From 0e3a1d7b7741cbaf28230d9e985a74d37fd51676 Mon Sep 17 00:00:00 2001 From: "Sergey E. Koposov" Date: Sun, 21 Jul 2024 19:20:33 +0100 Subject: [PATCH] update readme --- README.md | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/README.md b/README.md index d757f6e..83cf60f 100644 --- a/README.md +++ b/README.md @@ -86,7 +86,9 @@ This requires running ``` $ rvs_train_nn_interpolator --dir ${PREFIX}/ --setup myconf ``` - +instead of rvs_make_nd. +When doing the fitting, you may also set this +environmental variable RVS_NN_DEVICE=cpu to force the code to run on cpu. * Making the cross-correlation files @@ -231,7 +233,8 @@ To run on DESI data use rvs_desi_fit or rvs_weave_fit By default rvspecfit uses the multi-dimensional linear interpolation with Delaunay triangulation. It also supports the multilinear interpolation if -the input grid is organized as a true n-D grid without gaps +the input grid is organized as a true n-D grid without gaps. +The alternative interpolation method is Neural Network based. ## Other template libraries