Nikolas Garske

  • My journey into embedded programming

    My journey into embedded programming

    Now something completely different: Embedded! For some time now, we have been the owners of a small solar power plant, known in Germany as a „Balkonkraftwerk“. It is a limited version (only 800W) of its bigger brother, which means less paperwork when it comes to installation. The power output is measured with a Shelly. The…

  • Evaluating the Language Consistency of Local and Cloud-Based LLMs

    Evaluating the Language Consistency of Local and Cloud-Based LLMs

    One year ago, in my first article about OpenAI Whisper I made a prediction: I still think that one day we will have LLMs in the capability range between GPT-3 and GPT-3.5 (maybe up to GPT-4) – on our mobile devices. Now, let’s see how good the sentence has aged and how far we have…

  • Tokens, RAG und das Wissen von LLMs

    Tokens, RAG und das Wissen von LLMs

    Ich möchte diesen Blogartikel mit einer lustigen Begebenheit aus den vergangenen Tagen beginnen. Der Chevrolet Autohändler in Watsonville hat versucht den Kundenservice durch eine künstliche Intelligenz zu verbessern. Eingesetzt wurde das GPT-4, eingerahmt in einem klassischen Chatfenster, wie es viele Webseiten heutzutage verwenden. Aber wie sich herausstellte, konnten einige Nutzer den Chatbot austricksen. So wurde…

  • Zephyr, Mistral and LLMs in general

    Zephyr, Mistral and LLMs in general

    Exploring the role of foundation models and the emergence of smaller LLMs like Mistral 7B There was big news in the world of LLMs at the beginning of the month: Mistral 7B. This new LLM was unique for its incredible performance and for its relatively small size. This LLM does not have to hide behind…

  • Making OpenAI Whisper better

    Making OpenAI Whisper better

    We already looked at ways to make the original OpenAI Whisper model faster. We came across two different projects that aimed to deliver the best performance in terms of transcription speeds. But Automatic Speech Recognition is actually not about speed. The primary focus is on factors like quality and functionality. The overall quality of Whisper…

  • Making OpenAI Whisper faster

    Making OpenAI Whisper faster

    Explore faster models of Whisper with reduced transcription times, lower memory consumption, and use of TPUs.

  • Evaluating the Expense of OpenAI Whisper: API or Self-Hosted?

    Evaluating the Expense of OpenAI Whisper: API or Self-Hosted?

    OpenAI’s Whisper is one of the best Speech-to-Text models currently available on the market. The transcriptions are highly accurate and include information such as pauses and emphases. It outperforms in terms of quality major competitors. It can also process a variety of languages. In addition to automatic transcription, it is able to perform language translation…

  • Automating Stock Market Updates with Python

    Automating Stock Market Updates with Python

    It can be complex to keep track with all the developments in the stock market. Luckily, as a developer, we can help ourselves with automation. However, sometimes complexity can make this a daunting task. This does not need to be the case all the time. In this blog post, I explain how to write a…

  • How to use Rust and MQTT in your Next Project

    How to use Rust and MQTT in your Next Project

    Why do you want to buy an IoT device? In most of the cases, you want to obtain more control over your home. Maybe you just want to monitor things or even control them remotely. Of course, you can rely on the supplied apps of the manufacturers. But as a programmer, this is not your…

  • Tauri’s use of JavaScript and Rust

    Tauri’s use of JavaScript and Rust

    Tauri is a very interesting framework. On the one hand, it is a framework solving a huge problem in the world of desktop development. It enables cross-platform development without the disadvantages of shipping a separate Chrome browser each single time. This saves resources and reduces the bundle size. On the other hand, it is one…