1. Hey, TensorFlow community:
Do you think there is value in something like Learn TensorFlow The Hard Way (cc: @zedshaw)
-
Show this thread
-
2. I don't mean yet another: How to build a MNIST classifier. People using tf generally understand backprop, layer types, & architectures.
1 reply 0 retweets 0 likesShow this thread -
Replying to @generativist
What's the hard way? Implementing multi-GPU training??
2 replies 0 retweets 0 likes -
Replying to @drenerbas
Eventually, sure. But, more like Zed's LCTHW-style. Lots and lots and lots of simple examples, not all ANN.
1 reply 0 retweets 0 likes -
Replying to @generativist
GPFlow kind of stuff? Solving engineering pde's with tf? Cool!!
1 reply 1 retweet 1 like -
Replying to @drenerbas
Yes! It's cool. The pure focus on ANN is pedagogically bad (IMO). You can do lots of stuff in TF. It's not just an ANN hammer.
1 reply 0 retweets 0 likes -
Replying to @generativist
Yes when I first came across Theano years ago I was trying to speed up some matrix differential equations. Neural nets seemed like a pivot.
2 replies 0 retweets 2 likes
Yep! My thread was largely self-criticism, too. I wrote some code that did work out-of-graph because I didn't know ops for in-graph.
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.