search

accosmin / nano

C++ library [machine learning & numerical optimization] - superseeded by libnano
MIT License
1 stars 0 forks source link

Benchmark trainers with varying number of threads #117

Closed accosmin closed 8 years ago

accosmin commented 8 years ago

Goal: assess if most processing time is spend in accumulator_t

accosmin commented 8 years ago

BUG: The training does not scale nicely with the number of threads