Homepage
Close
Menu

Site Navigation

  • Home
  • Archive(TODO)
    • By Day
    • By Month
  • About(TODO)
  • Stats
Close
by jeremyphoward on 2020-01-24 (UTC).

Interesting analysis suggesting that the reason for the disappointing performance of many modern CNN architectures is that their depthwise convolutions are memory-bound. https://t.co/C55Q6tv3LN pic.twitter.com/nXafyOseH3

— Jeremy Howard (@jeremyphoward) January 24, 2020
misclearning
by jeremyphoward on 2020-01-25 (UTC).

Fascinating. I'm really surprised Google is using depthwise convs so much in their research, when it performs so badly on their TPUs. https://t.co/nXqtrbSWzM pic.twitter.com/f51mI4LPEq

— Jeremy Howard (@jeremyphoward) January 25, 2020
misclearning
by jeremyphoward on 2020-01-25 (UTC).

Wow this is obvious in hindsight but I'd never thought of it! Seems like a big issue... https://t.co/RQLfRfrlif

— Jeremy Howard (@jeremyphoward) January 25, 2020
misc

Tags

learning tutorial misc nlp rstats gan ethics research dataviz survey python tool security kaggle video thought bayesian humour tensorflow w_code bias dataset pytorch cv tip application javascript forecast swift golang rl jax julia gnn causal surey diffusion
© Copyright Philosophy 2018 Site Template by Colorlib