The local chapter of Women in Data Science (WiDS) here in Madrid offered a prize to the best ‘Spanish’ team in last year’s datathon of a Data Science Boot Camp at IE University: a 15 week in-person course.
Now I am 12 weeks in and really enjoying working with other people offline. Of course, I’ll bring my classmates to fastai once the course is over; so far it’s DL free.
Hi everyone,
I’m Satya Prateek, I’m the data scientist at Marigold Health (https://www.marigoldhealth.com) where we do a lot of NLP for mental health use cases. Our data sizes are relatively small so we get a lot of work done out of transfer learning techniques and smarter features.
This course is interesting as a means to check out the new fastai library. This is also a neat way to get me to do more Deep Learning stuff. Needs of early stage startups means a lot of my work these days is product engineering, data visualization and grant management/reporting (ughh).
Hi everyone,
After yesterday After-Christmas gift, I hope you are having a jolly day like @mrfabulous1 like to say.
My name is Farid and I live in the wonderful city of Montreal (despite the -12 degrees today which is pretty mild in comparison to -30 sometimes). I did my master and my PhD in biomedical engineering in France, moved to Montreal where I did a postdoc at the University of Montreal, worked a couple of years as a researcher, then started a company where I developed StatMap3D a Statistical Brain Electrical Activity Mapping software (it looks like a weather forecast map but for brain electrical activity), and DataFinder (a multidisciplinary clinical and research database), gave several Java courses around 2000 when Java was a thing.
A couple of years ago, I accidentally discovered fastai course while I was following Andrew Ng’s excellent courses. I followed all fastai deep learning courses and I learned so much. Every fastai course is a gem packed with concentrated useful information. In my opinion, fastai courses are the best.
When Jeremy started the fastai v2 walkthrus last September, I got hooked to this new library once and for all. Fastai v2 is really a Deep Learning Library on Steroids. I cannot thank enough both @jeremy and @sgugger for their hard work and their vision to transform an already popular fastai v1 into such a powerful and promising library as this new one. I am pretty confident that they will do it again in a near future (with swift4tf maybe 
)
I am actually porting (to fastai v2) the excellent timeseries modules that @oguiza built, he is the one who started the fantastic timeseries study group. I already have a working version. I am in the process to clean-up and polish my notebooks and right after that I will share them with you.
I would like to also start a timeseries forecasting study group where I will post what I learned and the resources (articles, libraries, blogs, etc.) that might be useful for others. Is there anybody interested in this area?
Thank you all for contributing in building such a fantastic community!
Good idea. I’ve added that request to the top post.
Oh wonderful! Can’t wait to see it…
Hi farid hope your having a beautiful day to day!
Hearing stories like yours and that you have been inspired by the generosity that people like jeremy, @sgugger and the many other wonderful people in the fastai community show, is what inspires myself to be beautiful to everyone in the community. I am still working on my super project, learning fastai and helping in the community. Glad you like my salutation, when I joined the community I decided I would always only send beautiful messages!
Have a jolly and beautiful day mrfabulous1
Hi everyone 
I started programming in August 2018 and fastai in October 2018. I knew nothing about deep learning at the time, so the course blew me away in two key ways: what deep learning can do, and how accessible fastai makes it.
I’m interested in the intersection of AI and creativity, especially with regard to filmmaking. I’d like to build tools that empower creators in ways we couldn’t imagine pre deep-learning. The course inspired me to build my own dataset to start doing so. I wrote an extensive blog post explaining the concept of visual language in filmmaking, deep learning 101, the dataset w/ sources in depth, and learnt front-end web programming to build it from scratch. I’d recommend seeing the sliding heatmaps if not the entire post:
https://rsomani95.github.io/ai-film-1.html
This blog post got me some attention, and recently, I joined the Synopsis team and am a key contributor to CinemaNet, their deep learning system. If you do any sort of video related work, Synopsis is worth a look.
I’m currently moving through Deep Learning from the Foundations at a snail’s pace; the course is packed with advanced techniques, and Jeremy’s communication skills make it a delight to go through. Fastai has been a very rewarding experience. I’ve made some contributions to the v1 library and also to torchvision since taking the 2019 version of the course.
I’m convinced that fastai is flexible enough to extend it’s functionality to pretty much anything you can imagine. I’d like to reach a point where I can contribute along two lines particularly:
video module for fastaiONNX and CoreML.Pleased to be here and read all your inspiring stories!
Congratulations! Very well deserved - it was a brilliant post.
I would love to see that.
Please keep doing that @mrfabulous1. You always put a smile on my face when I read your greetings. Apart of being @mrfabulous1, you are Mr Deployment on Render . Thank you for helping so many people in their deployment experiments. You are so patient and generous with your time.
Always puts a smile on my face personally
Hi, this is Manikanta Sangu from Bengaluru, India.
I work as a computer scientist at Adobe around C++, iOS.
I was part of the fast ai v1. Thanks to Jeremy, I have started blogging and written a couple of blogs around my course time. I am very interested to see this v4 and play around with fast ai v2. I also want to restart blogging this time especially with the github fast template jeremy has shared.
I feel good coming back and typing in forums. I love to get connected with more people and collaborate around mobile and ML.
Hi Everyone,
My name is Arvind and I recently relocated to my hometown, Bangalore after spending more than a decade in the US. I’ve joined a robotics company, Invento Robotics as Chief Architect and am actively recruiting deep learning engineers. We use deep nets extensively for robot vision, conversational interfaces and indoor autonomous navigation. I’m here to learn more about deploying compressed, quantized models on low powered edge devices.
Thanks Jeremy for the invite. It’s great to be back here again!
I just read up on the post. Really neat stuff. 
I was thinking of all the various shots in mr. robot the whole time, esp. the final season, and saw that you’ve included a shot with a remark. 
Hi, I’m Yonatan. I was lucky enough to hit fast.ai v1 in the early stages of my dive into ML (from pure physics) and I think of it as a major contribution to my successful ML career since then.
I have since participated in all previous versions of the basic and intermediate fast.ai courses and there wasn’t a single lesson where I didn’t learn something new. I’m quite confident it will be the same now.
I recently started working on a project involving AI and bee-hives, using ML to improve current methods of beekeeping. This course came right in time for me - I have a specific computer vision project which i’d like to try with the new 2.0 library.
I’d like to express again my gratitude and love for the fast.ai core team and members - Even if you don’t know it, I think of you as my mentors… Whenever I confront a tough project, I have higher confidence because I know that you (and all of the amazing community members here) are behind me!
Wayde here. I’ve been around for awhile.
fast.ai is truly a second home to me and it always feels like Christmas morning when I see that invite to the next course. I can’t think of a better community for deep/machine learning nor a better group of folks than Jeremy, Rachel, and Sylvain to learn from and work with. For me, more than anything, your work is a reminder of the value of being a good and selfless human being above all else.
Thanks again and looking to reconnect with all the folks I’ve had to pleasure of getting to know here these last 3-4 years.
-wg
Thank you!
Just checked twitter, cheers for the shoutout
Thank you
Oh yeah, Sam Esmail sure knows how to catch your eye. Shots like those really stand out.
Hey everyone,
I’m John. I’ve been following fastai since v1 and am looking forward to the latest iteration. My main focus has been on audio, and I made a blog post about a year ago with an experimental audio module for fastai: Audio Classification using FastAI and On-the-Fly Frequency Transforms
Lately I’ve been very interested in reinforcement learning, and how to apply deep learning techniques in more types of problems outside of classification. I’m also very interested in scaling deep learning to multi-gpu setups, and doing parallel processing on the CPU to complement DL for things like preprocessing.
Anyway, I’m really looking forward to fastaiv2 – it just keeps getting better and better!
John
So cool!!
As someone who started out in marine bio, I would have never have thought to do this!!! That is absolutely incredible to me