etam89

Results 3 issues of etam89

You put flops for inference and training. FLOPS are Floating point operations per second. should the unit be GLOP/TOP without the "S" ? With "S", those number means operation per...

good job! Is there a way to change plot size? maybe a parameter one can set? Thanks!

can these python code uses trained model from Meta? I meant, is there a way to extract parameters from Meta's Llama 2 model and plug into this code to run...