PyTorch vs TensorFlow: Which One Should Beginners Learn First?

Hey everyone, I'm your myth. I'm back again. I believe uh you all got the data set downloaded and unzipped in your local right. So we are all ready. Let's directly jump and start working with our project. I hope you have installed Jupyter environment or if you are using any other environment that's fine but I suggest Jupyter environment. So in Jupyter environment I suggest creating a new folder. So you can organize things better. So you can name anything. Once your folder is created, have a new Python environment. You can u name this environment anything. And your environment will be ready. I'll just give animal classification. Ignore my spells. That's fine. Okay. Before starting our project, we need to understand what exactly neural networks are. Before that, let's understand PyTorch and TensorFlow. What exactly they are? PyTorch and TensorFlow are deep learning frameworks. PyTorch is built by Meta, TensorFlow by Google. Let me give an detailed example to understand the two better. I do have a line image. Now I give this image to model. Now we all know that models won't understand images, text or anything. We need to convert that to Nimbus. Now one or two images I can convert to Nimbus and I can do matrix multiplications and I can make model to learn the patterns. This I could do if it takes time to what if I do have 28,000 images like our data set that's very huge right humans for sure you can't do it. Obviously this models came into the picture. What exactly it will do is this will provide you the tools. So this frameworks are mainly used to train your neural networks. Your neural networks will learn patterns. All the key matrix multiplications will be done with this frameworks. Again you have an flexibility here like what to choose pyarch or tensorflow. I suggest going with pyarch because pyarch is very simple to learn. Right now in many models they are using pyarch. And coming to TensorFlow, it is like bit complex and it's production grade. Uh in many enterprises they use it. So we'll start with pyarch later we can use tensorflow. I'll show you in the next video how exactly we install pyarch into our environment. So we can start working with pyarch. Until then, uh, see