Is your feature request related to a problem? Please describe.
When working with multiple experiments in Visdom, it becomes difficult to organize and keep track of them. There is currently no simple way to group or categorize experiments based on their purpose or configuration.
Describe the solution you'd like
I would like to propose adding support for experiment-level tagging.
This would allow users to:
- Assign tags (e.g., cnn, resnet, baseline) to environments
- View tags directly in the UI
- Filter experiments based on tags
Tags can be stored along with environment data and exposed through simple API endpoints.
Describe alternatives you've considered
Currently, users rely on naming conventions to organize experiments, but this is not flexible and becomes hard to manage as the number of experiments grows.
Additional context
I have already explored an initial implementation for tagging (UI, backend, and API) and would be happy to refine it based on feedback and align it with the project’s direction.
Is your feature request related to a problem? Please describe.
When working with multiple experiments in Visdom, it becomes difficult to organize and keep track of them. There is currently no simple way to group or categorize experiments based on their purpose or configuration.
Describe the solution you'd like
I would like to propose adding support for experiment-level tagging.
This would allow users to:
Tags can be stored along with environment data and exposed through simple API endpoints.
Describe alternatives you've considered
Currently, users rely on naming conventions to organize experiments, but this is not flexible and becomes hard to manage as the number of experiments grows.
Additional context
I have already explored an initial implementation for tagging (UI, backend, and API) and would be happy to refine it based on feedback and align it with the project’s direction.