Hi, I was wondering if it would be possible to save the compiled GradientTape that is used in ReverseDiff.jl. Using the gradient example from the package, I’d like to save compiled_f_tape # some objective function to work with f(a, b) = sum(a' * b + a * b') # pre-record a GradientTape for `f` …

Possible to save compiled GradientTape after compile in ReverseDiff

xxxxx March 8, 2023, 12:24am 7

The number of parameters is around 50,000, and the size of the data is about 20M.

Topic		Replies	Views
ReverseDiff.GradientTapes for functions of the form x -> f(x, y) when gradient is to be calculated wrt x only? General Usage	2	822	November 4, 2019
ReverseDiff Incorrect Result Numerics	5	596	July 19, 2018
ReverseDiff resultst do not fit theory Numerics	6	1249	June 8, 2018
ReverseDiff.jl Community package , announcement	9	1737	December 13, 2017
Unexpected performance mismatch in gradients for "compiled-tape-in-tape" experiment Performance autodiff	0	191	June 25, 2023