Hacker Blogs
Work in progress: 181,232 of 181,233 sites checked so far, more every week · updated 2 Nov 2026Missing a blog? Tell us

Top all-time · #12565

L

Lernapparat recent posts

Personal blog · lernapparat.deaiprogramming

Lernapparat recent posts is a personal blog at lernapparat.de about ai and programming. Since 2018, 6 different Hacker News readers have linked it 10 times, most often “Optimizing models using the PyTorch JIT”. It has a working feed you can follow.

Visit the blog
10mentions?
6HN people?
2018–2022first – last linked?
Links on Hacker News per year · peak 6 in 2020

Too few sampled comments (3) to say how often HN recommends it.?

Why HN links it

HN readers link this blog for its recipe on compiling PyTorch against AMD's ROCm stack on Debian, with no Docker needed. The posts are practical build instructions for running PyTorch on Radeon GPUs.

Summarised by Claude from a sample of 3 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.

Most-linked posts?

  1. Optimizing models using the PyTorch JIT
    3 people · 3 mentions · first linked 2020-12-22
  2. Fast LSTMs in PyTorch
    2 people · 2 mentions · first linked 2019-03-18

What Hacker News says?

  1. t-vi · 2020-08-01
    PyTorch supports ROCm quite well if you can follow three step build instructions instructions ( https://lernapparat.de/pytorch-rocm/ ). TVM.ai also supports ROCm well. TF I didn't check. FP32 performance for the VII is comparable to GTX20280Ti e.g. on Resnet training. Tensor Cores are cool, but 90% of people don't use FP16 that much. I'm not sure that is the best reply that stackoverflow has ever seen.

    Comment linking /pytorch-rocm/

    on “Nvidia is reportedly in ‘advanced talks’ to buy ARM for more than $32B”
  2. homarp · 2019-03-19

    Submitted Fast LSTMs in PyTorch as a story · 34 points

    HN discussion
  3. t-vi · 2020-10-28
    I think PyTorch support is decent, I use it on my Radeon VII a lot. - You can compile on https://lernapparat.de/pytorch-rocm/ (disclaimer my own link) - Arch has support out of the box: https://aur.archlinux.org/packages/python-pytorch-rocm/ - AMD also has a ROCm docker with PyTorch ( https://rocmdocs.amd.com/en/latest/Deep_learning/Deep-learni... ) - If you know the secret URL (which you can infer from the other UR…

    Comment linking /pytorch-rocm/

    on “AMD Reveals the Radeon RX 6000 Series, Coming November 18th”
  4. ghosthamlet · 2019-03-18

    Submitted Fast LSTMs in PyTorch as a story · 4 points

    HN discussion
  5. t-vi · 2020-04-25
    You can compile PyTorch this pretty easily, I wrote a recipe here: https://lernapparat.de/pytorch-rocm/ No docker involved, just Debian + the ROCm packages from their repository. I'm doing this relatively regularly, and it has worked well for me for the past half year or so. I guess I could upload wheels somewhere, but I'm never sure how tight the dependency on ROCm has to be. Over the last release cycle, PyTorch/RO…

    Comment linking /pytorch-rocm/

    on “PyTorch 1.5”
  6. t-vi · 2020-12-22

    Submitted Optimizing models using the PyTorch JIT as a story · 3 points

    HN discussion
  7. naillo · 2022-09-12

    Submitted Optimizing models using the PyTorch JIT as a story · 2 points

    HN discussion
  8. homarp · 2020-07-14

    Submitted Bridging PyTorch and TVM as a story · 2 points

    HN discussion
  9. sebg · 2018-08-01

    Submitted A selective excursion into the internals of PyTorch as a story · 2 points

    HN discussion
  10. briggers · 2020-12-28

    Submitted Optimizing models using the PyTorch JIT as a story · 1 points

    HN discussion

Search all of HN for lernapparat.de