Skip to content
This repository was archived by the owner on Dec 9, 2024. It is now read-only.
This repository was archived by the owner on Dec 9, 2024. It is now read-only.

The data visualized by tensorboard is all 0 #551

Description

@nordelund

I used Perfzero to perform tensorflow performance testing, and used Dockerfile Dockerfile_ubuntu_2204_tf_cuda_12 to build the container used for the test. The tensorflow version is 2.18.0-dev20240722, the Python version is python 3.11, and the dataset used is CIFAR-10

The Benchmark command is as follows, specifying --profiler_enabled_time

python3.11 perfzero/lib/benchmark.py \
--git_repos="https://github.com/tensorflow/models.git;benchmark" \
--python_path=models \
--benchmark_methods=models.official.benchmark.keras_cifar_benchmark.Resnet56KerasBenchmarkReal.benchmark_1_gpu_no_dist_strat \
--profiler_enabled_time=1

After that, I used TensorBoard to visualize the profiler data, but the display was abnormal. Is there something wrong with my step?
image

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions