Choose specific values from dataset

I have the following dataset consisting of two columns, interrupted by random integers. How can I read the dataset (attached below) and keep only the first set of values (colored red in the picture) after each integer?

Answer 1

[Stephen23]

txt = fileread('demodata1.txt');
rgx = '^\s+\d+\s+(\S+)\s+(\S+)';
tkn = regexp(txt,rgx,'tokens','lineanchors');
mat = str2double(vertcat(tkn{:}))

mat = 4×2

0.3164 -0.3921 -0.1385 0.4998 0.4545 -0.2000 -0.1175 -0.1774

[Jorg Borren]

It works, thank you!