[06x12] How to use your GPU for Machine Learning using CUDA.jl and Flux.jl (CUDA.jl 101 Part 3 of 3)

แชร์
ฝัง
  • เผยแพร่เมื่อ 9 พ.ย. 2024

ความคิดเห็น • 7

  • @leroywalton4348
    @leroywalton4348 2 ปีที่แล้ว +3

    Thank you so much for creating these videos.

    • @doggodotjl
      @doggodotjl  2 ปีที่แล้ว +1

      You're welcome! Thanks for watching!

  • @tubesteaknyouri
    @tubesteaknyouri ปีที่แล้ว +1

    Thank you for the videos. Very informative and easy to follow.
    I do have a suggestion for timing. @btime removes compilation time is useful for fast functions. @time is sufficient for slow functions (such as training a neural network). However, when using @time, you should run it twice and use the second run. During the first run, the function is compiled, but the compile time should not affect subsequent runs.

  • @fireji8449
    @fireji8449 2 ปีที่แล้ว

    thank you so much for your video, could you update some video about Distributed? i want use Julia to simulate turbulence, i don't know whether should use mpi or Distribute.

  • @laalbujhakkar
    @laalbujhakkar 2 ปีที่แล้ว +1

    I tried installing the env for one of the julia academy courses and it installed all of conda on my machine it’s about 5gb of install … lol

    • @MrDoraHo
      @MrDoraHo 2 ปีที่แล้ว

      it will just install the whole cuda kit and the library inside your julia lib lol

    • @laalbujhakkar
      @laalbujhakkar 2 ปีที่แล้ว

      @@MrDoraHo Yea, well that whole Julia nonsense is off my drive now. Got me 10GB back!