@rshakoor12 other user mentioned greasing the groove, highly recommend thisโ like 8-10 sets of half of your total max pull-ups once you can do like 4-6
@nicoritschel interesting point, iโve only seen people highlighting output tokens/s but now i wanna read more about prompt eval speed across quantization typesโ
@daniel_c0deb0t what scale of data do you deal with? is the hard part processing raw data at scale or just storing a shit ton of it? any cool up and coming bioinfo frameworks for data stuff youโve seen or are you rolling your own stuff?