You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Is your feature request related to a problem? Please describe.
The Progress Viewer tab currently displays logs related to model training, but it lacks comprehensive integration with third-party logging services (e.g., TensorBoard, Weights & Biases). This limitation makes it difficult for users to access detailed logs and visualizations directly within the Progress Viewer, forcing them to switch between multiple platforms to track training progress and performance metrics. This disrupts the workflow and makes monitoring and debugging model training cumbersome.
Describe the solution you'd like
Enhance the Progress Viewer by integrating third-party logging services, allowing users to view detailed logs, graphs, and performance metrics directly within the tab. This could include displaying real-time updates, loss curves, accuracy plots, and other visual data provided by these logging tools. The integration should be seamless, with clear options to connect third-party services, select which logs to display, and customize the view to suit the user’s preferences.
Describe alternatives you've considered
Providing a configurable API endpoint within the Progress Viewer to pull in logs from third-party sources without deep integration, giving users basic but direct access to external logging.
Allowing users to download logs from third-party services and manually upload them into the Progress Viewer for visualization, though this would still lack real-time functionality.
Additional context
Improving third-party logging within the Progress Viewer will streamline the model training workflow, reduce context switching, and provide a more unified view of training progress. Below are examples of other platforms that effectively integrate third-party logging, enhancing the usability and monitoring capabilities of the training process.
The text was updated successfully, but these errors were encountered:
Is your feature request related to a problem? Please describe.
The Progress Viewer tab currently displays logs related to model training, but it lacks comprehensive integration with third-party logging services (e.g., TensorBoard, Weights & Biases). This limitation makes it difficult for users to access detailed logs and visualizations directly within the Progress Viewer, forcing them to switch between multiple platforms to track training progress and performance metrics. This disrupts the workflow and makes monitoring and debugging model training cumbersome.
Describe the solution you'd like
Enhance the Progress Viewer by integrating third-party logging services, allowing users to view detailed logs, graphs, and performance metrics directly within the tab. This could include displaying real-time updates, loss curves, accuracy plots, and other visual data provided by these logging tools. The integration should be seamless, with clear options to connect third-party services, select which logs to display, and customize the view to suit the user’s preferences.
Describe alternatives you've considered
Providing a configurable API endpoint within the Progress Viewer to pull in logs from third-party sources without deep integration, giving users basic but direct access to external logging.
Allowing users to download logs from third-party services and manually upload them into the Progress Viewer for visualization, though this would still lack real-time functionality.
Additional context
Improving third-party logging within the Progress Viewer will streamline the model training workflow, reduce context switching, and provide a more unified view of training progress. Below are examples of other platforms that effectively integrate third-party logging, enhancing the usability and monitoring capabilities of the training process.
The text was updated successfully, but these errors were encountered: