I hear you Vikram, i felt like the content felt rushed and lacking, so dropped off the first week and saved my money - but the Slack community was vibrant and fun
I did finish Andrew Ng’s old coursera course long back, so i think deeplearning.ai is a derivative! As for the Udacity nano-degrees, thanks again for clarifying. I’ll invest all my energy into this course.
Your personal experience and immense knowledge in this field and your(+Rachel’s) availability.
The top down approach of explaining concepts with examples and allowing real time testing with Kaggle
Structure of the course, meaning, interactive learning rather than recorded sessions. Special mention Forum, it allows people to connect and learn (I found it useful while following part 1 & 2 on youtube)
Discussion of the latest papers and improvising accordingly, allowing students to absorb the best as of today (Like W-GANs from part 2, using pytorch in this course)
Last is a little philosophical, your purpose of this course is to educate! All things with purpose grow big!
So in contrast, other courses may achieve 2,3 and 4 but 1 & 5 makes fast.ai what it is!
Others, please share your views.
Hello all! I’m from the US but I live in Mexico.
I have an MS in Electrical Engineering and worked for a defense contractor for a while. I got out of defense and am working remotely now. I took a couple of classes in graduate school (10 years ago!) about Neural Networks, so I have some training and background in it. I’m here to get the modernized take on it and get Kaggling! My hope is to finish a nice portfolio of work and build up my Kaggle-muscles and get a deep learning job somewhere.
Self-driving car nano-degree covers basic image classification, transfer learning and semantic segmentation. It’s more of an introduction to Transfer Learning (similar to Part1 lesson 1) and the content for sematic segmentation is same as https://nvidia.qwiklab.com/focuses/preview/2193.
Basically that is all the DL in Self-Driving Udacity degree.
And Deep learning Nanodegrees covers simple introductions to image classification, RNN (sentiment analysis), intro GAN on MNIST etc. They are similar to introductory tutorials on Kera’s blogs and don’t really go into papers, neat tips-n-tricks or advanced materials as you did for Part I and Part II.
Like most people i started out with the Andrew NG’s course on ML when i was in college. I am eagerly looking forward to this course, since Deep Learning is something i am still relatively new to. I have experience in building traditional ML models for production , but i feel now my knowledge slightly inadequate w.r.t Deep Learning.
So this year , it has all been about learning and getting my fundamentals strong. My learning plan or syllabus is entirely based on MOOCS (https://github.com/whiletruelearn/data-science) . I have so far completed the Pytorch course on Udemy and The first course in deep learning specialization. I also took a shot at the neural style transfer problem to build a prisma like app for a hackathon. I look forward to learning a lot from this course and getting to know all of you a lot more.
how did you like the Udemy Pytorch course? i signed up but had no time to dive in
It is good. The course doesn’t delve in detail to all the topics but is a good intro course to PyTorch. Emphasis was on getting learners familiar with the PyTorch API, yet the instructor gives a high level overview on most topics.
I don’t have any experience with other DL libraries such as Keras and Tflearn , so found the ease of implementing things in PyTorch to be really cool.
Hello all, very interesting stories here!
As for me I left my job as an Android developer almost 1 year ago and I spent the last year learning everything I can on machine learning/deep learning/AI. I’m from France but I originally come from French Polynesia. I’m very excited to start this course, I completed part 1 & 2 already, thanks a lot for opening these new lessons to us for free @jeremy . You and Rachel truly are great an inspiring people.
If you guys by any chances are interested by my story and want to spend some time listening to a poadcast here is my story on superdatascience with Kirill Eremenko (author of the famous courses on ML/DL/AI on Udemy) .
Thanks for sharing @Ekami! Since you’ve already got a lot of experience, I’m sure you’ll be able to help us improve the fastai library during the course, and help other students too!
For sure @jeremy . I already started writing mine but I believe I better work on the the one from fast.ai . I didn’t take a look at fast.ai lib code yet but that will be amazing to see it working on Pytorch as well as openmined as they both have more or less the same API
Who is working on what?
It seems we have something in common. I am very interested in applied DNN to computer communications and networking problems, which includes synthetic data generation.
Hello all, My name is James Birchfield (Birch), I live in Frankfort, KY (USA) but work out of San Fransisco, CA (USA). I’ve been a programmer for 25+ years and primarily work on large back-end systems. My primary languages are Java, Scala, and quickly learning Python more and more.
I have dabbled with ML is some sense for the last 5-6 years but never really dove in. I created a Java/WEKA tutorial for the Kaggle Titanic competition, and I have started playing around with Kaggle Spooky Author Identifaction here.
I do not have a strong formal math background but have spent part of the last few months on Khan Academy brushing up, and leaning python with tutorials on numpy and pandas.
I have been looking forward to this course for a while now and am excited to get started!
My name is Rajat. I started my machine learning journey in October 2016. I like applying ML/DL algorithms to Social Media as you get to work with real time data and problems belonging to domains such as NLP, Computer Vision etc. .
I am excited to learn more about Pytorch, Numba, LSTM, GANs, RNN etc. .
I’ve spent time with Machine Learning ND and AI ND. There’s a lot of theoretical knowledge for all the traditional branches of AI and ML and little on NN and the latest advances in deep learning. The hands on approach. This forum the practicality of both part 1 and part 2 sets fast.ai apart.
Wiki: Lesson 1
I started studying DL/ML in the past year, reading papers, working through various tutorials and online classes. I have a personal interest in audio applications and video. I had been working through the 4th week of the online version of course 1 when I saw the exciting opportunity to join this v2 offering. I found the practice of applying these techniques and rewriting the notebooks really useful in feeling more comfortable and proficient in applying these techniques. I’m looking forward to learning more, getting better and getting to know others in the class over the next 7 weeks.
I am a software engineer from Montreal, Canada. I have been learning ML/DL on my own and on-and-off for the past 4 years. This is an amazing opportunity to actually focus and follow along. I am very much excited by the fact that this is geared towards coders and more of an applied course. Doing has always been the best way for me to learn.
Best of luck to everyone and happy learning !
Hello deep learners!
My profession is primarily in web design but I have an affinity for AI related topics and have seen andrew ng’s ML series on youtube. I also help out kids with robotics so all this is up my alley. Here’s my linkedin if you want to connect.
Is that through some program you’re involved in? Would love to hear more!