4 ms·

I'm now in week 4 of the first round of this MOOC (here: http://course.fast.ai/ http://course.fast.ai/ ) and it's incredibly intuitive, highly recommended. It's

I'm now in week 4 of the first round of this MOOC (here: http://course.fast.ai/ http://course.fast.ai/ ) and it's incredibly intuitive, highly recommended. It's less about the mathematics, more about what has been proven to work, and based on these scores I've gotten quite high in the Kaggle competitions the classes go through (Cats & Dogs, State Farm, Titanic) The Jupyter noteboks are sometimes still quite messy and deserve some cleanup, but it's all open anyway, should submit a PR...

kriro 10y ago
Completely agree. I love the pragmatic nature of the course and the general attitude (anyone can do DL). It's probably the one MOOC I'd recommend to every developer or CS student. It's really mind blowing how good you can perform compared to what was state of the art not too long ago. Not quite finished with #1 either but can't wait for part 2. The pitch/promise for part 2 is that it'll basically take you to the bleeding edge of current (2017) research...that sounds rather exciting.
technics256 10y ago
How high have you gotten on dogs VS cats? Just curious, am submitting there as well.:)
- a_bonobo 10y ago
  I was #150 but moved down a bit, one guy I'm doing the course with got into the 90s (but I think moved down a bit too)
TACIXAT 10y ago
For the course setup they say to use an AWS instance. I have a machine with a decent Nvidia GPU in it. What software does it use for the deep learning? Do you use the Nvidia suite directly [1]? Or does Theano [2] have everything built in and you just need drivers? 1. https://developer.nvidia.com/deep-learning-software https://developer.nvidia.com/deep-learning-software 2. http://deeplearning.net/software/theano/index.html http://deeplearning.net/software/theano/index.html
- zitterbewegung 10y ago
  All of the popular deep learning libraries need cuda drivers with cudnn from nvidia which are free . This course uses theano and keras which are open source .
- jpuderer 10y ago
  If you've got a decent NVidia GPU, you should be alright. I understand that Theano can support other (non-NVidia) GPUs using OpenCL, but I haven't tried this. Part 2 of the course requires TensorFlow (which doesn't support OpenCL, AFAIK), so you're better off sticking with NVidia unfortunately.
- t0mk 10y ago
  if you are comfortable with Docker, check out https://github.com/t0mk/ultimate-mldev https://github.com/t0mk/ultimate-mldev
mcintyre1994 10y ago
Do you have to spend money on AWS for this course?
- BrandiATMuhkuh 10y ago
  Yes you do. But you can turn off your machine after you are done with your calculations. You probably end up with not more than $20 a month.
  - a_bonobo 10y ago
    That's a good guess! I spent nearly exactly $20 until week 4, including a whole Sunday of fiddling around with the Cats & Dogs dataset
- dminor 10y ago
  If you want you can just watch the videos. You won't learn quite as deeply but he goes over his solution to the homework.
- jpuderer 10y ago
  No. You can build your own machine, use Google Computer Engine (https://cloud.google.com/gpu/ https://cloud.google.com/gpu/), or even just stick with CPU while you get your feet wet. I started the course with CPU only, and then switched to a GCE instance (which I run sparingly) when I got tired of waiting for my models to train.
jpuderer 10y ago
I'm in Part 2 of the course now, and it's shaping up to be really really good as well.

kriro 10y ago

Completely agree. I love the pragmatic nature of the course and the general attitude (anyone can do DL). It's probably the one MOOC I'd recommend to every developer or CS student. It's really mind blowing how good you can perform compared to what was state of the art not too long ago. Not quite finished with #1 either but can't wait for part 2. The pitch/promise for part 2 is that it'll basically take you to the bleeding edge of current (2017) research...that sounds rather exciting.

technics256 10y ago

How high have you gotten on dogs VS cats? Just curious, am submitting there as well.:)

a_bonobo 10y ago
I was #150 but moved down a bit, one guy I'm doing the course with got into the 90s (but I think moved down a bit too)

TACIXAT 10y ago

For the course setup they say to use an AWS instance. I have a machine with a decent Nvidia GPU in it. What software does it use for the deep learning? Do you use the Nvidia suite directly [1]? Or does Theano [2] have everything built in and you just need drivers? 1. https://developer.nvidia.com/deep-learning-software https://developer.nvidia.com/deep-learning-software 2. http://deeplearning.net/software/theano/index.html http://deeplearning.net/software/theano/index.html

zitterbewegung 10y ago
All of the popular deep learning libraries need cuda drivers with cudnn from nvidia which are free . This course uses theano and keras which are open source .
jpuderer 10y ago
If you've got a decent NVidia GPU, you should be alright. I understand that Theano can support other (non-NVidia) GPUs using OpenCL, but I haven't tried this. Part 2 of the course requires TensorFlow (which doesn't support OpenCL, AFAIK), so you're better off sticking with NVidia unfortunately.
t0mk 10y ago
if you are comfortable with Docker, check out https://github.com/t0mk/ultimate-mldev https://github.com/t0mk/ultimate-mldev

mcintyre1994 10y ago

Do you have to spend money on AWS for this course?

BrandiATMuhkuh 10y ago
Yes you do. But you can turn off your machine after you are done with your calculations. You probably end up with not more than $20 a month.
- a_bonobo 10y ago
  That's a good guess! I spent nearly exactly $20 until week 4, including a whole Sunday of fiddling around with the Cats & Dogs dataset
dminor 10y ago
If you want you can just watch the videos. You won't learn quite as deeply but he goes over his solution to the homework.
jpuderer 10y ago
No. You can build your own machine, use Google Computer Engine (https://cloud.google.com/gpu/ https://cloud.google.com/gpu/), or even just stick with CPU while you get your feet wet. I started the course with CPU only, and then switched to a GCE instance (which I run sparingly) when I got tired of waiting for my models to train.

jpuderer 10y ago

I'm in Part 2 of the course now, and it's shaping up to be really really good as well.