Tensorboard initializer
The new Tensorboard version requires Polyaxon CE or Polyaxon Agent with a version higher than v1.18.
This new version can be used both by single runs and multiple runs and it leverages a new Tensorboard initializer.
Starting Tensorboard from the UI
For single runs:
As soon as the tensorboard is created it will show in the sidebar:
For multiple runs:
Use run’s names to initialize the Tensorboard logs, to enable run names:
...
params:
...
user_names:
value: true
...Or using the CLI:
polyaxon run --hub tensorboard:v2 -P user_names=true -P ...Pre-installing plugins:
...
params:
...
plugins:
value: [tensorboard-plugin-profile, another-plugin, ...]
...Or using the CLI:
polyaxon run --hub tensorboard:v2 -P plugins=tensorboard-plugin-profile,another-plugin -P ...Usage with single runs:
...
params:
...
uuid:
value: UUID
...Or using the CLI:
polyaxon run --hub tensorboard:v2 -P uuid=UUIDUsage with multiple runs:
...
params:
...
uuids:
value: [UUID1, UUID2, ...]
...Or using the CLI:
polyaxon run --hub tensorboard:v2 -P uuids=UUID1,UUID2,...Learn More about Polyaxon
This blog post just goes over a couple of features that we shipped in v1.18. To learn more about all the features, fixes, and enhancements, please visit the release notes and the short term roadmap.
Polyaxon continues to grow quickly and keeps improving and providing the simplest machine learning abstraction. We hope that these updates will improve your workflows and increase your productivity, and again, thank you for your continued feedback and support.
