Skip to content
Snippets Groups Projects
Commit ec0579dc authored by Hashim Sharif's avatar Hashim Sharif
Browse files

Changing global_knobs.txt speedups for knobs

parent 6a9b519c
No related branches found
No related tags found
No related merge requests found
fp32,11 -1 1.0 tensorConvolution tensorConvolution
fp16,12 -1 1.5 tensorConvolution tensorHalfConvolution
perf,21 1,2,0 2.25 tensorConvolution tensorConvPerfCuda
perf,22 1,2,1 2.25 tensorConvolution tensorConvPerfCuda
perf,23 1,3,0 1.88 tensorConvolution tensorConvPerfCuda
perf,24 1,3,1 1.88 tensorConvolution tensorConvPerfCuda
perf,25 1,3,2 1.88 tensorConvolution tensorConvPerfCuda
perf,26 2,1,0 2.25 tensorConvolution tensorConvPerfCuda
perf,27 2,1,1 2.25 tensorConvolution tensorConvPerfCuda
perf,28 3,1,0 1.88 tensorConvolution tensorConvPerfCuda
perf,29 3,1,1 1.88 tensorConvolution tensorConvPerfCuda
perf,30 3,1,2 1.88 tensorConvolution tensorConvPerfCuda
samp,31 2,0 2.25 tensorConvolution tensorConvSampSim
samp,32 2,1 2.25 tensorConvolution tensorConvSampSim
samp,33 4,0 1.8 tensorConvolution tensorConvSampSim
samp,34 4,1 1.8 tensorConvolution tensorConvSampSim
samp,35 4,2 1.8 tensorConvolution tensorConvSampSim
samp,36 4,3 1.8 tensorConvolution tensorConvSampSim
perf,21 1,2,0 2.0 tensorConvolution tensorConvPerfCuda
perf,22 1,2,1 2.0 tensorConvolution tensorConvPerfCuda
perf,23 1,3,0 1.5 tensorConvolution tensorConvPerfCuda
perf,24 1,3,1 1.5 tensorConvolution tensorConvPerfCuda
perf,25 1,3,2 1.5 tensorConvolution tensorConvPerfCuda
perf,26 2,1,0 2.0 tensorConvolution tensorConvPerfCuda
perf,27 2,1,1 2.0 tensorConvolution tensorConvPerfCuda
perf,28 3,1,0 1.5 tensorConvolution tensorConvPerfCuda
perf,29 3,1,1 1.5 tensorConvolution tensorConvPerfCuda
perf,30 3,1,2 1.5 tensorConvolution tensorConvPerfCuda
samp,31 2,0 2.0 tensorConvolution tensorConvSampSim
samp,32 2,1 2.0 tensorConvolution tensorConvSampSim
samp,33 4,0 1.5 tensorConvolution tensorConvSampSim
samp,34 4,1 1.5 tensorConvolution tensorConvSampSim
samp,35 4,2 1.5 tensorConvolution tensorConvSampSim
samp,36 4,3 1.5 tensorConvolution tensorConvSampSim
reduction_samp,41 1 1.5
reduction_samp,42 1 2.25
reduction_samp,43 1 1.4
......
0% Loading or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment