Episode 13: Jonathan Frankle, MIT, on the lottery ticket hypothesis and the science of deep learning
Generally Intelligent
00:00
Is There Any Scale Set for Machine Learning?
I have never regretted doing good softor engineering, not once. Undergond that scale set is my secret weapon. I can build experimental frameworks such that i can manage a hundred and trained and pruned networks. That's the reason i'm able to do this kind of empirical work. Any student whom i might hire to work with at some point if i'm a professor, that's a scale stet that i'm going to need to learn how to teach to graduate students.
Transcript
Play full episode
