https://github.com/Microsoft/CNTK
Revision 19719a656bba269ace4c3b3bdee85119d26d4ae1 authored by KeDengMS on 09 February 2018, 06:22:27 UTC, committed by KeDengMS on 09 February 2018, 06:22:27 UTC
- Accelerates some common tensor ops in Intel CPU inference for float32, especially for fully connected networks
- Can be turned on/off by cntk.cntk_py.enable_cpueval_optimization()/cntk.cntk_py.disable_cpueval_optimization()
1 parent 08bcbdc
History
Tip revision: 19719a656bba269ace4c3b3bdee85119d26d4ae1 authored by KeDengMS on 09 February 2018, 06:22:27 UTC
Fast tensor ops using MKL
Tip revision: 19719a6

back to top