This is an example data-loader plugin that lets you view a TFRecord of Events (i.e., Tensorboard log files) in the Rerun Viewer. It uses the external data loader mechanism to add this capability to the viewer without modifying the viewer itself.
rerun_opens_tfevents.mp4
External data loaders are executables that are available to the Rerun Viewer via the PATH
variable, with a name that starts with rerun-loader-
.
This example is written in Python, and uses TensorFlow to read the files. The events are then logged to Rerun.
NOTE: Not all events are supported yet. Scalars, images, text, and tensors should work. Unsupported events are skipped.
The simplest option is just:
pip install rerun-sdk
Read this guide for more options.
The most robust way to install the plugin to your PATH
is using pipx.
If you don't have pipx
installed on your system, you can follow the official instructions here.
Now you can install the plugin to your PATH
using
pipx install git+https://github.com/rerun-io/rerun-loader-python-example-tfrecord.git
pipx ensurepath
Note: you can use the --python
argument to specify the Python interpreter to use with pipx.
On unix-like systems --python $(which python)
will use the currently active Python.
Make sure it's installed by running it from your terminal, which should output an error and usage description:
rerun-loader-tfrecord
usage: rerun-loader-tfrecord [-h] [--recording-id RECORDING_ID] filepath
rerun-loader-tfrecord: error: the following arguments are required: filepath
curl -OL https://github.com/rerun-io/rerun-loader-python-example-tfrecord/raw/main/events.tfevents.example
You can either first open the viewer, and then open the file from there using drag-and-drop or the menu>open… dialog, or you can open it directly from the terminal like:
rerun events.tfevents.example