Hi! I find MeZO-adam code in medium size folder, but it uses the Adam from pytorch.optim. Its not like the case in large_models that author re-write the inner_loop. Can you please explain it? Thank you every much for your time and efforts.
We found in the medium model experiments that Adam is only comparable or worse compared to SGD. Hence we only ran it as an ablation (on the medium-sized models) and did not implement the efficient version.
Hi! I find MeZO-adam code in medium size folder, but it uses the Adam from pytorch.optim. Its not like the case in large_models that author re-write the inner_loop. Can you please explain it? Thank you every much for your time and efforts.