4 ms·

Wouldn't we rather trend model memory requirements down? This way models/devices meet in the middle.

• brucethemoose2 3y ago
  The thing about ML is that memory helps everything. Take almost any given model that fits in X memory pool for some task. Double that, and everything gets better. Training quality goes up, inference quality potentially goes up, you gain the capacity for more batching which dramatically increases performance. Maybe more caching, depending on the app. You can add more models to the pipeline without much fuss.
  • eachro 3y ago
    why is this? I would have thought that more memory would help up to some point (wherever the entire model can fit in RAM) and then no gains after that. I'd also have thought that things like accuracy wouldnt be affected.
    • brucethemoose2 3y ago
      Every model and ecosystem has different compromises, and obviously small models hit diminishing returns quick. But the big models that are in fashion now... GPUs are basically always memory capacity limited to some extent.