Announcement

Collapse
No announcement yet.

PyTorch 1.8 Released With AMD ROCm Binaries

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • coder
    replied
    Originally posted by Sin2x View Post
    TensorFlow is slowly dying in the academic community. I would advise against using it, unless you have a very specific reason for it:

    https://paperswithcode.com/trends
    Wow, I'd gotten the sense that PyTorch was becoming more popular, but I didn't realize it was dominating!

    As for caffe2, isn't it part of PyTorch? I guess they're counting only what papers are directly using, but it makes me wonder whether libtorch or PyTorch always use it, or whether it's just an optional backend or something like that.

    Leave a comment:


  • Aeder
    replied
    Good to see support being added to different frameworks, too bad I still have no idea how to get ROCM to a working state in any distro that isn't the two supported ones. Has there been any news when it comes to packaging it?

    Leave a comment:


  • Paradigm Shifter
    replied
    Originally posted by Sin2x View Post

    TensorFlow is slowly dying in the academic community. I would advise against using it, unless you have a very specific reason for it:

    https://paperswithcode.com/trends
    Oh, that link pokes an open wound with a salt-covered stick. I had great "fun" recently examining some options already available for a particular "AI" task, and more of 11 options, only two had public code.

    Why is TensorFlow losing ground, though? Inferior speed? More difficult to code? I've never used PyTorch, and only tinkered around with TensorFlow...

    Leave a comment:


  • Sin2x
    replied
    Originally posted by plinkyplonky View Post
    I wasn't sure whether to go with Tensorflow or PyTorch. I hope that PyTorch supports AGPUs as well, like the laptop Zen+ CPUs with built-in GPU.
    TensorFlow is slowly dying in the academic community. I would advise against using it, unless you have a very specific reason for it:

    Papers With Code highlights trending Machine Learning research and the code to implement it.

    Leave a comment:


  • Paradigm Shifter
    replied
    Originally posted by coder View Post
    Nah, just curious.
    It's fun seeing patterns in things.

    ...

    Broader support for hardware on these things are always good.

    Leave a comment:


  • boffo
    replied
    Originally posted by bridgman View Post
    AFAIK tensorflow support for ROCm has been upstream for over a year now, with regular builds off the upstream source code.



    The "tensorflow-rocm" repo is now called "tensorflow-upstream" and used for ongoing development. We do build "tensorflow-rocm" docker images, however.

    Pytorch upstream support is relatively new though.
    Nice!

    Leave a comment:


  • coder
    replied
    Originally posted by Qaridarium
    i know you are not but really man you sound a little bit paranoid at this point...
    Nah, just curious.

    Leave a comment:


  • coder
    replied
    Originally posted by Steffo View Post
    To be honest: Your kind of your perception is a little creepy. I don't think, you do yourself a favour with selective perception which has basically no point.
    If we're being honest, it really was a funny coincidence to have two similar-sounding accounts post similar sentiments, right at the start of the thread, and with registration dates and activity level that are further closely aligned. I was mostly having a laugh about it, since it's almost too inconsequential to be nefarious.

    Originally posted by Steffo View Post
    I don't even have an AMD graphic card for reasons like power consumption and bad support for neural networks. - The latter is changing now, the former hasn't changed much. NVIDIA cards are still more efficient than AMD cards.
    RDNA and RDNA2 have both made a lot of progress on power-efficiency, at least for gaming. Meanwhile, Nvidia has been delivering sub-par efficiency improvements since Pascal, which have been masked by their addition of Tensor Cores. AMD introduced Matrix Cores in their M100, but we can only hope and wait for them to trickle down to more affordable GPUs.

    At work, I've had to use Team Green for the reasons you cite. We jumped on that bandwagon back in the Pascal era and haven't yet had cause to reconsider, though we're also dabbling with OpenVINO (Intel).

    Leave a comment:


  • coder
    replied
    Originally posted by boffo View Post
    LOL!!! You got the OCD eye!
    That's not even all. I further computed that your average post interval is 8.75 days, while Steffo's is 9.05 days. I wondered how close the ~7 months between your account registrations came to compensating for your different post counts.

    Leave a comment:


  • bridgman
    replied
    AFAIK tensorflow support for ROCm has been upstream for over a year now, with regular builds off the upstream source code.



    The "tensorflow-rocm" repo is now called "tensorflow-upstream" and used for ongoing development. We do build "tensorflow-rocm" docker images, however.

    Pytorch upstream support is relatively new though.

    Leave a comment:

Working...
X