You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We should achieve 80% intel-caffe single node performance, and achieve same top-1(68%) and top-5(88%) error compare to GPU version on local mode and cluster mode.
The text was updated successfully, but these errors were encountered:
…d some testcases (intel-analytics#12)
* fix:refactor: getNumThreads return the omp_num_threads enabled.
And add some testcases for mkl settings, which will be converted to
4 seperated methods.
* fix: unsupported arguments will throw exception
* fix: magic numbers to final static variables and fix comments of native method by javah
* fix: private vars to public about MKL
Le-Zheng
pushed a commit
to Le-Zheng/BigDL
that referenced
this issue
Oct 20, 2021
Googlenet_V2 Model reference this paper.
A caffe model is here
We should achieve 80% intel-caffe single node performance, and achieve same top-1(68%) and top-5(88%) error compare to GPU version on local mode and cluster mode.
The text was updated successfully, but these errors were encountered: