Speaker 1: Transcribing can be pretty important part of video creation. It has benefits of accessibility, search engine optimization and monetization, improved content delivery, compliance and many more things. It is a win-win situation for both creator and for the viewer. However, captioning videos is not so easy. Or let's say creating accurate captions is not easy. For example, on YouTube, you can easily create captions and even make translations. However, you can spot many mistakes. So, in this video, we will be looking at a very powerful alternative, It is a neural network trained by the OpenAI and it can accurately transcribe even the very difficult examples. Check this out. As you can see, it is pretty amazing and we will be using Whisper for transcription. However, we cannot use it as is. We need a middle layer to correlate Whisper to YouTube. Luckily, it is already done and it is called YT Whisper. It essentially combines YouTube DL, FFmpeg and Whisper together. So, further ado, hit the like button and let's get started. Now, first, we head to the project page, which you can find its link in the description down below. Once we were there, under the installation section, we can see the project requirements, which are Python and FFmpeg. It is not mentioned, but we also need GITs because this is how we fetch the project. Installing those are pretty straightforward, so I skip over them. However, just to verify them, we can open up PowerShell and type git //version python //version ffmpeg We should see outputs like those and we are all good to go. At this point, we are ready to install YT Whisper. To do so, we go back to the project page and copy this command. Then we head back to the PowerShell, paste it and run the command. It should start installing the YT Whisper. Once that's done, we can again verify the installation by typing YT Whisper –h. It should return output like this and that's basically it for the installation. We are now ready to transcribe our videos. Doing so is pretty straightforward. We first type YT Whisper, then follow that by the video URLs that we want to transcribe. In this example, I put two URLs separated by a comma. Now, this can be enough and if I hit enter, as we can see, it started. And while it's going, let's quickly mention the other options. First, we have the model. It is the option that we can specify which model to use. By default, we are using the small model. However, we can change it to one of those as well. Using a bigger model can provide better results. So before bulk transcribing, I suggest you to try out a few different models in order to find which suits best for your case. This is a very useful option, but we have more. Assume that we want to go from French to English in the captions. Normally, this requires additional translation on top of transcription. Luckily, we can do all at once. All we have to do is add –task and translate. It should be good to go. Anyways, as we can see, our transcription job is now complete. The program has returned us two VTT files which contain our captions. Now, a quick note. The output location of these files are dependent on the directory of your terminal. In simple terms, if I were to create a folder named yt-whisper, head into it in my terminal and execute yt-whisper there, then the output files should be inside that. However, since my initial directory was desktop, it created those files on desktop. And speaking of it, what are those VTT files? Well, if I open one of them, we can see that they include transcription with the timestamps. So this means that we can use these files directly to add captions on the YouTube video or we can use it on our editing software. Either case, it's very useful. And that said, this is essentially how you can transcribe your videos using OpenAI's Whisper. As always, I hope you've enjoyed it and find it somewhat useful. If so, make sure to hit the like button and subscribe. And until next time, take care.