AUTHOR: Chao Xiong fancysimon@gmail.com
python lda.py -a 0.01 -b 0.1 -k 2 --train_name=train2.txt --compute_likelihood
--total_iterations=100 --burn_in_iterations=70
mpiexec -n 2 -hosts "192.168.1.2,192.168.1.3" python mpi_lda.py -a 0.01 -b 0.1
--train_name=train2.txt --compute_likelihood --total_iterations=100
--burn_in_iterations=70
- Check point for mpi.
- Comment.
- Asymmetric alpha(document topic distribution prior).
- Accumulative model for MPI.