Training material seems to be especially rich re GPU kernels and SIMD.
I wonder if there is extra effort put into this because they are useful for the researchers working on the models or just a sub-domain that language models are a great fit for and humans have trouble with?
I wonder if there is extra effort put into this because they are useful for the researchers working on the models or just a sub-domain that language models are a great fit for and humans have trouble with?