<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Whisper on Datengraben</title>
    <link>https://datengraben.com/tags/whisper/</link>
    <description>Recent content in Whisper on Datengraben</description>
    <generator>Hugoo -- gohugo.io</generator>
    <language>de-DE</language><atom:link href="https://datengraben.com/tags/whisper/rss.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Whisper.cpp default weights transcribes german audio</title>
      <link>https://datengraben.com/til/llm/whisper-cpp-lang-de/</link>
      <pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate>
      
      
      <guid>https://datengraben.com/til/llm/whisper-cpp-lang-de/</guid>
      <description>&lt;p&gt;Today I learned that using &lt;a href=&#34;https://github.com/ggerganov/whisper.cpp&#34;&gt;whisper.cpp&lt;/a&gt; and your CPU, you can transcribe german audio via &lt;code&gt;-l de&lt;/code&gt;, using the &lt;a href=&#34;https://huggingface.co/ggerganov/whisper.cpp/tree/main&#34;&gt;standard model weights&lt;/a&gt;, available via huggingface.
For location in the manual documentation &lt;a href=&#34;https://github.com/ggerganov/whisper.cpp/blob/b87494bb8f1e2b5843ec606294e8c370aa25a368/README.md?plain=1#L139&#34;&gt;see&lt;/a&gt;.&lt;/p&gt;
&lt;p&gt;Suppose your repo is at &lt;code&gt;/path/to/repo/whsiper.cpp&lt;/code&gt;, then full command to execute in the root folder of whisper.cpp repo is: &lt;code&gt;./main -m /path/to/repo/whisper.cpp/models/ggml-medium.bin -l de -t 12 -pc -oj -f samples/your-audio.wav&lt;/code&gt;&lt;/p&gt;</description>
    </item>
    
  </channel>
</rss>
