Would attending "Full Stack Deep Learning Bootcamp" taught by Pieter Abbeel, Sergey Karayev, and Josh Tobin be worth it?

(Joshua Lelon Mitchell) #1

I stumbled across:


It looks enticing for sure (and, of course, it’s got some pretty big names…) but jeez it’s expensive. It sounds like it’d be worth it, but if there was high employer visibility somehow that’d make the decision a no-brainer for me.

Those on the job market can choose to take a certification exam, aimed to ensure that you are ready for deep learning engineer technical interviews. The exam covers pre-requisites as well as course content, and can be taken on your own time.

I don’t know how much weight the “certification” has or what I can do with it. Do they refer me or help me find a job? Do they put me in some pool where employers look at me? Or maybe put my name on a public list on their website vouching for me?

Do you folks think it’d be worth it? Thoughts?

(Even Oldridge) #2

I’m debating this as well. I’ve asked about pytorch deployment and haven’t gotten a response.

(Naveen Pandey) #3

I received my acceptance today and I am also debating whether I should attend it or not as the cost of attending is exorbitant. Any suggestions and thoughts welcome.

(Christina Young) #4

Very interesting boot camp, seems to be oriented towards productionizing deep learning models.
One cheaper alternative I think is the Google Cloud Machine Learning specialization on Coursera. The major limitation of this one is that it only teaches how to do it on Google Cloud using Google tools, of course.

(Ignacio Perez) #5

Same situation. It’s a long (and very expensive) trip from where I live, but it might actually be worth it. Hit me up if you’re going!

(Sanyam Bhutani) #6

I got the email of selection as well!

There are two things on my mind:

  1. The child in me wants to go and learn from the innovators of our field, meet new enthusiasts.

  2. The logical and poor, realisitic guy says:

  • Even if I meet them, I don’t have much to network with them. I’m not a good kaggler or even a noob. I haven’t implemented a single Paper to show anyone. I don’t have a super impressive github.
  • Money: Even if I go over there to learn, the lectures may be uploaded to YT later.
  • The money could be spent on hardware which is something that helps a lot.
  • I already have the privelage of being able to talk to one of the leaders of the Free AI World: Jeremy, on here. Use that platform instead.
  • Maybe one day later, some years down the road I would have accomplished something to be able to not ask questions to Jeremy but even discuss a cool project that even he would find cool. And so would Pieter Abbeel.
  • Also, I really want to spend that money on attending Fast.ai in person. If I’ll be spending that much amount.

There is still a fight between point 1. and 2. inside my head.


(Thomas) #7

Personally, I tend to do fall in your 2nd category.
But here are three things:

  • What do you specifically accomplish by going there vs. somewhere else? If you learn one great thing you won’t learn elsewhere, it probably is a good deal.
  • So this is “learning from popstars”. Is your inclination to go more because of the learning opportunity or more of the seeing popstars aspect? Either is OK, but probably the benchmark to look at are other things you might do of that type.
  • Remember Jeremy’s mantra (if I may call it so - it is close to my heart, so I might exaggerate his emphasis) that you really only learn by doing. Now how much serious deep learning can you do on site as opposed to “you start a model and check in x hours what it has done”.

Personally, I found meetups in my town to be quite rewarding and there is no shortage of enthusiasts there - and if you really want to get into a subject, you can do your own talk there. Because you really only learn by doing and then explaining what you are doing to someone else.



(WG) #8

Just got accepted and will be attending.

From the list of speakers, both TF and PyTorch look to both be well represented.

Most of the bootcamp seems structured around lifecycle and productionizing ML solutions. Topics that are both complimentary to what is learned here through the fastai course/community and useful regardless of chosen framework.

Will report back on the event, in particular its value as a follow-up to the fastai MOOC, once things get wrapped up in early August.


How was it? Technically/for networking/career?