Speaker 1: Let's say you have a bulk number of audios in mp3 file and you want to generate their transcript, their subtitle files, then I believe this script will be helpful for you. In this script, there is a file in variable, in it, it contains where you specify your audio files, where you want to get your output for the plain transcript and it is for search functionality. For example, if you want to search these words in all audios, then you can specify these and which mode this script should be run. Do you want simple transcript of audio? Do you want to also perform search on the search or and do you want to generate subtitle only? Let's say I will start from the subtitles, I make the subtitle true and the input directory is input, so in my input directory, you can see there is two audios. Now I will run this and there is a different type of models you can specify, all of you may be aware of Whisper which is recently released by the OpenAI, the same company that released the GPT. So just you run python.main.py, now it will generate subtitle because we are only in subtitle mode, it will only generate subtitles, SRT files. You have to wait, it will take some time for generating these subtitles. Now it is 0%, you can see the folder is created here, this folder contains the subtitles files, different type of subtitles. Here it goes, the video you can see is approximately 3 minutes, it is a famous poet, Khalil Gibran famous poet, now you can see this is running and here you will get your subtitles. After when the script will finish running, you will have your subtitles here, for example, this is the first file, now it is running, generating the subtitles for second file, that is second mp3. It has already generated subtitles for this file and here it is done. Now you can see two folders are created with the same name of audio and it has a transcript in different formats and also has subtitle files. For example, this file, this has a subtitle, you can see. Now if you also, these subtitles, these are the SRT files, but if you want word level subtitles that are also generated here, word by word, you can see here. Now this is a subtitle and there are other files that you can use as well and both of the audios have generated the same subtitle files for both of the audio. Now if we, for example, you just want to, want a plain script, then what will you do? Just go, what will you do, just go to here and for example, you don't need subtitle for now, set it false, you want only transcript and set it to true. Now you will run the script and what it will do, it will do transcription, it will transcribe the audio, same as before, but it will, this time, it will generate the output in output folder. You have to wait for some minute, the audio is approximately 3 minutes, so it will take a little bit time, till then I will tell you about the, how such works. If I set, set to true and all things to false, what it will do, it will cite these words, you can specify n number of words, for example, word 11, word 12, which is equal to I am here, for example, here the first transcript has been generated, now it is generating the second transcript, second transcript. Now if you want, just want to search these words in all of the audio, for example, will this audio contain this word? If you want to search that, you can set cite to true, now for example, I have set cite to true, and I will, I want to, I have set trusting to false. Now if I run the script, it will generate the file in answer folder, the time stamp when the file was generated, now if you open the file, you will get a cv file, it will tell you, this word was not found, this word was not found, not found, it was found, and in which file, mp3 pxt file, same, same here, it found in, this word is, my word is found where my was written, and these, how many times these words are found, for example, my is this many times found in this word, this audio, and same goes for here, and I am here is not found, so this all for you, if you want to me, to give access to, to contact me or subscribe and leave the comment, thank you.