Skip to content
GitLab
Explore
Sign in
Primary navigation
Search or go to…
Project
H
hpvm-release
Manage
Activity
Members
Labels
Plan
Issues
Issue boards
Milestones
Wiki
Code
Merge requests
Repository
Branches
Commits
Tags
Repository graph
Compare revisions
Snippets
Build
Pipelines
Jobs
Pipeline schedules
Artifacts
Deploy
Releases
Model registry
Operate
Environments
Monitor
Incidents
Analyze
Value stream analytics
Contributor analytics
CI/CD analytics
Repository analytics
Model experiments
Help
Help
Support
GitLab documentation
Compare GitLab plans
Community forum
Contribute to GitLab
Provide feedback
Keyboard shortcuts
?
Snippets
Groups
Projects
Show more breadcrumbs
llvm
hpvm-release
Commits
f8577472
Commit
f8577472
authored
5 years ago
by
hsharif3
Browse files
Options
Downloads
Patches
Plain Diff
Adding pareto confs for LeNet
parent
d79d8707
No related branches found
Branches containing commit
No related tags found
Tags containing commit
No related merge requests found
Changes
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
llvm/test/VISC/DNN_Benchmarks/benchmarks/lenet/data/autotuner_data/tuner_pareto_confs_batch220.txt
+904
-0
904 additions, 0 deletions
...lenet/data/autotuner_data/tuner_pareto_confs_batch220.txt
with
904 additions
and
0 deletions
llvm/test/VISC/DNN_Benchmarks/benchmarks/lenet/data/autotuner_data/tuner_pareto_confs_batch220.txt
0 → 100644
+
904
−
0
View file @
f8577472
+++++
conf1 1 0 99.69 0
1 gpu conv fp32 1 add fp32 1 pool_max fp32 1 tanh fp32 1
2 gpu conv fp32 1 add fp32 1 pool_max fp32 1 tanh fp32 1
3 gpu mul fp32 1 add fp32 1 tanh fp32 1
4 gpu mul fp32 1 add fp32 1 tanh fp32 1
5 gpu softmax fp32 1
-----
+++++
conf1 2.01610051566 0 99.400002 0.6899979999999971
1 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf2 2.01610051566 0 99.040001 0.974998499999991
1 gpu conv perf 26 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf3 2.00016617632 0 99.68 0.4099999999999909
1 gpu conv perf 23 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf4 2.00016617632 0 99.660004 0.42999599999999705
1 gpu conv perf 29 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf5 1.97610564729 0 99.599998 0.4900019999999984
1 gpu conv fp16 1 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf6 2.00016617632 0 99.599998 0.4900019999999984
1 gpu conv perf 25 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf7 2.00016617632 0 99.080002 0.9149970000000067
1 gpu conv perf 30 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf8 2.00016617632 0 99.239998 0.6750029999999967
1 gpu conv perf 30 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf9 2.00016617632 0 99.199997 0.7350045000000023
1 gpu conv perf 28 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf10 1.99590274244 0 99.099998 0.8850029999999975
1 gpu conv samp 36 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf11 2.01610051566 0 99.559998 0.5300020000000046
1 gpu conv perf 27 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf12 1.99590274244 0 99.540001 0.549998999999994
1 gpu conv samp 33 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf13 2.00016617632 0 99.639999 0.45000099999999466
1 gpu conv perf 30 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf14 1.99590274244 0 99.580002 0.5099980000000045
1 gpu conv samp 33 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf15 2.01610051566 0 99.099998 0.8850029999999975
1 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf16 2.01610051566 0 99.160004 0.7949939999999955
1 gpu conv perf 27 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf17 2.00016617632 0 99.379997 0.46500449999999205
1 gpu conv perf 29 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf18 1.99590274244 0 99.639999 0.45000099999999466
1 gpu conv samp 36 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf19 2.01610051566 0 99.580002 0.5099980000000045
1 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf20 1.97610564729 0 99.660004 0.42999599999999705
1 gpu conv fp16 1 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf21 1.99590274244 0 99.440002 0.6499979999999909
1 gpu conv samp 33 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf22 1.99590274244 0 99.260002 0.6449969999999965
1 gpu conv samp 36 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf23 2.00016617632 0 99.360001 0.49499850000000123
1 gpu conv perf 23 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf24 2.01610051566 0 99.32 0.5550000000000068
1 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf25 2.00016617632 0 99.519997 0.5700029999999942
1 gpu conv perf 30 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf26 1.97610564729 0 99.379997 0.46500449999999205
1 gpu conv fp16 1 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf27 2.01610051566 0 99.68 0.4099999999999909
1 gpu conv perf 27 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf28 2.00016617632 0 99.559998 0.5300020000000046
1 gpu conv perf 23 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf29 2.00016617632 0 99.080002 0.9149970000000067
1 gpu conv perf 23 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf30 1.97610564729 0 99.660004 0.42999599999999705
1 gpu conv fp16 1 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf31 2.01610051566 0 99.599998 0.4900019999999984
1 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf32 1.97610564729 0 99.080002 0.9149970000000067
1 gpu conv fp16 1 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf33 2.01610051566 0 99.620003 0.4699970000000008
1 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf34 2.00016617632 0 99.620003 0.4699970000000008
1 gpu conv perf 28 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf35 2.00016617632 0 99.599998 0.4900019999999984
1 gpu conv perf 25 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf36 1.99590274244 0 99.599998 0.4900019999999984
1 gpu conv samp 36 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf37 2.01610051566 0 99.540001 0.549998999999994
1 gpu conv perf 26 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf38 2.00016617632 0 99.339996 0.5250059999999976
1 gpu conv perf 25 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf39 2.00016617632 0 99.599998 0.4900019999999984
1 gpu conv perf 24 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf40 1.97610564729 0 99.379997 0.46500449999999205
1 gpu conv fp16 1 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf41 2.00016617632 0 99.559998 0.5300020000000046
1 gpu conv perf 28 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf42 1.99590274244 0 99.459999 0.6300010000000015
1 gpu conv samp 34 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf43 1.99590274244 0 99.400002 0.6899979999999971
1 gpu conv samp 34 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf44 2.00016617632 0 99.599998 0.4900019999999984
1 gpu conv perf 29 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf45 2.01610051566 0 99.599998 0.4900019999999984
1 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf46 2.01610051566 0 99.080002 0.9149970000000067
1 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf47 2.01610051566 0 99.660004 0.42999599999999705
1 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf48 2.00016617632 0 99.639999 0.45000099999999466
1 gpu conv perf 24 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf49 2.01610051566 0 99.480003 0.6099970000000013
1 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf50 2.00016617632 0 98.400002 1.9349969999999956
1 gpu conv perf 23 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 26 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf51 2.01610051566 0 98.540001 1.724998499999991
1 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 26 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf52 2.01610051566 0 99.080002 0.9149970000000067
1 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf53 2.00016617632 0 99.660004 0.42999599999999705
1 gpu conv perf 29 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf54 2.01610051566 0 99.660004 0.42999599999999705
1 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf55 1.97610564729 0 99.599998 0.4900019999999984
1 gpu conv fp16 1 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf56 2.01610051566 0 98.900002 1.1849969999999956
1 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf57 1.99590274244 0 99.099998 0.8850029999999975
1 gpu conv samp 36 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf58 2.01610051566 0 99.580002 0.5099980000000045
1 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf59 1.97610564729 0 99.080002 0.9149970000000067
1 gpu conv fp16 1 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf60 2.01610051566 0 98.959999 1.0950015000000022
1 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf61 2.01610051566 0 99.220001 0.7049985000000021
1 gpu conv perf 26 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf62 2.01610051566 0 98.839996 1.2750059999999976
1 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf63 1.99590274244 0 98.940002 1.1249969999999863
1 gpu conv samp 34 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf64 1.97610564729 0 99.379997 0.46500449999999205
1 gpu conv fp16 1 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf65 2.00016617632 0 99.559998 0.5300020000000046
1 gpu conv perf 23 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf66 2.00016617632 0 99.239998 0.6750029999999967
1 gpu conv perf 30 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf67 2.01610051566 0 99.459999 0.6300010000000015
1 gpu conv perf 26 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf68 2.00016617632 0 99.360001 0.49499850000000123
1 gpu conv perf 24 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf69 2.01610051566 0 99.559998 0.5300020000000046
1 gpu conv perf 27 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf70 1.99590274244 0 99.440002 0.6499979999999909
1 gpu conv samp 33 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf71 2.00016617632 0 99.339996 0.5250059999999976
1 gpu conv perf 25 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf72 2.01610051566 0 99.32 0.5550000000000068
1 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf73 1.97610564729 0 99.379997 0.46500449999999205
1 gpu conv fp16 1 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf74 2.00016617632 0 99.019997 1.0050044999999912
1 gpu conv perf 29 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf75 1.99590274244 0 99.260002 0.6449969999999965
1 gpu conv samp 36 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf76 2.01610051566 0 99.099998 0.8850029999999975
1 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf77 1.97610564729 0 98.440002 1.8749969999999863
1 gpu conv fp16 1 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 27 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf78 2.01610051566 0 98.440002 1.8749969999999863
1 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 27 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf79 2.01610051566 0 99.160004 0.7949939999999955
1 gpu conv perf 27 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf80 1.97610564729 0 98.480003 1.814995500000002
1 gpu conv fp16 1 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 26 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf81 2.00016617632 0 99.360001 0.49499850000000123
1 gpu conv perf 23 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf82 1.97610564729 0 99.660004 0.42999599999999705
1 gpu conv fp16 1 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf83 1.99590274244 0 99.540001 0.549998999999994
1 gpu conv samp 33 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf84 2.00016617632 0 99.199997 0.7350045000000023
1 gpu conv perf 28 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf85 1.97610564729 0 98.440002 1.8749969999999863
1 gpu conv fp16 1 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 27 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf86 2.00016617632 0 99.0 1.0349999999999966
1 gpu conv perf 28 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf87 1.99590274244 0 98.519997 1.7550044999999912
1 gpu conv samp 35 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf88 2.01610051566 0 99.400002 0.6899979999999971
1 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf89 2.01610051566 0 97.760002 2.8949969999999965
1 gpu conv perf 26 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 26 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf90 2.01610051566 0 99.519997 0.5700029999999942
1 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf91 2.01610051566 0 99.32 0.5550000000000068
1 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf92 2.01610051566 0 99.580002 0.5099980000000045
1 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf93 2.01610051566 0 99.480003 0.6099970000000013
1 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf94 2.01610051566 0 98.480003 1.814995500000002
1 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf95 2.01610051566 0 98.540001 1.724998499999991
1 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 26 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf96 2.01610051566 0 97.82 2.805000000000007
1 gpu conv perf 27 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 26 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf97 2.01610051566 0 98.959999 1.0950015000000022
1 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf98 2.01610051566 0 98.459999 1.8450015000000022
1 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 26 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf99 2.01610051566 0 99.660004 0.42999599999999705
1 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf100 2.01610051566 0 99.620003 0.4699970000000008
1 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf101 2.01610051566 0 97.699997 2.9850045000000023
1 gpu conv perf 27 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 27 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf102 2.01610051566 0 99.040001 0.974998499999991
1 gpu conv perf 26 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf103 2.01610051566 0 98.0 2.5349999999999966
1 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 27 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf104 2.01610051566 0 99.160004 0.7949939999999955
1 gpu conv perf 27 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf105 2.01610051566 0 99.540001 0.549998999999994
1 gpu conv perf 26 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf106 2.01610051566 0 99.519997 0.5700029999999942
1 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf107 2.01610051566 0 99.099998 0.8850029999999975
1 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 21 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf108 2.01610051566 0 98.120003 2.354995500000001
1 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 26 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf109 2.01610051566 0 99.459999 0.6300010000000015
1 gpu conv perf 26 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 31 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf110 2.01610051566 0 99.68 0.4099999999999909
1 gpu conv perf 27 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf111 2.01610051566 0 98.839996 1.2750059999999976
1 gpu conv samp 32 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
+++++
conf112 2.01610051566 0 98.18 2.2649999999999864
1 gpu conv perf 22 add fp16 1 tanh fp16 1 pool_max fp16 1
2 gpu conv perf 27 add fp16 1 tanh fp16 1 pool_max fp16 1
3 gpu mul fp16 1 add fp16 1 tanh fp16 1
4 gpu mul fp16 1 add fp16 1 tanh fp16 1
5 gpu softmax fp32 1
-----
This diff is collapsed.
Click to expand it.
Preview
0%
Loading
Try again
or
attach a new file
.
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Save comment
Cancel
Please
register
or
sign in
to comment
