LLH value in <mcmc_samples.txt> seems differ in different arch

[go-ahead-maker]

Hello,
Recently I work on the phylowgs and observed a weird phenomenon.
When the command python2 multievolve.py --num-chains 2 --ssms ssm_data.txt --cnvs cnv_data.txt --random-seeds 3407 3407 || return 1 is performed, I got different LLH results in file mcmc_samples.txt on arm and x86 arch.
The difference starts in the eighth decimal place, is this caused by the different system arch?

Yours,