Collective Knowledge Aggregator proof-of-concept
Crowd results Raw CK browser Graphs Reports Datasets Models Home

This page is outdated! New version is available here.


Share
1099e435bb397e85
Date: 2017-05-16 14:28:49.599114
License:
Module: experiment.bench.dnn
Repo: upload
CID (DOI-like distributed ID): 1eb2f50d4620903e:1099e435bb397e85

DNN engine and model evaluation statistics per layer (crowd-tuning)


DNN engine name: BVLC Caffe framework (cpu)
DNN engine version: trunk-eeebdab
DNN engine type: cpu
DNN engine dependencies: cmake=3.5.1
compiler=5.4.0
lib-blas=0.2.19-85636ff
lib-boost=1.62.0
lib-gflags=2.2.0
lib-glog=development-77a0e36
lib-hdf5=1.10.0
lib-lmdb=development-4d21543
lib-opencv=3.2.0
lib-protobuf-host=3.1.0
DNN model name: Caffe model (net and weights) (bvlc, alexnet)
DNN model version: trunk
Batch size: 10
Platform: Microsoft Corporation 7.0 (Virtual Machine)
OS: Ubuntu 16.04.2 LTS
CPU: Intel(R) Xeon(R) CPU E5-2673 v3 @ 2.40GHz
GPU: Microsoft Corporation Hyper-V virtual VGA

Name Direction Min time (ms.): Expected time (ms.): Max time (ms.): Repetitions:
03: norm1 backward 130.1 136.0 139.7 3
03: norm1 forward 119.5 124.8 130.0 3
07: norm2 backward 84.4 84.7 85.0 3
07: norm2 forward 77.7 83.2 86.8 3
16: fc6 forward 65.2 68.0 71.5 3
05: conv2 forward 49.0 50.8 54.7 3
05: conv2 backward 47.8 49.3 50.5 3
19: fc7 forward 31.3 31.8 32.1 3
16: fc6 backward 29.9 31.7 34.1 3
09: conv3 forward 29.0 33.1 35.8 3
01: conv1 backward 28.1 30.9 36.7 3
11: conv4 forward 27.6 28.3 28.7 3
01: conv1 forward 26.5 26.8 27.5 3
09: conv3 backward 23.1 25.5 32.1 3
13: conv5 backward 22.2 23.8 26.8 3
11: conv4 backward 20.9 23.4 32.4 3
13: conv5 forward 20.4 22.6 25.6 3
04: pool1 forward 15.1 16.0 17.3 3
19: fc7 backward 14.2 16.1 17.4 3
08: pool2 forward 9.8 11.8 14.2 3
22: fc8 forward 7.2 7.4 7.5 3
22: fc8 backward 6.8 12.3 15.5 3
02: relu1 forward 3.5 3.7 3.9 3
15: pool5 forward 2.6 2.7 2.9 3
06: relu2 forward 2.2 2.3 2.6 3
10: relu3 forward 0.8 0.8 0.8 3
12: relu4 forward 0.7 0.8 0.9 3
14: relu5 forward 0.5 0.5 0.5 3
21: drop7 forward 0.3 0.3 0.3 3
23: prob forward 0.3 0.3 0.3 3
18: drop6 forward 0.2 0.3 0.3 3
23: prob backward 0 0 0 3
17: relu6 forward 0 0 0 3
20: relu7 forward 0 0 0.4 3
00: input forward 0 0 0 3
00: input backward 0 0 0 3
06: relu2 backward 0 0 0 3
08: pool2 backward 0 0 0 3
12: relu4 backward 0 0 0 3
15: pool5 backward 0 0 0 3
18: drop6 backward 0 0 0 3
02: relu1 backward 0 0 0 3
04: pool1 backward 0 0 0 3
10: relu3 backward 0 0 0 3
14: relu5 backward 0 0 0 3
17: relu6 backward 0 0 0 3
20: relu7 backward 0 0 0 3
21: drop7 backward 0 0 0 3
View entry in raw format

Developed by Grigori Fursin           
Implemented as a CK workflow 		                         
   
                      Hosted at