 So, we need to talk a little bit about cost logistics. The first thing is there's multiple points of contacts between you and the cost. The primary here and that we really focus on is Google Colab. The tutorials that we prepare for you run in Google Colab. The videos I embedded into the Colab and play right out of it. You can also find them on a YouTube playlist where you can view them full screen if you want to. The questionnaires that we have will run on Google Colab. We feel that it's really helpful for your learning experience to when we ask you questions to actually have to write them in text right into the Colab. We will use those questionnaires to know that you're with us and we will ultimately also use that for grading. The homework is also presented in the Colab, so it's all we try to as much as possible use Google Colab for our link. The second point of contact is your TA slash part. You will be meeting with your TA slash part in Zoom and you will have the chance to all talk together there and to talk in group with breakout groups. The third point of contact is we will use Piazza. Most of your questions you should direct to your TA and your group because by discussing those questions everyone learns. But for some questions you might want to ask that in more of a forum and that's what Piazza will be helpful for. Lastly we will have some crowdcast sessions that are live sessions where you get to ask Lyle or me or anyone else who participates. Interesting questions and we believe that the ability to ask questions to the professors is also essential for having fun in this context. Lastly feel free if something's not right or if there's something that you can't learn with the other approaches to just email me directly. I'm carding at gmail.com. Now for the people at U Penn taking this course I imagine a lot of people are interested in grading. Your participation which includes sharing what you do not understand with your group will make up for half of the grade. We will be able to automatically track using the question asked for which we use air table and other means if you're there if you participate. We will equally track your homeworks which count towards participation as well. The second one is the project that you do at the end will make up the other half of the grade. Now what's the philosophy there? In a way grades are an ugly thing. We want you to learn but we want to make sure that a good grade is correlated with you trying really hard to learn the interesting things here. So that's why the combination of participation and the project is what will make up the grade for this. Testing and grading however is not the important part of this course. The important part of this course is that you learn and get to be good at solving the eblon in France. Let's briefly say a few things about projects. Towards the end of the course you will get together in groups of two to three students always from the same part and you'll tackle a truly interesting deep learning problem and that problem must be chosen so that it relates to social impact and then you will produce a video and a write-up about it. My experience with such projects is that if people put their heart into these projects they're often very cool and they lead to new applications of deep learning and machine learning towards real problems. Now let's have one more question for your part. Tell us how important theory is relative to practice to you in the context of deep learning. Discuss with your part where you're in the continuum and ultimately share it with us.