TTS - Voice Cloning, GPT-Sovits, OpenVoice, DeepFake, Ghibli Style Image and many Python projects!

1. Introduction Real Time Voice Cloning: Greetings everyone, I am chosen for them. And today I will share how you can install real-time voice morning on a Windows PC. This Python program can convert text to audio using a library or even your own voice. We will be installing this report from GitHub on our Windows ten computer. The only objective of this course is to make the real-time voice morning work on your PC, which is a Python program. What will you learn in this course? In this course, you will get a basic idea of installing Python, like creating the working environment, installing various modules, Sandra. And you will also learn to install and use the real-time voice cloning application. What's included, creating a working environment, installing real-time voice cloning, and using the real-time voice cloning application, we will use the audio library as well on our own voice. What are the requirements for insulation? For distillation? You will require a Windows computer and a high-speed Internet connection. No coding skills are required. Everything has been properly explained so that you can install and run the voice floating application. At the end of this course, you will be able to install and use real time voice cloning on your Windows computer. In addition to this, you will gain a brief knowledge about the Python and know how you can resolve various uterus while installing the application. Let's get started. 2. Real Time Voice Cloning: Greetings everyone. In today's video, we will quickly learn how you can install real-time voice cloning on a Windows computer. We will be using this repo from GitHub and we'll be installing this Python application, something like this. Let's get started now to install real-time voice cloning. Open the browser and search for real-time voice cloning. This is the one. Click on it. So basically, we will be installing these to our Windows ten PC. All the information of the application can be found here in this space. Anyway. Here you can find all the installation requirements and we will do it from the basics, targeting those who have absolutely no idea about how this can be installed. Please note, we will have to install various software on the PC and we will be doing it for Windows ten computer. Anyway. Now to get started, search for Anaconda. This will create their working environment. This is the one. Click on it. You can directly download it from here. Or you can also look for additional installers. Click on it. Since this PC runs at 64 beat, I will download this if your PC runs on 32-bit installed, this one, you can see the size is about 594 MB. Let me quickly download it. So I can close this. While this is downloading here. Python 3.7 is recommended, but we will be installing Python 3.9. As I have worked with it. It works. All great. So the download has been completed. Show in folder. Here is the file. Let's install it. Double-click on it. Click Next. I agree. Choose where they just for you or for all users, I will choose just me. Next, choose the folder. I will leave it to default. So we'll install Python 3.9 and install. The installation took about five-minutes. Click on Next. Next. Don't need to open this, so we'll undertake it and finish. So we have successfully installed Anaconda. Let me close this all. Now let's install Python. On this top menu. You can see Anaconda Prompt here. If not, simply type Anaconda prompt here, right-click and run as administrator. So this will open. Please note this is not a regular command prompt. Now the first thing we need to do is to create a working environment for that. Simply type conda. Create double dash, name, give any name you want. I will write voice clone. Space. Python is equal to 3.9. Don't worry, I will provide this code and press Enter. Yes, and then top. So now we have created a working environment with Python 3.9. Now let's activate voice clones. Simply copy this from here. Control plus C, and then Control V. Please copy this and save it. Please note this base name would change to voice globe and press. So you can see this has changed. Now after Python installation, it says to install FFM PZ. If I click on this link, here are all the details, but this might be a little confusing. So instead of following this, I suggest you to go to this site. And I kinda.org slash Conda, desk forage slash FFM BZ. Now on this website, you can find the codes we can use. So let's install. Ff and BZ. This is what we want. So let me copy this, Control C, open this, make sure this is the working environment and not based, but voice cloned based control plus V and press Enter. So it's installed. Let me close this. So we have successfully installed FFM, be easy. Now let's install dark open this. The previous version worked great for me. So click here. Now, depending upon your PC and graphics card, use what works for you. If your PC uses never die graphics and want to use it, you can install cuda 10.2 or 11.31. I only want to use the CPU. Let me copy this code, control plus C and paste it here, control plus V, and then tap into. So we have installed dark. Let me close this. So now we have install FFM peasy BY dirge. Now let's install the requirements for that. We need to download this click on code and download the zip file. So it's done, showing the folder. Double-click to open it and extract it to detect stop. Here, is it this too? Since it has a long name, let me rename it to voice. The changing folder's name is not required. Anyway. Again, go back here. So now we need to install the remaining files using this command. Now we need to change this location where the facilities. So if we write cd space thick stop. It will go to the text file folder. But the folder does not exist. It **** stuff usually is inside Windows. So type cd, double-dot and enter. So now we are on C Windows. Now in most BC dict sub is inside Windows. So typing CD specific stock should take you to the deck stuff folder. Please use CD or CD double-dot to change the folders. The text up could be inside the Windows, uses a spectra if you have no idea where the exact folder is like this BC, and simply open the folder and click here to see the exact location. So let me go out of the windows with cd double dot. Then CD uses CD. Cd, ****, stop. Sorry. Cd OneDrive and CD deck star. And finally to the voice folder. So we're finally to the voice folder. Please make sure this link is correct to where your folder is. You can use CD or cd double dot as accordingly, disconfirming this link. And this link is the same. Anyway, after coming to this location. Now we need to run this command, copy control plus C, control plus V. And then you can see they're quite some years in these areas may differ from PC to PC. So let's investigate the era for building will far we've Odyssey vet, and the other eateries each acquires Microsoft Visual C plus plus to be more than 14. Let's first resolve this issue with Microsoft Visual C plus plus. Simply Google it. Click on this link. And here's the offline link. Click on it. So did download is done showing folder and double-click to install. Continue. Now we can choose the component we want. Go to individual components and sets for Windows ten SDK. So I will install this version and click on Install. Please note, this will take some time. So this has and installed. Let me close this and close this two. Now, we can test if it is installed properly or not. Copy this Control plus C and paste it here. So you can see the model liberal side is not found. So let's install it for that type. Conda. Install dass sie, Conda, forge space lib Rosa, and press Enter. Basically, we are testing if everything is all good or not. If there are any areas or missing modules, we will need to resolve this. We will have to install all the missing modules as required. Yes, and enter. Libra psi is installed. Again, let's run the demo command and see if anything is missing or not. So again, you can see that we have a missing module, unique code. Let's install it. Type PIP, install you in IDE code, and enter. So it's installed. Let's again run the demo. Again. You can see inflict model is missing. Let's install it. Pip, install, inflict an intact. It's installed. Let's try again to run the day rule. Now, again, we see two here is one, year is unable to load this one web RTC VAT. And the other eater is failure to download synthesizer dot P D synthesized adopt PET actually is downloaded from Google Drive. And if many users have downloaded eat, it might be temporarily on every level. Anyway, let's first fix the web RTC vet for that type PIP, install Web RTC VAD and press Enter. So it has failed to install Web RTC vet. Let's try to install Web RTC vet desk wheels, PIP install Web RTC VAT wheels. This one is successful in stock. Now let's try to run the demo. So this one is resolved. So one issue is that synthesized or cannot be downloaded. Normally, it is auto download it. If it is not downloaded, let's see a weekend fix it. Let's download it manually. So here you can see pre-trained models. Simply click here. And awkward into this, it tests fail to download, synthesize a dot p d. So let's download this synthesizer. Let me click here. So the reason why it cannot be downloaded is that the limit was passed as many users downloaded it. Anyway, this may not happen in your case. However, if it happens, then I will provide an alternative download link to four. Now I have the backup of the file. Let me quickly insert the pen drive. This is the one. Let me copy this. Again, a reminder, if the Google link works, you possibly want half an issue, but if it fails, then download it from other links and keep it manually. Like we're going to do, open the voice folder, open save on the skull models folder, default. These two have been downloaded. Let me paste it here. So here's the synthesizer. It be close this. Now again, let's run the demo. Let me copy this. And paste it here and enter. So it's basically saying that all tastes best. You can now synthesized speech. Now let's run the program. I will close this so that it's easier for you to navigate later, as well as the program's needs to be run like this every time. Anyway. Now, here's the audio libraries speech. You can download it and use it to or use your own voice or other sample voices, which I will show you later. For now, let's open Anaconda Prompt. Now, whenever you open Anaconda, please note this goes back to this space. You will need to change it. For that, we need to activate it. If you have saved it earlier, you can copy paste and activated. But if you're unsure how to activate, then go to Start Menu. Type Anaconda on Anaconda Navigator, right-click and open file location. Again, right-click and open the file location. Scroll up and you will see ENV folder. Open this. So here you can see you environments name. I will simply copy the name by renaming and control plus C here, right? Conda, activate folder name, and then times. Now the best name he changed. Let me close this. Now again, we need to get back to this folder as we did it earlier. So we are in this folder. So earlier we successfully passed the test configuration. Now let's run this one. Copy control C and paste control plus V. And then again, another model is missing. This SAN, device model is missing. Let's install it. Bab installed some device into. So that's done. Now, again, tried to run this. Again. You map model is missing. Let's install it. Pip, install you map, and then it's done. Again, run the program. Again, you wake, you, D5 is missing. Let's install it. Pip installed BY Q D5 and enter. It's done. So again, let's run the demo. So now we have successfully run the program. Now, if you see here the data is missing, it cannot be clicked. Now we can either browse the voice here and choose an audio file and use it for our own text. Or you can download these data sets to use it in your project. Please note the size is nearly six CV. For this video, I've actually downloaded it. It's on my pen drive. So this is the whole data set. You will receive it as a zip file. I only unzip this. So to use it, copy to our folder. Let's close this. On the voice folder. Simply paste it. It's nearly six CB Insights. Please note, when you deleted download from here, you will receive a zip file and you will simply have to extract it. So this has been copied. Let me close this to open with the data set. So if I open voice, you can see we have these livers folder. If I open this, you can see there are lots of voices. And you can use this one on the toolbox. Now to display all these voices on our tool, click on Anaconda prompt. Please note that to launch the tool box, every time you will use this command. This command will launch the toolbox without the data set. Use the data set we added earlier. We can use this. Here we need to specify the location of the audio files. I'll copy this to the North path to make it clear. So this and the data says roots is basically the folder location without livery speech folder. Simply copy this address space and paste it. So here's the full command. Now copied control plus C and paste it Control plus v and Enter. So now you can see here all the data sets, speakers and choose accordingly. And here you can write custom message is select the speaker and load them. Once it's loaded, click on synthesize invoke code. It will take some time. Hello, how are you doing? So that's the voice. You can change this utterance and try various vocals and use accordingly. So this is using the library. Now, if you want to use their own voice, let me close this. You can either add it to this livery speech folder. Like here. You can create another folder and call it maybe 18. And make sure you follow the exact patterns like this. Fuller's 19 and add subfolders with numbers and follow the same number format, like 191980000 and so on. I've already created a sample vocal five to 105 Recording. So here's 18 and sub folders, and I have numbered it accordingly. So this is my own vocal. So let me simply copy this to the library. Delete this one, and paste it here. Here are 18 and other voices on the same format. On the integral. No problem. Let me run the command with the data set. So here is the speak up. You will find the vocals you have added is number 18. And here are the audience weekend Jews and load this plate. And then you can synthesize and vocal. However, it seems like the voice like mine has not been cloned accurately, maybe because of the accent. Since I may not have the ingress accident, hope in the future, it may get more advanced. But for the native speakers, this might work very well and close, very close to your vocal. So please do give it a try anyway. So this is how you can install this toolbar and clone voices. There is other simpler option to use your or any vocal. Let me close this. You do not need to add the vocals here, nor you will have to download this whole library. You can simply run the program without the data site. Let me use this one, copy and paste. This will run without the data set. You can see there's no data safe, and it's not here. You can simply browse and choose a sample voice, which could be five to ten seconds. It takes hips, MP3, WAV and more. It might take some time. You might also see not responding. It depends on various factors. So it seems to have an issue again. You can see some warnings there. Let me try again. Still getting this era. If I check the Anaconda Prompt, there are some more eras. Eras are basically because of the audio I chose. Let me change the audio and try with another sample. Let's browse to new sample audio. So this seem to work. If I play this, this is how you can make the necessary. So this is my vocals. Let me try with some texts. Synthesize and vocal. How are you? So that's the vocals. As I mentioned earlier, the cloning depends on your vocal type, accent and many other factors, but please do give it a try and see if it works for you. I really hope this video is useful. Thank you very much. 3. GPT SoVITS Clone Your Voice: Hi, guys. Here is how you can clone your voice using GPT Soviet. Here is how you can password protect your PDO. So basically, we will be installing this GPT Soviets to PC. With only 1 minute of your voice data, you can train a good TTS model to clone your voice. So let's install this. It's super easy for Windows users. No need for detailed commands. Simply scroll down, and you will find Windows here. Click on Download the intrigtd package. The file size is about six to seven GV, so it will take some time. Simply download this, and we will be using this on Windows 11. Anyway, we have already downloaded this. So here is a zip file, open it, and let me extract it to the desk stay. So this is done. By the way, to extract this, you might need seven Z. If you do not have it, simply Google seven G. This is the one you can simply download and install on your PC, and you should be able to extract it. Anyway, so here is the ZPT soviets folder. You can delete the zip file if you want. Let me open this. And in here, we need to open this go web dot bat file. Simply double click on it, and this will open in a browser, but this will take some time. So this is the interface, but all this is in Chinese. Let me quickly change this to English. Let me close this. Let me close this too. Right click on the go WebUI and edit it notepad and simply remove this Zach Esco CN. Everything remains the same and save it. Now again, open the web UI. So now everything is in English. Let me close this skitter paz, as well. So this is GPT Soviets interface. Now we want to clone our own voice. So here is our sample voice. Make sure your voice is at least 1 minute. But for this tutorial, I have made this a little shorter, but make sure you have at least 1 minute of your recordings. Let me p this audio inside a folder. Let me put it in the new folder. Now copy this folder address. After copying it in here, we need to add our folder link. So paste it here. You do not need to change anything else, let everything else be on default, and then click on speed slicing. You can see it is processing here. You can also open CMD to see the progress. Since I am using CPU, it is taking much time. So this is now working, and you can see it on the CMD too. So it is now finished. You can also see this in CMD. After this is done, go back to the Soviets folder. Go to output and on Slicer underscore OPT, you will find all the voices. Please note it is better to use at least a minute audio. I've only used a small audio. Now after this, go to denoising tool. It will remove the noise from the audio. You will see the folder path here. It is auto selected, but if it is missing or empty, then simply copy this address. There, now click on open speech denoising. You can see the progress on CMD. So you can see it is done. You can also see this denoising finished. After this, go to a speech recognition tool. Make sure the folder path is correct. It should automatically be here. If not, make sure to copy this denoise path folder. Anyway, this is already here, everything set to default and click on open speech recognition. This will take some time. You can see the progress on the CMD. So this is done too. You can see it is finished here, too. Now it will automatically add the dot List pile here. If you cannot see the link here, Go to the output folder and go to the ASR underscore OPT folder. You will have to add this location along with the file name. You can see it is the location of the file name. After this, click on open audio leveling wbi this will list all the audio and corresponding text, and you can play it. Hi, guys. Here is how you can password protect your PDF file online for free. The text and the audio is not correct. Simply correct all the text versus audio. If there are any mistakes, let me quickly correct all of them. This is very important. So we have made all the changes. Now click on Summit text and save file. So this is done. Close this. Click here on close Audio Libling if it is not closed earlier. Now, let's recheck if our changes have been reflected or not. Click on Open Audio Living. So here all the changes are made, and it is working all fine. Let me close this, and you can close it here, too. After this is done, click on GPT SOVDTTs. Now give any model name you want for your audio. I will write my custom voice. This will auto use ZPU if you have any. This PC does not have it, so I will leave this now on the dataset formatting tool at the denoispt dot least path. If this is not added here, make sure to copy this file path, copy this path. And paste it here and make sure to remove the inverted commas. So this is all done. Leave everything to default and then click on Open training set one click formatting. And here is the progress, as you can see. You can leave everything to default. By the way, here is the audio dataset folder. If this is empty, use the denoise folder, and here is the least file we added. This takes quite a long time. You can always track progress on CMD. So it has been done. After this, click on One B Fine Tuning. You do not need to change anything. Now let's fine tune our model. Click on Open COVID Training. So it is opened, and you can see the progress in CMD two. So the training is finally done. You can also see it here, so it's training finished. Now after this, click on Open GPT training. It is opening, and this will take quite some time. I'm just fast forwarding it, so this is also done, and you can see it is all finished. After this, scroll up and click on this one inference and click on refreshing model paths. In here, you can find your own custom voice. So choose the one with the highest value. E 15, I will choose this one. Likewise, in the weit list, choose the one with the highest value. So 800, and this is done. Then take this enable parallel inference version and click on Open TTS inference webi. If you're using a CPU, and opening this will take some time. So here you can see my custom voice and also in the Witist. You can upload a reference audio here, too. I will use the one in the DnoisOPtfolder, choose any audio you want. Hi, guys. Here is how you can password protect your PDF file online for free. So here you can write the text from our reference audio. You can type it or you can go back to the main GPT so Bits folder, right click, and it with Notepad. So here you can see the text. Let me copy this. And pay sit here. So this text is reference for the audio only. You can also add the second audio for reference, but I will skip the spot. Now, in here, write the text. You want to convert to audio using your voice. So the text is done, it will be converted to the audio. And here, choose the language that's English. Also choose the language here, English. You can leave these setups to default or change it if you want to experiment. You can also use the snow slice or with English punctuations. You can also use a segmentation, but I will skip this. Anyway, now click on Start Inference. Again, this will take some time. You can check the CMD for progress. Now simply wait till you see the audio. Just a quick reminder, this whole process of initial training and cloning of your voice takes quite a long time, especially if you're using a CPU. By a lot of time, we mean two to 3 hours depending on your processor. If you're using ZPU then it will be much faster. The text to audio, however, is not that slow, but again, it depends on the length of your text. So it is all done and you can play to listen. Hi, everyone. Please subscribe to our channel. Thank you. So this is my clone voice. Not as perfect, maybe because of a limited sample audio. And as well, I'm not a native English speaker, and I'm pretty sure it should sound good for our native speakers. Anyway, you can also experiment further with these settings and find the best match for your audio. We really hope this video was useful to you. Thank you very much. 4. OpenVoice - Clone Your Voice & Text to Speech: Greetings, everyone in this video, we quickly learn how you can clone your voice in seconds and as well convert text to speech using open voice. Basically, we will be installing this open voice to our Windows PC. We will be using Anaconda to run this application. So for that, go to anaconda.com and click on Free Download. You can give your email or skip registration. I will skip the registration and download the installer and install it to your computer. After Anaconda is installed, if you go to the Start menu and type Anaconda, you will find Anaconda Prompt. Let me open it. Now, the first thing we need to do is create a Python environment. For that, type Conda, create doubles name. I will use open voice. Python is equal to 3.10. We have tested with Python 3.10. That's why we will be using this one. This will create a Python environment, open voice with Python 3.10. Wile Enter. So the Python environment has been created. Now let's activate the environment. For that type Conda activate open voice. Make sure this buss name has changed to open voice. Now again, go back to the guitar page. And let's download all these files, go to code, and download a zip. I have actually already downloaded it. Anyway, showing folder, I would delete this one as I've already downloaded it earlier. So here is the open voice C file, open this. And let me extract it to the desktop. Let me close this. So here is a folder, open it. Now, we need to copy this address, copy this Go back to Anaconda Prompt type CD space Control plus V. So now we are inside this folder. Now let's install all the dependencies required to run Open Voice. For that, type PIP install. That's requirements dot TXT and Enter. This might take some time as it will download lots of files. You will be stuck at this screen for a few minutes, so please be a little patient. I am fast forwarding it. So this is all done. We also need to install FFMPZ to run open voice. So for that type Conda install FFMPZ and Enter. Y and Enter. So FFMPZ has also been installed, so this is all done. Now to run these files, we also required to install Jupiter Nodebook. For that type Kanda, install C, Conda Forge, Jupiter Lab and Enter. So there was a little typo with an extra T, so let me retype it again. It froze. Let me try to re run the command. And now this seems to work. Why and Enter. So this is done. Now, we also need to install some Widget for Jupiter Notebook. For that type PIP install notebook, IPI wicket. So now everything has been installed. Now again, go back to Guitar pays. Scroll down, and you will find this uses link. Click on it. And we need to download this V one and Vt checkpoints. Click on V one. For Demo Part one and Part two, we need this checkpoint. Simply click here and download it. I have already downloaded it. The file size is about 400 MB. Likewise, also download this Vito checkpoint. The file size is about 116 B, and I've already downloaded this one too. I will cancel this. Let me go to the downloads folder. Let me open this view one Checkpoint and extract it here. So it's here, likewise, let me open V to Checkpoint and extract it here. So it's here. Now let's move this to Open Voice folder. Let me close this Open this folder and paste it here. So we have the checkpoints here. Now go back to the anaconda prompt. Let me clear this CLS. Now, let's open the Jupiter notebook, type Jupiter Notebook and Enter. So this is the Jupiter Notebook. Let me close this all. Let's open this demo part one. Now, let's run this. Highlight this and click on play. This asterisk means it is working. So this is done. Now click here and run this again. You might see this, but it will work anyway. I will play all this. For now, we are using the reference AP three. Play all the So this is done, and now this is the text we want to transcribe. You can add your own text here. And if I go to open Voice folder and go to resources, this is the example reference voice. Let me quickly play it so that you can hear it. When I was a wanted man, the resistance gave me a lot of help. So this is the voice we are using for now. So let's try to see how this works. So let's run this one. So this is done. So the output name is output in default. In this folder, if you go to output, let me rephras this. So here is the output in default. Let me play this. Please subscribe to our channel. This is a voice, and it is working all good. Now in here, we can also add different styles to the voice, like friendly, cheerful, excited, sad, et cetera. For now, let's try cheerful. I will replace whispering. I will use the same earlier text. And let's run this again. So this is done. You can also change the speed here and let me repress this. So here is the audio. Let me play it. Please subscribe to our channel. So you can hear the difference. Likewise, this will use the Chinese language. Now let's try with our own voice. So here is our sample audio voice. Hi, guys, here is how you can password protect your PDF file. Here is just more than a minute audio. Let me cut this and paste it inside the resource folder. And here is my audio. And in here, replace with your audio name. It is audio dot mp three. Make sure the name and extension is correct. Let me run this. So it is all done. Now let's try with our own voice. Let me change the output name, output and default one, and run it. So this is done. So I'm in the output folder. Let me repress it. So here is it. Let me play it. Please subscribe to our channel. So that's the clone of my voice. It's definitely not so great, possibly because of my accent, but you can try with yours. And I can also add emotions to it. Let's try whispering. And rename the output to whispering one and run this. By the way, you can add as much text as you want. And let me repress this. So here is it. And let me play the audio. Please subscribe to our channel. So that's how it sounds. So this part is done. So this is using Demo part one. You can also use Demo part two and Demo part three likewise. Let me close Demo part one. Here is Demo part two and Demo path three. Let me open Demo Part two. For the Demo part two, it is suggested to add the Open AI API key. For now, I'll skip this. But if you want, you can get the APIkey from Open AI website. You might get free credit if you are using it for the first time, but eventually it is a paid service for API. Anyway, I will skip this and instead do Demo path three. Now, in order to use Demo path three, we need to install some more applications. Let me go back to Anaconda Prompt. Press Control plus C. This will stop the Jupiter notebook. I would clay this as well. Again, let's open the open voice guitar pase. This is the one and go to uses. And click on Version two. You will need to install Milo TTS to use Version two, copy this code and paste it to naconaPmpt. Press anyway, and it will install all the required files. And Demo Part three is a little advanced than Demo Part one. So please do give it a try. Again, presenter. It will download some files, so it will take some time. So all the downloads have been finished. Now we need to install furthermore dependencies. Without them, we had encountered some issues earlier. So we have this download NLTK file. This will be provided. This will basically import NLTK and all required files. Without installing this, we could not run Demoga three. So let's install this. I will put this inside the OpenoiceFolder. Let me copy the name. And now on Anaconda prompt, type Python trop w dot py and Enter. These steps are only required for Demopat three. If you are satisfied with Demo path one, you do not need to install all this. Anyway, click on Download. So the downloads have finished. You can close this. So we have installed everything. Now again, let's run Jupiter Notebook, type Jupiter Notepad. Sorry, there is an error. It's actually Jupiter Notebook. Let me close the previous ones. Now, let's run the Demopath three. Let's run this first. If it is running, you will notice this asterik. So this is done. Now let's run this as well. This will use the hick Point Version two. Earlier, we were using Version one. This is done too. So this is all done. In reference audio, use your own audio. So this is all done. So this is using the M TTS, the one we installed earlier. So this can use the actions of all these languages. Only use the ones you require. It can do it in multiple accent, but will take some time. Let me run this. By the way, the output folder had changed for this one, its output version two. This is working. Is the output version two folder. You will find all the outputs here. And this will take some time. It will download the checkpoints as it requires for languages, and it will take some time. Let me quickly check the output folder. So here are the outputs on various tones. Let's listen to it. Did you ever hear a folk tale about a giant turtle? This is with Indian accent. Did you ever hear a folk tale about a giant turtle? British accent. Did you ever hear a folk tale about a giant turtle? Now, let me stop this Control Class C. So this is how it works. I will stop all this. So this is how open voice works. Now, if you want to reopen open voice later, let me first close this all. Simply go to Stop Menu and open Anaconda Prompt. Open this, activate open voice. The base name should change. After this, navigate to open voice folder. Please do not delete any checkpoints, so copy this cd space Control plus V. So we are inside this folder and then run Jupiter Notebook. And you can run the Demo Part one, part two, or part three, as you want and always make sure to run itself. Anyway, we hope this video was useful to you. Thank you very much. 5. Kokoro tts Free Text to Speech: Hi, guys. Here is how you can convert your text to speech using Coco or TTS. And here are some of the sample voices. Hi, please subscribe to our channel. Thank you very much. Hi, please subscribe to our channel. Thank you very much. Hi, please subscribe to our channel. Thank you very much. Hi, please subscribe to our channel. Thank you very much. Now we will be installing this ocuro TTS to our PC, and it is very simple to do. We will be using Anaconda to create a Python environment. And if you do not have anaconda, go to anaconda.com and click on Free Download. Provide your email or skip the registration and download the installer and install it onto your computer. After this is done, go to the Start menu and type Anaconda Prompt. This is the one, open it. And first thing we need to do is create a Python environment. For that, type Conda, create name I will name it Cocoro. Python is equal to 3.12. This will create a Python environment, 3.12. Y and Enter. So the Python environment has been created. Let's activate it. Type Conda, activate Cocoro and Enter. You can see the base name had changed to Cocoro. After this, we need to install CcoroF that type PIP install Ccoro and Enter. So this is done. Now we need to install spaces for that type, PIP install spaces and Enter. So this is also done, and now we also need to install radio. For that type, PIP install radio. So this is also done. Now in this folder, we have created a custom file. So this is a Python file, and we have modified the Coco RTTs demo, and we will provide this. Anyway, copy this address. And on Anaconda, type CD, Control plus V, and Enter. So we are inside that folder. Now type Python, run that PY and Enter. You might see some warnings, but you can ignore it. And here is the Cocoro TTS. By the way, if this does not author opens, kind of copy paste this on any of the browser. Anyway, this is the Cokoro interface. Input your text here, and click on Generate. This is using this voice. Hi, please subscribe to your channel. Thank you very much. So this is how this sounds. And you can simply download this. Let me choose more voices for sample. You can also change the speed. I'm using CPU and it works all good. Hi, please subscribe to our channel. Thank you very much. Hi, please subscribe to our channel. Thank you very much. Hi, please subscribe to our channel. Thank you very much. Hi, please subscribe to our channel. Thank you very much. So it does works nice and should be all good for your projects. We hope this video was useful. Thank you very much. 6. Tortoise TTS - Setup & Installation: Greetings everyone. In today's video, we will quickly learn how you can install deltoid TTS on your Windows computer. Now, we will be installing this tortoise TTS to our Windows 11 PC. This one is developed by James Metcalf. Anyway, to install this, we will use Anaconda for those who do not have Anaconda, simply Google Anaconda. And this is the official website. Simply download this and install it to your computer. We have already downloaded install Anaconda to save time. So I will close this. After downloading and installing Anaconda. Also download these files, click on code and download zip. So it will be downloaded as a zip file. I have downloaded this one too, and it is on my download folder. So here is the file, open this, and then extract these to the deck style. So here are all our files. And this is the Anaconda installation, which I have already installed. After doing this, if you go to the Start menu and type Anaconda, you will see Anaconda prompt open this. Now the first thing we need to do is create a Python environment for that. Simply type conda, create deaths, en, give any name you want. I will write TTS. Python is called a three-point nine. We have tested with Python 3.9 and it works all good. So we will recommend this version anyway. And he turned to us and enter. The Python environment with 3.9 version has been created. So let's activate this type conda, activate T ds. So you will see this base name will change to TTS. I will type CLS to clear all this. Let me minimize this. Open the guitar bass. You can find all the installation instructions here. So let's first install BY Terje. Open this a new tab. Now select as per your BC viewer touch build, choose stable OS, choose windows. Back is, you can either use PAP or Conda. I will use Conda languages, Python platform, I will use CPU since this PC has no graphics card, if you have never die, graphics Jews could elaborate, went seven or 11.8 as a UPC. For us, we will use CPU, copy this and paste it here, and then this is tan, I would clear this so it looks clean after installing this, close this. Before running these requirements installed BY sound file, copy this and paste it here. Then. Now let's run this requirement x. We need to make few eddies to do this. Open the daughters folder. And open requirements. In here. Remove this number, which is twice here. And also remove these LLVM light. If required, we will install Newman manually. Save these, and it is done. Now navigate. We are taught to us folder is they've CD. We will navigate here, copy and paste. So we're inside this folder now run PIP install. That's our requirements, dot TXT. So it is done. Now, also run this setup.py, copy this, and paste it here. So it is all done. Now let's see if this works or not. We need to run this command. Basically. It will use this text. We will use random voice and the preset is fast. You can change the voice and as will the preset, as per your requirements. To explain better, go inside daughters folder, open this EPA dot PY file. I will use a notepad. Inserts fast. So you can use ultra fast as you can read it here. Use fast for decent quality, use standard for good-quality, and use high-quality for absolute best. However, this will take less time computed fast, and high-quality takes more time. For now, let's use fast. So let's copy this. Sit here. So the audio will have, I am going to speak this text. You can change it if you want. And for now, it will use a random voices and uses fast preset. Please note, this will take quite a long time. So this has been done. Let's check result. Open the daughters folder. Go to results. So here are the audio. Let me play them. I'm going to speak this. I'm going to speak this. I'm going to speak this. These are the random audio. So you can use this audio. You can also use your own voice Fourier. For instance, here is my audio recording. So these are my own voices. All you have to do is copy this folder, go to turquoise and open voices and simply paste it here. You can check all the audios and use them to. Anyway, let me clear this. Now let's try using our own voice. We need to run some command. Here's the command. Let me paste it in a Notepad. You can write any text here. And we will use our own voice. In order to do that. We can use this voice is equal to directory name. So right is a call to the folder name. And make sure to have the exact name of the folder. Copy the exact name, and paste it here. I will use the preset fast. You can also use high-quality, but it will take some time. Now, copy this code and then paste it here. Please note, this process will take a very long time. So this has been done. Now, let's check the results. And here is our audio. Subscribe to our channel. Subscribe to our channel. Subscribe to our channel does not exactly sound like my voice, possibly because my accent is different. Please give it a try. We hope this video was useful. Thank you very much 7. Voice Cloning Tips, advantages and more: Now after successfully installing the real-time voice cloning, Let's see how it can be practically used. Let me quickly open Anaconda app and activate voice and jump to the voice folder. By the way, here's the salt. Got to do it. Simply open this folder, copy this link, type cd, and control V and enter. So we are in default itself. Anyway, let me open the tool without the data set. One of the best use of this application could be for you to videos if you are a YouTube content creator or even create other videos than this tool can be very helpful. It saved loads of your recording time. Simply browse the sound, new one to use, or you can even use your own voice. And unless fancied himself at anytime unequal to company. Let me use this one and add new content here. You can type here or copy paste and sensitize invoke code. There's no limitation for words, characters, or whatsoever, and you will not need to pay. However, they can visa can take some time depending upon the text and UPC configuration. Everyone. Today we will learn how you can run it and use real-time voice cloning. This is a great application. Thank you. So this is how the vocals will sound like. The advantages is it's auto-generated Audio. Note texts limitation then works great. You also have the option to cheesy audio speed and more. Greetings everyone. Today we will learn how you can run and use real-time voice clotting. This is a great application. However, the downside is it may take quite a long time to load. And most importantly, it may not sound as human, but more robotic. But again, it really depends on what voices you are using. For me personally, I use it for various video projects. Smaller projects are very handy. And since I do not have a grid English action, this tool can definitely help for the sum to edit the quality of the audio. You can use tools like Audacity. You can simply google Audacity and install it and try to edit the audio. Anyway, there are countless opportunities you can work with this Python program, use for various video projects like YouTube, video tutorials and more. And you can also clone your own voice. It's definitely a time-saver and hope in future religious, it will be more accurate and more humanly. We really hope you can use this application at It's based. Thank you very much. 8. How to install Deep Live Cam: Hi, guys. Here is how you can install Deep Live Cam by Hesider. This is how the live webcam looks like, and this is the original video, and here is the video after face swapping. Now we will be installing this Deep Live cam by HeckDr on our Windows PC, and it can easily swap faces and even run on a live webcam. Now, in order to install this, we will be using Anaconda. But those who do not have anaconda, simply go to anaconda.com and click on Free Download. Provide your email address or skip registration, and then simply download and install on a computer. Once Anaconda is installed, go to the Start menu and type Anaconda. You will then find Anaconda prompt, so open it. The first thing we need to do is create a Python environment. And if you go back to the Github page, and here is the whole installation guide. It is recommended to install 3.10 Python. So let's do that. Let's create a Python environment. For that type Konda create, double dash, name, use any name you want. I will say Deep live Cam and then Python is equal to 3.10. So this will create a deep live cam environment with Python 3.10. Yes and Enter. So the Python environment has been created. Now let's activate it. For that, simply type Conda Activate Deep Live Cam. Here is the code, Conda activate Deep Live Cam and Enter. So this has been activated, and you can see the busname had changed to DeepfCam. Now we need to download some files, so let's head over to the Guitar page. We will require models to run Deepivecam. So download these models, simply click on it. It's about 332 B, and I've already downloaded it. Also download this In swapper. It's about 264 B, and I've already downloaded this one as well. Now we need to download this whole guitar file. So for that, go to code and download a Zip I've already downloaded this one as well. So these are the three downloads you will require. Let me head over to my download folder. So these are the files we just downloaded. Now let's open this file. And I'll extract it to the desktop. Here is a folder, open it. Now in Models folder, open it. Now simply copy these two models we downloaded earlier. I'll copy this, but you can cut this as well. And paste it here. Make sure to add it inside the models folder. Now go back to DeepbCamFolder, copy its URL and go back to NaconaPmpt type CD, space, row plus V. So we are inside this folder. Now we need to install all the required dependencies to run this application. For that, type PIP install this our requirements of TXT and Enter. So basically, this will install all the required dependencies to run this application, and this might take some time. You might be stuck here for five to 8 minutes, so please be a little patient. So this is all done. Now we also need to install FFMPZ. So for that type Conda install FFMPZ and Enter. Y and Enter. So we have installed everything. Now let's try if the application works or not. For that, type Python, run that PY and Enter. So this is a simple application, and it is working all fine. Now for some reasons, if this is not working and giving some errors, then go back to the Gita pase, scroll down. And here you might need to install Visual Studio 2022 runtime for Windows. If you do not have it, open this link and simply download these build tools and install it. And if you have NVDA graphics, then you will need to install Koda. Please install this cuda, go to this link. And download it for Windows and install it on your PC. After that, simply run these commands, PAP uninstall ONNX Runtime, and install this version of ONNX runtime with GPU. And while opening the application, instead of using Python run PY, you can execute it with Koda. So this is the code for NVDA graphics. We are using Anaconda, so we do not require to create this virtual environment, so we are skipping this. Now, everything is done. Let's see how this works. Now, this is a simple application. Here is an image, and here is a sample video. I download it from Pixels and Pexels. Click on Select the Face. So this is the image, open it, and you will see the image here. If you want to preview for the live chat, select the webcam here. I will use the integrated camera and then click on Live. So this is how it looks, and this is a live webcam video. It really depends on your PC on how smooth it works. Mine is a little sloppy. It's doing right, but this is a live webcam, and you can see the moments, too. So this is a live webcam feature. Let me close this. Now you can also use this app to swap faces on a video. So this is a sample video. Let me open this. So this is a sample video, and it's not that smooth. Anyway, let's select the target, and in here, instead of image, choose video. So this phase should come here. Use better images and videos. Now simply click on Start. Save the output to your location. I will choose Desktop, click on Save. This will create this temporary folder. And in here, you can see processing. This means it's working. You might also see this error frame processor not found. For us, this app is working all good, even with this warning. But if that does not work for you, then we will create another video to get rid of this error, but this worked all fine for us, and this processing can take lots of time depending upon your PC, graphics, and more. So the processing is done. You can also see processing to video succeed. So here is the output. Let me open this. So this is the processed video. And this was the original video. Now, the video itself was a bit glitchy, but this is how the app works. Let me close this, or you can click on Destroy here. So this is all done. Let me close this. If you want to reopen this app, it is very simple. Open the anaconda prompt. Here is it and activate the environment. Conda activate DepiveCam and Enter. So you will see the base change name to DepiveCam. Now we need to navigate to the DepivecamFolder, type CD and copy the URL. And paste it here and enter. Now we are inside the folder. After this, simply run Python run PY. And if you are using NVDer simply run it with Kuda. Now, here is the application, and you can use it as you like. We really hope this video was useful to you. Thank you very much. 9. Frame Processor face enhancer Not Found: Hi, guys. Here's a quick fix to frame processor, face enhancer not found on Deep Live cam. In our early video, you can find the full guidelines on how you can install Deep Live Cam. Anyway, so let me quickly open it. Let me open the anaconda prompt, activate the environment. Let me navigate to the folder. So I'm inside the folder. Now let's run the application. The application is running, but you can see there's an issue, frame processor, pace enhancer not found. So let's resolve this. Let me close this. Now we first need to uninstall Basic ARF that type PIP, uninstall Basic AR, that's Y and Enter. So this has been uninstalled. Now let's reinstall Basic AR with this command. So copy the code and paste it here. After that, again, unstallGFP JN, for that type PIP, uninstall GFP, GNI and Enter. So this has also been unstalled. Now, let's reinstall this again, copy the code, and run it here. So this is done. Now let's try to open Depive cam. Python Runner PY. And the application is running all good. And you can see there are no errors. So please do give it a try. Thank you very much. 10. Create Ghibli Style Images on your PC without ChatGPT or Grok AI: Greetings, everyone. In today's video, we will learn how you can create Zibl type images on your PC. And this is absolutely free. No hidden cost. Now, in order to create a Zibi type images, we will be using CompuI. I will provide all the links. This is on Github, and you can use it for various other purposes, too. And the best part is, it is absolutely free. Here are all the details on how to use this, and I will run through the quick setup process where you do not need to run any commands whatsoever. Anyway, on this Github page, simply click on these two legs. Scroll down. Download this comfy UI Window Sport table. To make the video shorter, we have already downloaded this. The file size is about 1.59 ZV, and you will require Seven Z to extract the files. If you do not have seven Z, simply Google seven Z. And you can download this for free and simply extract it. After downloading this, we also need models to run CFIUI. We need checkpoints and Laura's, and there are varieties of this available. You can experiment with many for better results. To download it, go to the CivitaR website and download this dream SEPA checkpoint. I will provide the direct link. Simply click here to download. It's absolutely free, and the file size is about two Z. After that, go to this link and download this ZB style Laura. Now this is one Laura's model. But if you search here Zibli you can find various models, and you can experiment with it accordingly. To download this one, simply click here and it will be downloaded. So these are the only files required. To recap, we download Comfy UI, Dream shaper, and Zibli style La. I have downloaded all of them. Let me quickly navigate to the Downloads folder. So here are all the downloads. Now let me extract the Sven c file, open it. And I will extract it to the dektf. You can extract it or you can simply drag it to the deckst. Please note this will take some time as the file size is quite big. So you will have to be a little patient. I will just fast forward it. Anyway, the file has been extracted. After that, go back to the Downloads folder and copy or cut this dream Saper checkpoint and open the CompuI portable folder, go to Comfy UI, find models folder. And open checkpoints and paste it here. So this one is here. Now go back to the models folder and in here, find Laura's. So here is the Laura's folder. Open it. And now copy this sibly style. Copy. And paste it here. So this is all done. Now go back to Comp UI portable folder. Now this is all the application files. You can run this CPU VAT file, or if you have graphics, then you can run this NVDA ZPUVtFle. This laptop does not have a ZPU nor any graphics card. So I will run the CPU model. Double click on it. Again, it might take some time to load all the application. So here we are, and at first, this may be blank, or you might see some defaults. After this, let me close this. Now, here is the dot Json file we have created for Zibl style images. Simply drag this file to CFUI and you will see all these. All this has been already set up, so you don't have to do anything. You only need to choose the image you want to Ziblify. Simply click here. Our sample image is on the desktop. This is a sample image. The sample image was downloaded from Pixa B, so this is done. So this is the checkpoint we used. As mentioned earlier, you can use other checkpoints as well. Simply download it and put it on the checkpoints folder. Civit AI offers lots of checkpoints. And in Las, you can customize it further as well. This is a Zibl style we downloaded. You can always experiment with other Las files as well. Different models will give you different results. So we do recommend experimenting with various checkpoints and Las. Also, you can adjust or experiment the strength. Again, CVD AR is one great website. You can simply search here. So there was a spelling mistake. Anyway, you can find many options here to experiment. Please do give it a try and as well let us know which one works the best. After downloading the Las, simply put it in Laura's folder, and you will have it as a job deer option. And here is the prompt you can play with this too. This one is the positive prompt, and this one is the negative prompt. And you can also adjust these settings, adjust the noise, adjust the steps, CFG, and more. And this is all done. Now let's run it. Since my PC does not have any graphics card, it takes incredibly longer time. To generate a single image, it takes more than 50 minutes for us. That's one reason we could not try with many models, as it is very time consuming. Anyway, in this CMD, you can see all the progress. I fast forwarded it. It really took quite a long time, more than 1.5 hour. Anyway, this is all done, and if you come back here, you will see the preview image. To see the output image, open the ConfuiPortable folder and go to Comp UI folder. And here you will find the output folder. And here is the output. So this is how the image looks like. And this was the original image. Now, again, you can experiment with various combinations, play with noises, you can increase the steps and more. You can try with different Laura's models and with different checkpoints. Each of them will give you different results. And this Civita website is a great place to search. Search for the models and various gibly styles. You can as well use this VUI to generate other images, too. So please do give it a try and let us know what works the best. Thank you very much. 11. One click face swap using Roop: Greetings everyone. In today's video, we'll quickly learn how you can face swap only using a single Emase. Now we will be using this sample Emas and video for this tutorial. This video has been downloaded from pick cells, so we will swap this face, and this image has been downloaded from pig cells. Anyway, to do this, we will be using this Guitar Repo group. Please note, this project has been discontinued, but it still works. Now to use this, we will be using Anaconda. Simply visit Anaconda.com Click on free download, download it, and install it on your PC. You can check other videos on how you can do this. After this is done, simply go to code and download the zip file. The file has been downloaded. Show in folder open the zip file, I will extract it to the dex star. Here is the file. Now go to the start menu and type Anaconda. You will see Anaconda prompt after installing Anaconda. Anyway, open in. Now let's create a Python environment for that type Conda. Create, give any name you want. I will write rub face Python is equal to 3.10 We tested with Python 3.10 so we'll be using this version and enter Y and enter the Python environment has been created now let's activate this Kunda activate rub phase and make sure the base name has changed. After this, we need to run some codes which will be provided. We need to install this on X run time and FFMPZ, copy this code and paste it here. And enter this is done. Now let's install FF MPC, copy this and paste it here. And enter this is done. Now we need to navigate to this folder, open this folder, copy the address and type CD space control bus fee. We are inside this folder. Now let's also install other required files for that type PIP, install R requirements XD. This will take some time as many files will be downloaded. Let me clear this all now. In order to face swipe, we need to run some commands. Here is the command, please replace the video link. I will copy the address and replace it here. And make sure the name is exactly the same. It is video pf file and the other one is JPZmae. Video quality is eight. Let's change the Ema source link. Also change the output link where you want to save the swap video. The output video name will be swap for I will be using the CPU. If you have a graphics card, you can also use GPU instead of CPU anyway. Now simply copy this whole code and paste it here and enter. This whole process can take minutes to hours, depending on the videos, your PC, et cetera. The video has been completed. Now let's check the video. This is the swapped video. This was the original video. The quality has decreased a bit, but please do give it a try. Thank you very much. 12. Installing Deepfake - First order model: Greetings everyone. In today's video, we will quickly learn how you can install the deepfake first order model, a Python application on a Windows computer. So basically, we will animate this MAs with this video to produce something like this. Let's get started. Now in this video, we will be installing this Python application by Alexander. Here are all the details for this Python application. We will make this application run on our PC. So the first thing we need to do is install Python. There are numerous application that can install Python, but we will use Anaconda, open the browser answers for anaconda. This is the one. Click on it and simply download it. This can install Python 3.9. Anyway, download. The download will begin. I will cancel this for now. As I've already downloaded it. I will close this all. So here is the Anaconda. Let me install it and navigate accordingly. This installation will take a few minutes. So it's almost done. I would untick this and finish. The Anaconda has been installed. Open it, simply search Anaconda. So this is the command file, run it as an administrator. So after this has opened, the first thing we need to do is create an environment and installed Python, do that type conda, create that sand and give any name you want. I will write def fake. Python is equal to 3.7. Please note, we did try with 3.9, but it did not work. And press Enter. This will create an environment named defect with 3.7 Python. Why? And enter. This has been installed. Now we need to activate the fake. You will have to activate it every time you want to use it, to activate type conda, activate the fake, and enter. So here you can see the fake. After this is done, go back to this space and download all of this code. Click on code, and download a zip. So it's downloaded. So in folder, open this and I will extract it to the deck star. So here are all the Python files. Now we need to get inside this folder on Anaconda prompt for that, simply copy the URL. And here type cd space and paste it and enter. So we are inside the folder. Now, we need to run this code to install all the requirements, copy this, and paste it here and enter. Now everything has been installed. However, there is some error for NumPy. So let's install numpy for that type Conda install Anaconda NumPy and enter Y. And Enter. It has been installed now to check if everything is working or not. Type Python demo dot PY. Here is an error. No module named YAML found. Actually, there is a list of modules that will be needed to install to run the application. Let's first install YAML for that type Conda. Install. See conda first BY YAML and enter. So it has been installed. You will required to run python demo dot p-y again and again to see the missing modules and install it, I will provide the list of missing modules that needs to be installed. But it may depend as per your PC anyway and install t q dm installed emails that you installed scikit. Please make sure to install 0.1 at version as newer versus did not work. Installed BY torch. Installed, SKLearn install, pandas installed. That reason. Install FFM. Pz installed emails IU, FFM peasy installed emails IU BY HIV. So after all this installation, let's run the demo again. If everything is installed correctly, you will see this message. We now need to specify the source MAC. Source video. Checkpoints has a CRA. After specifying everything, we can animate our emails with the video. Let's again go back to the instruction pays. Now, we need to run this code. Let me copy this and paste it to the notepad. We need to add this data set name, which is in the Python files folder. Let me open the folder. It is inside the config folder. You can use any of these data sets, but I will use this one. Let me copy the exact name by renaming and copying and replace it here. For the driving video. I have a video here that I want to use. You can create your own video as z. Like this is the expression I like to use, and this is the emails. Let me copy this and paste it in the same folder. So here is the email which is in the PNG format, and here is the video file in mp4 format. So replace the path to driving with the video name and its extension. You will not required to add this whole address here for this video address. Likewise, Let's add an e-mail address. For sources, the maze replaced with the e-mail name and add this extension. Now we also require the checkpoint. For the checkpoint, again, go to the web page. And here is the link to download the checkpoint. You can use either of these two links. For now, I'll click on Google Drive. And here you can download any checkpoints and you can use any checkpoint as you want. All you have to do is click on this icon to download. The file size is about 700 MB. So I've already downloaded it. So I have this file downloaded. Here is a file I downloaded earlier. Let me copy this to the same fight and folder. I will put it inside the config folder, but you can keep it anyway. Let me copy the file name and replace it in the pathname here. Also at config in front as it is inside the config folder. Don't forget to add the extension dot TR. Actually, you did not need this slash. Anyway, a quick recap. This is inside the config folder. We are using this one. Driving video is the video used to animate the emails. So this is a video here. The source image is the image we want to animate. Here is the MAs. And the checkpoint is what we downloaded earlier and kept inside the config folder. And do not forget to add.dr extension. Now let's copy this code and paste it here. Here, starch is not compiled with cuda enabled. This is because of the graphics card. Since this laptop does not have a graphic card, it is displaying an error. If you have a card and want to use that, simply Google. And I conduct, could I install and click on this link and install the cruder version as per the requirements. But since this PC does not have a card, I will not install cuda. So instead, I will use the CPU to create the video. To do that in the earlier code, add double dash, CPU and space. Also note that if any module is missing here, you will first need to install it. Anyway, Let's try to copy this code and paste it here. So this seems to work, but please note the processing takes some time. So this has been done. You will see the final video here with the name of result dot mp4. Let me open it. Here is the final output. So this is how you can install deep fake Python application to your Windows PC. Thank you very much. 13. How to transcribe Audio to text using Python application: Greetings everyone. In today's video, we will quickly learn how you can install free audio to text transcriber, a Python application to your computer. Let's get started. So in this video, we will learn to install a free Python transcriber whisper by OpenAI. Basically, we will be installing this Open AI whisper to our Windows computer and it works on both windows 11, but we are installing on Windows 11 anyway, the first step we need to do is install Python and we will be using Anaconda for that. Simply Google anaconda. This is the one. Click on it and click on get additional installers. And you can install Anaconda for Windows, Mac OS, and Linux. I have already downloaded and installed the Windows version. So I will not be downloading this. After installing Anaconda, go to the Start menu and search for Anaconda. We will be using the Anaconda Prompt, Open it. After this, we need to create an environment with Python installed for that type. Conda space, create dash n and give any name you want. I will write dex test speech. And Python is equal to 3.9. This will install Python 3.9. Please note, you can use any name for this text as speech and press Enter. So now we have successfully created the Python environment. Now we need to activate the environment we created. And here is the code. Conda, activate, takes death speech. Please remember, this is the exact environment we created earlier and enter. You will see that this base has changed to take Stat speech. After this is done, let me minimize this. Now, download this Python application files, click on code and download SZ. So this has been downloaded, show in folder. Open the zip file. I will close this all. And I will extract these to detect stop. Here is it encloses two. Now we need to navigate this prompt to the Python folder. The easiest way to do is to open the folder, copy this address, N here, type cd space and control plus V and enter. So now we are inside this folder. Now we will require to install all the required files for this application. For that, we can use this code, copy this URL, and paste it here, control plus V and Enter. Now, this will install all the required files. So it is all done. Now we also need to install FFM PZ for that simply type conda. Install dash xi, Conda, das fours, space F, F MPC, deaths Python. And press enter Y and Enter. So this is done. And now we might also require this setup tools, dash trust. I will install this to copy this and paste it here and enter. So now we have install all the required files. Now, let's see if this works or not. There are a few options to use. Tiny dot e. Bay's small, medium or large. I will use the medium model. So here is the code we're going to use. Now before that. Here is the sample audio I want to transcribe. This is a small sample audio I want to transcribe. And so my fellow Americans ask not what your country can do for you. Ask what you can do for your country. So an Anaconda, we need to navigate to that folder. Simply copy the address c, d, and paste it here and enter. So we are in the folder where our sample audio is. Now here is the command we want to use. Basically, we want to convert our sample audio to text using model media. So I will copy this and paste it to denote bad. Now on this code, we need to replace it with our sample or your name. So our sample audio is test, and it is a MP3 file. You can see it is dot mp3. So replace this all with our audio name, right? Inverted commas, and write this dot mp3. We will be using a medium model. Basically, tiny model is less accurate, but fast, and so on. For most of the audio, the media works great. And since this is the English language, I will also write double desks language space English. This field is not mandatory. If not mentioned, it will auto detect anyway. So this is the final command. We're going to use, copy this and paste it to the Anaconda Prompt. Also make sure we are in the location where we have this sample audio. Please know the conversion will take some time. I'm fast-forwarding it. So here is the transcribed audio. By the way, it might also download a model file if required. The file size is about 1.42 CB. So this might take a longer time for the first transcription, but later it will be a little quicker Anyway, here is all the transcribed text. Alternatively, you will see the files in the sample for the two, you can find the text here. So here is all the text here. You can also find SRT file open with Notepad. So here is the timestamp two, and it will also create a V D D file. So this is how you can convert audio to text file. Again, here are lots of options for the conversion. You can also act to translate by using this command for more information. You can use this help, Good to go to whisper does help. And here are all the information you want to know and other details. So you can explore and experiment more with all this information. After everything is done. The next time you want to run this simply Open anaconda. Make sure it is Anaconda Prompt, not this one. Open this one. Activate the text speech environment, so the best names will change. Then navigate to the folder where you have the sample audio. Copy the address, cd space, and paste it. And then simply use this command line. The file name, and type, as per your requirements. Change the language here. Or you can also remove this line. And it will auto detect the language and enter while it is working. You can replace media with tiny.en based on EN, Medium dot e and etc. As we did not specify the language, it is trying to detect it. The language has been detected. So you can see it is all done. And again, the folder will have all the files. Please note, this process will take some time anyway. We hope this video was useful. Thank you very much. 14. Hindi TTS using Python: Hi guys. Here is how you can install text to speech for Indian languages. Now we will be installing the Python application to our PC. This is called indic TS, and here are all the required files and processes on how you can install it. In order to install this application, we will be using Anaconda. For that, go to Anaconda.com and click on free download and download the application and install it. The installation process is very simple. If you are unsure how to do it, please check our other video on how you can install it. After Anacona is installed, go to the Start menu and type Anaconda. You will find an Anaconda prompt, open it. To get started, we will need to create a Python environment for that type dare give any name you want. I will write Hindi and Python is equal to 3.11 This will create an environment name Hindi with Python 3.11 we tested with Python 3.11 Yes. And enter the environment has been created. Now let's activated for that type, conda activate Hindi. Please note you will see the change in base name to Hindi. Now, let's download these files. Click on code and download zip. The files are downloaded. Show in folder, open the zip file, I will extract it to the Dexter. Now, open this folder and copy this sewer address, and go back to Anaconda and type CD space Troll. Plus, basically we are inside this folder. Now, now we need to install all the dependencies to run this application. Everything is inside this requirements text. Anyways, on Anaconda, simply write PIP install R requirements dot. This will install all the dependencies for this application, since it will download many files. This will take some time. You will be on the same screen for a couple of minutes. Please be patient. Now, everything has been installed. Now there's only one thing we need to do. Go back to the guitar pas now we need some models to make this work properly. For that, click on this link to download. Here are several models and you can see the sizes which is nearly 1.5 ZB. Let me download this in English and Hindi, the size is about 1.4 ZB. It has been downloaded. Show in folder and open the z file. Let me close this. Now we need to extract this file. Open this folder again, and create a folder and name it as models. Open this folder again, and create another folder, V one, and open this folder. Now, extract all files to this folder. This is done. I'll just rename these to I. This is done. Now again, go back to this model's folder and create another folder with the name Fast Pach. Open this folder and create a new folder V one. Inside this V one folder, again, create a new folder EN. Now inside the EN folder, we need to extract a file. Open the Z file inside the fast pitch folder extracts Speakerph. This is done. So basically we created a models folder and saw folders. We have done everything. Now, let me close this all. Now here is the command you need to run to get the Hindi audio. You can write any text here, make sure the text is in Hindi. The model path is already allocated, the config file is all done. Please note, if you change the folder name to something else, make sure to change it here too. The output name is this one. You can change this one too. Basically, the output would be generated here. Let's copy this and paste it here. Control plus V and Enter. It is done. If I open this folder, here is the audio. Let's listen to it. Oh, the audio was muted. Channel subscribe, Ripa, Youtube channel subscribe, so this is how it sounds. Do give it a try. And by the way, you can also tag this to mail and it will generate a male voice. We hope this video was useful. Thank you very much. 15. How to update packages and uninstall unused packages Anaconda?: Greetings everyone. In today's video, we will quickly learn how you can free up some space by cleaning the old packages from the Anaconda. And as we'll update all the packages, let's get started. Now on Windows PC, if you go to the Start Menu and says for storage settings, this is the one. You can see all the details of the space taken on your computer by apps, temporary files, etc. When I quickly checked, this is my app and it's taking up reasonable space and the rest looks fine to button others. And that condyle was taking about 23 stories. This is very high because I only have one environment taking about 5.5 dB if I click on it. And if I check the back is a store is it is taking up more than 25 years space. This environment is occupying about five ZB. Anyway. Let's update all the packages and as we'll clean the unused PAC is, is. Let me close this. Simply Open anaconda, n type Conda update, dash stash all and Conda. Clean dash p. This command will update packages and this will clean the unused packages. You can only run this command if you don't want to update bag it is anyway, press Enter. So the update is done. Why? And enter. So here you can see there are about nine z v of unused packages which is about to be deleted. So it is done. Close this. Let me quickly check the storeys. Other so you can see the Anaconda size and reduce nearly by AGB And as well, all packages are updated. We hope this video was useful. Thank you very much.