{"id":179,"date":"2023-03-04T23:48:17","date_gmt":"2023-03-04T14:48:17","guid":{"rendered":"https:\/\/fukugyouhistory.tokyo\/?p=179"},"modified":"2023-03-27T12:05:17","modified_gmt":"2023-03-27T03:05:17","slug":"whisper%e3%80%80%e9%9f%b3%e5%a3%b0%e3%83%bb%e5%8b%95%e7%94%bb%e3%81%ae%e8%87%aa%e5%8b%95%e6%9b%b8%e3%81%8d%e8%b5%b7%e3%81%93%e3%81%97ai%e3%82%92%e7%b0%a1%e5%8d%98%e3%81%ab%e4%bd%bf%e3%81%8a%e3%81%86","status":"publish","type":"post","link":"https:\/\/fukugyouhistory.tokyo\/?p=179","title":{"rendered":"Whisper\u3000\u97f3\u58f0\u30fb\u52d5\u753b\u306e\u81ea\u52d5\u66f8\u304d\u8d77\u3053\u3057AI\u3092\u7121\u6599\u3067\u3001\u7c21\u5358\u306b\u4f7f\u304a\u3046"},"content":{"rendered":"\n<p>OpenAI\u304c\u516c\u958b\u3057\u3066\u3044\u308bWhisper\uff08\u97f3\u58f0\u30fb\u52d5\u753b\u306e\u81ea\u52d5\u66f8\u304d\u8d77\u3053\u3057AI\uff09\u3092<br>\u7c21\u5358\u306b\u8a66\u3059\u65b9\u6cd5\u304c\u3042\u308a\u307e\u3057\u305f\u306e\u3067\u3001\u8a18\u9332\u3068\u3057\u3066\u6b8b\u3057\u307e\u3059\u3002<\/p>\n\n\n\n<p>\u526f\u696d\u306e\u300c\u6587\u5b57\u8d77\u3053\u3057\u300d\u300c\u30c6\u30fc\u30d7\u8d77\u3053\u3057\u300d\u300c\u52d5\u753b\u306e\u30c6\u30ed\u30c3\u30d7\u4f5c\u6210\u300d\u306a\u3069<br>\u975e\u5e38\u306b\u6709\u52b9\u306a\u6b66\u5668\u3068\u306a\u308a\u305d\u3046\u306a\u306e\u3067\u3001\u8eab\u306b\u4ed8\u3051\u308b\u3079\u304f\u3002<\/p>\n\n\n\n<!--more-->\n\n\n\n<p><\/p>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 class=\"wp-block-heading\">\u52d5\u304b\u3057\u3066\u307f\u3088\u3046<\/h2>\n\n\n\n<p>\u81ea\u5206\u306e\u30ed\u30fc\u30ab\u30eb\u306e\u30d1\u30bd\u30b3\u30f3\u306b\u3001python\u306e\u74b0\u5883\u3092\u7528\u610f\u3057\u3066\u5b9f\u884c\u3082\u3057\u307e\u3057\u305f\u304c\u3001<br>\u5727\u5012\u7684\u306bGoogle Colaboratory\u3067\u5b9f\u884c\u3059\u308b\u65b9\u304c\u7c21\u5358\u3067\u3057\u305f\u3002<\/p>\n\n\n\n<p>\u3057\u304b\u3082\u3001Google Colaboratory\u306e\u5229\u7528\u306f\u30011\u65e5\u306e\u4f7f\u7528\u91cf\u306b\u5236\u9650\u306f\u3042\u308a\u307e\u3059\u304c\u7121\u6599\u3067\u3059\u3002<\/p>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h3 class=\"wp-block-heading\">Google Colaboratory\u306b\u30a2\u30af\u30bb\u30b9<\/h3>\n\n\n\n<p>Google\u30a2\u30ab\u30a6\u30f3\u30c8\u3092\u53d6\u5f97\u3057\u3066\u3044\u308b\u524d\u63d0\u3067\u3059\u304c\u3001<br><a rel=\"noreferrer noopener\" href=\"https:\/\/colab.research.google.com\/?hl=ja\" target=\"_blank\">Google Colaboratory<\/a>\u306b\u30a2\u30af\u30bb\u30b9\u3057\u307e\u3059\u3002\u300c\u30ce\u30fc\u30c8\u30d6\u30c3\u30af\u306e\u65b0\u898f\u4f5c\u6210\u300d\u3092\u9078\u629e\u3057\u307e\u3059\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-large is-resized\"><img decoding=\"async\" src=\"https:\/\/fukugyouhistory.tokyo\/wp-content\/uploads\/2023\/03\/Colaboratory-Colaboratory-Brave-2023_03_04-23_16_47-944x1024.png\" alt=\"\" class=\"wp-image-181\" width=\"419\" height=\"454\"\/><\/figure>\n\n\n\n<p>\u6700\u521d\u306b\u3001\u300c\u7de8\u96c6\u300d\u2192\u300c\u30ce\u30fc\u30c8\u30d6\u30c3\u30af\u306e\u8a2d\u5b9a\u300d\u304b\u3089\u3001<br>\u30cf\u30fc\u30c9\u30a6\u30a7\u30a2\u30a2\u30af\u30bb\u30e9\u30ec\u30fc\u30c8\u3092\u300cGPU\u300d\u306b\u8a2d\u5b9a\u3057\u307e\u3057\u3087\u3046\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-full is-resized\"><img decoding=\"async\" sizes=\"(max-width: 880px) 100vw, 880px\" src=\"https:\/\/fukugyouhistory.tokyo\/wp-content\/uploads\/2023\/03\/image-35.png\" alt=\"\" class=\"wp-image-182\" width=\"420\" height=\"164\" srcset=\"https:\/\/fukugyouhistory.tokyo\/wp-content\/uploads\/2023\/03\/image-35.png 880w, https:\/\/fukugyouhistory.tokyo\/wp-content\/uploads\/2023\/03\/image-35-300x117.png 300w, https:\/\/fukugyouhistory.tokyo\/wp-content\/uploads\/2023\/03\/image-35-768x299.png 768w\" \/><\/figure>\n\n\n\n<p>\u4ee5\u4e0b\u306e\u30b3\u30de\u30f3\u30c9\u3092\u5165\u529b\u3057\u3001Shift+Enter\u3092\u62bc\u3057\u3066\u5b9f\u884c\u3057\u307e\u3057\u3087\u3046\u3002<br>GPU\u306e\u60c5\u5831\u304c\u51fa\u3066\u304f\u308c\u3070\u3001\u6b63\u5e38\u306b\u8a2d\u5b9a\u5909\u66f4\u3067\u304d\u3066\u3044\u307e\u3059\u3002<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>!nvidia-smi<\/code><\/pre>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h3 class=\"wp-block-heading\">Google\u30c9\u30e9\u30a4\u30d6\u3078\u306e\u30a2\u30af\u30bb\u30b9\u8a31\u53ef<\/h3>\n\n\n\n<p>\u8aad\u307f\u8fbc\u307e\u305b\u308b\u97f3\u58f0\u30d5\u30a1\u30a4\u30eb\u3092Google\u30c9\u30e9\u30a4\u30d6\u7d4c\u7531\u3067\u53d6\u5f97\u3067\u304d\u308b\u3088\u3046\u306b\u3001<br>\u30de\u30a6\u30f3\u30c8\uff08\u30c9\u30e9\u30a4\u30d6\u306e\u8a8d\u8b58\uff09\u3092\u3055\u305b\u307e\u3059\u3002<\/p>\n\n\n\n<p>\u4ee5\u4e0b\u306e\u30b3\u30fc\u30c9\u3092\u5165\u529b\u3057\u3001Shift+Enter\u3092\u62bc\u3057\u3066\u5b9f\u884c\u3057\u307e\u3057\u3087\u3046\u3002<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>from google.colab import drive\ndrive.mount('\/content\/drive')<\/code><\/pre>\n\n\n\n<p>Google\u30c9\u30e9\u30a4\u30d6\u3078\u306e\u63a5\u7d9a\u3092\u8a31\u53ef\u3057\u3066\u304f\u3060\u3055\u3044\u3002<br>\u300cMounted at \/content\/drive\u300d\u3068\u51fa\u308c\u3070\u3001\u6b63\u5e38\u306b\u5b8c\u4e86\u3057\u3066\u3044\u307e\u3059\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-large is-resized\"><img decoding=\"async\" src=\"https:\/\/fukugyouhistory.tokyo\/wp-content\/uploads\/2023\/03\/Colaboratory-Colaboratory-Brave-2023_03_04-23_30_50-944x1024.png\" alt=\"\" class=\"wp-image-185\" width=\"412\" height=\"447\"\/><\/figure>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h3 class=\"wp-block-heading\">Whisper\u306e\u30a4\u30f3\u30b9\u30c8\u30fc\u30eb<\/h3>\n\n\n\n<p>\u5fc5\u8981\u306a\u30d1\u30c3\u30b1\u30fc\u30b8\u306e\u30a4\u30f3\u30b9\u30c8\u30fc\u30eb\u3068\u3001\u4f7f\u7528\u3059\u308b\u30e2\u30c7\u30eb\u306e\u30b5\u30a4\u30ba\u3092\u6307\u5b9a\u3057\u307e\u3059\u3002<br>\u4ee5\u4e0b\u306e\u30b3\u30fc\u30c9\u3092\u5165\u529b\u3057\u3001Shift+Enter\u3092\u62bc\u3057\u3066\u5b9f\u884c\u3057\u307e\u3057\u3087\u3046\u3002<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>! pip install git+https:\/\/github.com\/openai\/whisper.git\nimport whisper\nmodel = whisper.load_model(\"medium\")<\/code><\/pre>\n\n\n\n<p>model\u306e\u5024\u306f\u3001\u4ee5\u4e0b\u306e\u901a\u308a\u3067\u3059\u3002<br>medium, large\u306e\u3069\u3061\u3089\u304b\u304c\u73fe\u5b9f\u7684\u306a\u9078\u629e\u80a2\u3067\u3059\u3002\u305d\u308c\u4ee5\u4e0b\u3060\u3068\u624b\u76f4\u3057\u304c\u5927\u5909\u3067\u3059\u3002<br>\u203b\u30ed\u30fc\u30ab\u30eb\u306e\u975e\u529b\u306a\u30d1\u30bd\u30b3\u30f3\u3067\u306f\u3001\u9006\u306bsmall, base\u3042\u305f\u308a\u3067\u306a\u3044\u3068\u52d5\u304d\u307e\u305b\u3093\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-full is-resized\"><img decoding=\"async\" sizes=\"(max-width: 843px) 100vw, 843px\" src=\"https:\/\/fukugyouhistory.tokyo\/wp-content\/uploads\/2023\/03\/image-36.png\" alt=\"\" class=\"wp-image-186\" width=\"583\" height=\"167\" srcset=\"https:\/\/fukugyouhistory.tokyo\/wp-content\/uploads\/2023\/03\/image-36.png 843w, https:\/\/fukugyouhistory.tokyo\/wp-content\/uploads\/2023\/03\/image-36-300x86.png 300w, https:\/\/fukugyouhistory.tokyo\/wp-content\/uploads\/2023\/03\/image-36-768x221.png 768w\" \/><\/figure>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h3 class=\"wp-block-heading\">Google\u30c9\u30e9\u30a4\u30d6\u306b\u97f3\u58f0\u30d5\u30a1\u30a4\u30eb\u3092\u7f6e\u304f<\/h3>\n\n\n\n<p>\u66f8\u304d\u8d77\u3053\u3057\u305f\u3044\u97f3\u58f0\u30d5\u30a1\u30a4\u30eb\u3092Google\u30c9\u30e9\u30a4\u30d6\u306e\u30de\u30a4\u30c9\u30e9\u30a4\u30d6\u76f4\u4e0b\u306b\u4fdd\u5b58\u3057\u307e\u3059\u3002<\/p>\n\n\n\n<p>\u7d9a\u3051\u3066\u3001\u4ee5\u4e0b\u306e\u30b3\u30fc\u30c9\u3092\u5165\u529b\u3057\u3001Shift+Enter\u3092\u62bc\u3057\u3066\u5b9f\u884c\u3057\u307e\u3057\u3087\u3046\u3002<br>\u203b\u30d5\u30a1\u30a4\u30eb\u540d\u306f\u9069\u5b9c\u4fee\u6b63\u3057\u3066\u304f\u3060\u3055\u3044\u3002<\/p>\n\n\n\n<p>\u4ee5\u4e0b\u306b\u30b5\u30f3\u30d7\u30eb\u30d5\u30a1\u30a4\u30eb\u300cg_03.mp3\u300d\u3092\u7f6e\u304d\u307e\u3059\u3002<\/p>\n\n\n\n<figure class=\"wp-block-audio\"><audio controls src=\"https:\/\/fukugyouhistory.tokyo\/wp-content\/uploads\/2023\/03\/g_03.mp3\"><\/audio><\/figure>\n\n\n\n<p>\u203bhttp:\/\/pro-video.jp\/voice\/announce\/\u3088\u308a\u5f15\u7528<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>file_name = \"g_03.mp3\"<\/code><\/pre>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h3 class=\"wp-block-heading\">Whisper\u306e\u5b9f\u884c<\/h3>\n\n\n\n<p>\u4ee5\u4e0b\u306e\u30b3\u30fc\u30c9\u3092\u5165\u529b\u3057\u3001Shift+Enter\u3092\u62bc\u3057\u3066\u5b9f\u884c\u3057\u307e\u3057\u3087\u3046\u3002<br>\u300cverbose=True\u300d\u3092\u6307\u5b9a\u3059\u308b\u3068\u3001\u51e6\u7406\u6700\u4e2d\u306e\u7d50\u679c\u3092\u8868\u793a\u3055\u305b\u308b\u3053\u3068\u304c\u3067\u304d\u307e\u3059\u3002\u304a\u597d\u307f\u3067\u3002<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>result = model.transcribe(\"\/content\/drive\/MyDrive\/\" + file_name)\n#result = model.transcribe(\"\/content\/drive\/MyDrive\/\" + file_name, verbose=True)<\/code><\/pre>\n\n\n\n<pre class=\"wp-block-code\"><code>import pandas as pd\npd.DataFrame(result&#91;\"segments\"])&#91;&#91;\"id\", \"start\", \"end\", \"text\"]]\n\n#segments = result&#91;\"segments\"]\n\n#for data in segments:\n#   print(data&#91;\"text\"])\n<\/code><\/pre>\n\n\n\n<p>\u4ee5\u4e0b\u306e\u3088\u3046\u306b\u51fa\u529b\u3055\u308c\u308c\u3070\u3001\u6210\u529f\u3067\u3059\u3002\u3086\u3063\u304f\u308a\u4e01\u5be7\u306a\u306e\u3067\u3001\u6b63\u78ba\u306b\u8a8d\u8b58\u3067\u304d\u3066\u3044\u307e\u3059\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-full is-resized\"><img decoding=\"async\" sizes=\"(max-width: 746px) 100vw, 746px\" src=\"https:\/\/fukugyouhistory.tokyo\/wp-content\/uploads\/2023\/03\/Untitled0.ipynb-Colaboratory-Brave-2023_03_04-23_57_25-1.png\" alt=\"\" class=\"wp-image-195\" width=\"541\" height=\"178\" srcset=\"https:\/\/fukugyouhistory.tokyo\/wp-content\/uploads\/2023\/03\/Untitled0.ipynb-Colaboratory-Brave-2023_03_04-23_57_25-1.png 746w, https:\/\/fukugyouhistory.tokyo\/wp-content\/uploads\/2023\/03\/Untitled0.ipynb-Colaboratory-Brave-2023_03_04-23_57_25-1-300x99.png 300w\" \/><\/figure>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 class=\"wp-block-heading\">\u5fdc\u7528\u3057\u3066\u307f\u307e\u3057\u3087\u3046<\/h2>\n\n\n\n<h3 class=\"wp-block-heading\">\u52d5\u753b\u304b\u3089\u306e\u6587\u5b57\u8d77\u3053\u3057<\/h3>\n\n\n\n<p>Youtube\u52d5\u753b\u306e\u6587\u5b57\u8d77\u3053\u3057\u3092\u3059\u308b\u3088\u3046\u306b\u3001\u30b3\u30fc\u30c9\u3092\u5909\u3048\u3066\u307f\u307e\u3059\u3002<br>model\u306e\u5024\u3082large\u306b\u5909\u3048\u3066\u307f\u307e\u3059\u3002<\/p>\n\n\n\n<p>yt-dlp\u3067\u3001Youtube\u52d5\u753b\u306e\u97f3\u58f0\u30c7\u30fc\u30bf\u3092\u30c0\u30a6\u30f3\u30ed\u30fc\u30c9\u3057\u3066\u3001input.mp3\u3067\u4fdd\u5b58\u3057\u3066\u3044\u307e\u3059\u3002<br>\u6b8b\u308a\u306f\u3001\u540c\u3058\u624b\u6cd5\u3067\u3059\u3002<\/p>\n\n\n\n<p>\u3072\u308d\u3086\u304d\u3055\u3093\u306e\u52d5\u753b\u3092\u3001\u30b5\u30f3\u30d7\u30eb\u3068\u3057\u3066\u53d6\u308a\u6271\u3044\u307e\u3059\u3002\u65e9\u53e3\u306e\u5834\u5408\u306e\u4f8b\u306b\u306a\u308b\u304b\u3068\u3002<br>https:\/\/youtu.be\/u9eXBAnETVI<\/p>\n\n\n\n<figure class=\"wp-block-embed is-type-video is-provider-youtube wp-block-embed-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"wp-block-embed__wrapper\">\n<iframe loading=\"lazy\" title=\"\u3072\u308d\u3086\u304d\u3055\u3093\u2605\u5207\u308a\u629c\u304d\u52d5\u753b \u52aa\u529b\u3068\u624d\u80fd\u306b\u3064\u3044\u3066\uff5c\u3072\u308d\u3086\u304d\u3055\u3093 The DIGEST\" width=\"500\" height=\"281\" src=\"https:\/\/www.youtube.com\/embed\/u9eXBAnETVI?feature=oembed\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share\" referrerpolicy=\"strict-origin-when-cross-origin\" allowfullscreen><\/iframe>\n<\/div><\/figure>\n\n\n\n<pre class=\"wp-block-code\"><code>! pip install git+https:\/\/github.com\/openai\/whisper.git\nimport whisper\nmodel = whisper.load_model(\"large\")<\/code><\/pre>\n\n\n\n<pre class=\"wp-block-code\"><code>! pip install yt-dlp\n! rm input.mp3 \n! yt-dlp -x --audio-format mp3 https:\/\/youtu.be\/u9eXBAnETVI -o \"input.mp3\"<\/code><\/pre>\n\n\n\n<pre class=\"wp-block-code\"><code>result = model.transcribe(\"input.mp3\")\n#result = model.transcribe(\"input.mp3\", verbose=True)<\/code><\/pre>\n\n\n\n<pre class=\"wp-block-code\"><code>import pandas as pd\npd.DataFrame(result&#91;\"segments\"])&#91;&#91;\"id\", \"start\", \"end\", \"text\"]]\n\n#segments = result&#91;\"segments\"]\n\n#for data in segments:\n#   print(data&#91;\"text\"])\n<\/code><\/pre>\n\n\n\n<p>\u4ee5\u4e0b\u306e\u7d50\u679c\u304c\u51fa\u529b\u3055\u308c\u308c\u3070\u3001\u6210\u529f\u3067\u3059\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-full is-resized\"><img decoding=\"async\" sizes=\"(max-width: 904px) 100vw, 904px\" src=\"https:\/\/fukugyouhistory.tokyo\/wp-content\/uploads\/2023\/03\/Untitled0.ipynb-Colaboratory-Brave-2023_03_05-2_43_40.png\" alt=\"\" class=\"wp-image-236\" width=\"584\" height=\"422\" srcset=\"https:\/\/fukugyouhistory.tokyo\/wp-content\/uploads\/2023\/03\/Untitled0.ipynb-Colaboratory-Brave-2023_03_05-2_43_40.png 904w, https:\/\/fukugyouhistory.tokyo\/wp-content\/uploads\/2023\/03\/Untitled0.ipynb-Colaboratory-Brave-2023_03_05-2_43_40-300x217.png 300w, https:\/\/fukugyouhistory.tokyo\/wp-content\/uploads\/2023\/03\/Untitled0.ipynb-Colaboratory-Brave-2023_03_05-2_43_40-768x556.png 768w\" \/><\/figure>\n\n\n\n<p>\u65e9\u53e3\u3067\u8a71\u3057\u3066\u3044\u308b\u306e\u3067\u3001\u8aa4\u8a8d\u8b58\u304c\u591a\u304f\u306a\u3063\u3066\u3044\u307e\u3059\u306d\u3002<br>model\u306e\u5024\u306f\u3001large\u306b\u3059\u308b\u3068\u3001\u6587\u306e\u533a\u5207\u308a\u304c\u4e0a\u624b\u306b\u306a\u308a\u307e\u3057\u305f\u3002<br>\u4eba\u540d\u306f\u53b3\u3057\u305d\u3046\u3067\u3059\u3002\u3057\u304b\u3057\u3001\u307b\u3068\u3093\u3069\u5408\u3063\u3066\u3044\u308b\u3068\u8a00\u3063\u3066\u826f\u3044\u51fa\u6765\u3067\u3059\u3002<\/p>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h3 class=\"wp-block-heading\">\u5b57\u5e55\u30d5\u30a1\u30a4\u30eb\u3092\u4f5c\u308b<\/h3>\n\n\n\n<p><a rel=\"noreferrer noopener\" href=\"https:\/\/self-development.info\/%E3%80%90python%E3%80%91ai%E9%9F%B3%E5%A3%B0%E8%AA%8D%E8%AD%98whisper%E3%82%92%E4%BD%BF%E3%81%A3%E3%81%9Fsrt%E5%AD%97%E5%B9%95%E3%83%95%E3%82%A1%E3%82%A4%E3%83%AB%E3%81%AE%E8%87%AA%E5%8B%95%E4%BD%9C\/\" target=\"_blank\">\u30b8\u30b3\u30ed\u30b0<\/a>\u3055\u3093\u306e\u30d6\u30ed\u30b0\u3092\u53c2\u8003\u306b\u3057\u307e\u3057\u305f\u3002<br>\u6700\u7d42\u7684\u306b\u3001test.srt\u3068\u8a00\u3046\u5b57\u5e55\u30d5\u30a1\u30a4\u30eb\u304c\u4f5c\u6210\u3059\u308b\u30b3\u30fc\u30c9\u3067\u3059\u3002<\/p>\n\n\n\n<p>\u307e\u305f\u3001\u65e9\u53e3\u306a\u306e\u30671\u3064\u306e\u51fa\u529b\u6587\u7ae0\u304c\u9577\u304f\u306a\u3063\u3066\u3057\u307e\u3046\u5bfe\u7b56\u3068\u3057\u3066\u3001<br>whisper\u3092\u6539\u826f\u3057\u30661\u3064\u306e\u51fa\u529b\u6587\u7ae0\u306e\u9577\u3055\u3092\u5909\u66f4\u3067\u304d\u308b\u30bd\u30fc\u30b9\u3092\u898b\u3064\u3051\u307e\u3057\u305f\u3002<br>https:\/\/github.com\/nyanta012\/whisper<br>\u30bd\u30fc\u30b9\u4e2d\u306e\u300csegment_length_ratio\u300d\u21d2\u5909\u6570\u300cslr\u300d\u3067\u8abf\u6574\u3057\u3066\u3044\u307e\u3059\u3002<br>\u203b\u3053\u3061\u3089\u306e\u30bd\u30fc\u30b9\u3060\u3068\u300cverbose=True\u300d\u30aa\u30d7\u30b7\u30e7\u30f3\u306f\u30a8\u30e9\u30fc\u304c\u51fa\u305f\u306e\u3067\u3001\u7559\u610f\u3057\u3066\u304f\u3060\u3055\u3044\u3002<\/p>\n\n\n\n<p>\u5b57\u5e55\u30d5\u30a1\u30a4\u30eb\u304c\u81ea\u52d5\u3067\u4f5c\u6210\u3067\u304d\u308c\u3070\u3001\u52d5\u753b\u306e\u30c6\u30ed\u30c3\u30d7\u4f5c\u6210\u306f\u5927\u5e45\u306b\u77ed\u7e2e\u3067\u304d\u307e\u3059\u3002<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>! pip install git+https:\/\/github.com\/nyanta012\/whisper.git\nimport whisper\nmodel = whisper.load_model(\"large\")<\/code><\/pre>\n\n\n\n<pre class=\"wp-block-code\"><code>! pip install yt-dlp\n! rm input.mp3 \n! yt-dlp -x --audio-format mp3 https:\/\/youtu.be\/u9eXBAnETVI -o \"input.mp3\"<\/code><\/pre>\n\n\n\n<pre class=\"wp-block-code\"><code># \u51fa\u529b\u3055\u308c\u308b1\u6587\u306e\u9577\u3055\u306e\u8abf\u6574\uff08\u30c7\u30d5\u30a9\u30eb\u30c8\uff1d1.0\uff09\nslr = 0.3\n# \u51fa\u529b\u3055\u308c\u305f\u5b57\u5e55\u30d5\u30a1\u30a4\u30eb\u306e\u30bf\u30a4\u30e0\u30e9\u30b0\u8abf\u6574\uff08\u79d2\uff09\ntime_lag = 0.0\n\nresult = model.transcribe(\"input.mp3\", segment_length_ratio=float(slr))<\/code><\/pre>\n\n\n\n<pre class=\"wp-block-code\"><code>#import pandas as pd\n#pd.DataFrame(result&#91;\"segments\"])&#91;&#91;\"id\", \"start\", \"end\", \"text\"]]<\/code><\/pre>\n\n\n\n<pre class=\"wp-block-code\"><code>! pip install srt\n\nfrom datetime import timedelta\nfrom srt import Subtitle\nimport srt\n\nsegments = result&#91;\"segments\"]\n\nsubs = &#91;]\n \nfor data in segments:\n    index = data&#91;\"id\"] + 1\n    start = data&#91;\"start\"] + time_lag\n    end = data&#91;\"end\"] + time_lag\n    text = data&#91;\"text\"]\n    sub = Subtitle(index=1, start=timedelta(\n                            seconds=timedelta(seconds=start).seconds,\n                            microseconds=timedelta(seconds=start).microseconds),\n                            end=timedelta(\n                            seconds=timedelta(seconds=end).seconds,\n                            microseconds=timedelta(seconds=end).microseconds),\n                            content=text, proprietary='')\n \n    subs.append(sub)\n\nwith open(\"test.srt\", mode=\"w\", encoding=\"utf-8\") as f:\n    f.write(srt.compose(subs))<\/code><\/pre>\n\n\n\n<pre class=\"wp-block-code\"><code>!cat test.srt<\/code><\/pre>\n\n\n\n<p>\u4ee5\u4e0b\u306e\u7d50\u679c\u306b\u306a\u3063\u3066\u3044\u308c\u3070\u6210\u529f\u3067\u3059\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-full is-resized\"><img decoding=\"async\" sizes=\"(max-width: 709px) 100vw, 709px\" src=\"https:\/\/fukugyouhistory.tokyo\/wp-content\/uploads\/2023\/03\/image-91.png\" alt=\"\" class=\"wp-image-446\" width=\"405\" height=\"268\" srcset=\"https:\/\/fukugyouhistory.tokyo\/wp-content\/uploads\/2023\/03\/image-91.png 709w, https:\/\/fukugyouhistory.tokyo\/wp-content\/uploads\/2023\/03\/image-91-300x198.png 300w\" \/><\/figure>\n\n\n\n<p>segment_length_ratio\u306e\u5024\u306f\u3001\u52d5\u753b\u306b\u3088\u3063\u3066\u9069\u5b9c\u5909\u66f4\u3059\u308b\u3068\u3088\u3044\u3068\u601d\u3044\u307e\u3059\u3002<\/p>\n\n\n\n<p>\u203b\u53c2\u8003\u307e\u3067\u306b\u3001\u6539\u826f\u3057\u305f\u30bd\u30fc\u30b9\u306e\u4e2d\u8eab\u3092\u898b\u3055\u305b\u3066\u9802\u3044\u305f\u3068\u3053\u308d\u3001\u4ee5\u4e0b\u306e\u901a\u308a\u3067\u3057\u305f\u3002<br>\u3000\u3059\u3054\u3044\u3067\u3059\u306d\u3002<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>whisper\/decording.py<\/code><\/pre>\n\n\n\n<pre class=\"wp-block-code\"><code> 99 +    # segment length\n100 +    segment_length_ratio : Optional&#91;float] = 1.0\n101 +<\/code><\/pre>\n\n\n\n<pre class=\"wp-block-code\"><code>403 -        self, tokenizer: Tokenizer, sample_begin: int, max_initial_timestamp_index: Optional&#91;int]\n406 +        self, tokenizer: Tokenizer, sample_begin: int, max_initial_timestamp_index: Optional&#91;int],\n407 +        segment_length_ratio: Optional&#91;float]<\/code><\/pre>\n\n\n\n<pre class=\"wp-block-code\"><code>412 +        self.segment_length_ratio = segment_length_ratio<\/code><\/pre>\n\n\n\n<pre class=\"wp-block-code\"><code>440 -            if timestamp_logprob &gt; max_text_token_logprob:\n445 +            if timestamp_logprob*self.segment_length_ratio &gt; max_text_token_logprob:<\/code><\/pre>\n\n\n\n<pre class=\"wp-block-code\"><code>463 +        self.segment_length_ratio = options.segment_length_ratio\n464 +<\/code><\/pre>\n\n\n\n<pre class=\"wp-block-code\"><code>496 -                ApplyTimestampRules(tokenizer, self.sample_begin, max_initial_timestamp_index)\n503 +                ApplyTimestampRules(tokenizer, self.sample_begin, max_initial_timestamp_index, self.segment_length_ratio)<\/code><\/pre>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<p>\u4eca\u56de\u306f\u3053\u3053\u307e\u3067\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"<p>OpenAI\u304c\u516c\u958b\u3057\u3066\u3044\u308bWhisper\uff08\u97f3\u58f0\u30fb\u52d5\u753b\u306e\u81ea\u52d5\u66f8\u304d\u8d77\u3053\u3057AI\uff09\u3092\u7c21\u5358\u306b\u8a66\u3059\u65b9\u6cd5\u304c\u3042\u308a\u307e\u3057\u305f\u306e\u3067\u3001\u8a18\u9332\u3068\u3057\u3066\u6b8b\u3057\u307e\u3059\u3002 \u526f\u696d\u306e\u300c\u6587\u5b57\u8d77\u3053\u3057\u300d\u300c\u30c6\u30fc\u30d7\u8d77\u3053\u3057\u300d\u300c\u52d5\u753b\u306e\u30c6\u30ed\u30c3\u30d7\u4f5c\u6210\u300d\u306a\u3069\u975e\u5e38\u306b\u6709\u52b9\u306a\u6b66\u5668\u3068\u306a\u308a\u305d\u3046\u306a [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[21],"tags":[],"class_list":{"0":"post-179","1":"post","2":"type-post","3":"status-publish","4":"format-standard","6":"category-ai"},"_links":{"self":[{"href":"https:\/\/fukugyouhistory.tokyo\/index.php?rest_route=\/wp\/v2\/posts\/179","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/fukugyouhistory.tokyo\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/fukugyouhistory.tokyo\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/fukugyouhistory.tokyo\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/fukugyouhistory.tokyo\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=179"}],"version-history":[{"count":67,"href":"https:\/\/fukugyouhistory.tokyo\/index.php?rest_route=\/wp\/v2\/posts\/179\/revisions"}],"predecessor-version":[{"id":850,"href":"https:\/\/fukugyouhistory.tokyo\/index.php?rest_route=\/wp\/v2\/posts\/179\/revisions\/850"}],"wp:attachment":[{"href":"https:\/\/fukugyouhistory.tokyo\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=179"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/fukugyouhistory.tokyo\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=179"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/fukugyouhistory.tokyo\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=179"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}