undertherain / benchmarker

44%

master: 44%

LAST BUILD ON BRANCH gpus-issue-gh48

branch: gpus-issue-gh48

CHANGE BRANCH

x

Reset

• gpus-issue-gh48
• add-keras-applications-models
• attention
• auto-perf
• bert
• bert-large
• cblas
• cleanup
• coco-dataset
• coco-synthetic-gh34
• contributors
• conv2d_operations
• cosmoflow
• cublas
• cublas16
• cudnn
• cudnn-benchmark-param
• deepcam
• deeplab
• delete-old-ssd
• duplicate-imgnet-data-gh18
• edgetpu
• emil-WIP
• emil-ssd-TMP
• epochs_in_blas
• failing-tests-cpuinfo-issue-gh75
• fast_tests
• fix-ci
• fix-failing-tf-tests
• fix-fapp
• fix-framework-err
• fix-missing-samples-per-second
• fix-output-issue-gh66
• fix-time-logs
• fix-to-dense
• fix_bert
• fp16
• master
• mkldnn
• move_power
• ncf-WIP
• power
• precision
• random-seed-issue-gh65
• scripts
• sparse-categorical-crossentropy-issue-gh82
• ssd-downloader-issue-gh64
• ssd300
• testing
• tf-conv123d-WIP
• tf_precision
• torch-parallel-issue-gh47
• total_time
• tpu
• train_loss
• unparsed_args
• vis-scripts

Build # 502

Build Type

push

travis-ci

Committed by vatai

Commit Message

Remove unreachable code

The assertion above from one of the previous comment, ensures the GPU
specified in the params, and the GPUs "seen" by TF are the same.

Run Details

93 of 393 branches covered (23.66%)

Branch coverage included in aggregate %.

889 of 1822 relevant lines covered (48.79%)

0.49 hits per line