell_type": "code", "execution_count": null, "id": "849cb2ad", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "code", "execution_count": null, "id": "55706e93", "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "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": "3.8.10" } }, "nbformat": 4, "nbformat_minor": 5 }