Re:Ask HN: Full-on machine learning for 2020, what are the b...
by throwlaplace on Tuesday December 31, 2019 @04:20PM
(#21924825)
> Implement their methods from scratch (i.e. numpy not pytorch)
lol this is basically impossible and completely pointless. please show me a numpy implementation of BERT or CycleGAN or deformable convolutions (note that jax != numpy). it's like suggesting implementing a kernel to someone who wants to learn about virtual memory or scheduling.
better advice would be take a paper and implement the model using pytorch without looking at their implementation and fiddle with that.
lol this is basically impossible and completely pointless. please show me a numpy implementation of BERT or CycleGAN or deformable convolutions (note that jax != numpy). it's like suggesting implementing a kernel to someone who wants to learn about virtual memory or scheduling.
better advice would be take a paper and implement the model using pytorch without looking at their implementation and fiddle with that.