Jonathan Giannuzzi
Jonathan Giannuzzi
The tables that display runs in each explorer can become really sluggish when there are many columns to display. Instead of displaying all available params and metrics by default, we...
New users looking at our README wouldn't know what makes FastTrackML unique and why they should bother trying it over MLFlow or other experiment trackers. We should better describe what...
In the Runs tab, when trying to export the table we only get the rows that are rendered in the UI, we have to scroll down to reveal more rows....
https://github.com/G-Research/fasttrackml-ui-mlflow/pull/18
Once #6 and #224 are implemented, artifacts will become visible in the MLFlow UI. This issue is about making them visible in the Aim UI. Aim treats artifacts like metrics...
We are not currently testing whether pagination works in our S3 artifacts integration tests. We need to add a test that creates multiple thousands of artifacts for a given run...
The various Aim streaming integration tests do not seem to take ordering into account. We should check that we received exactly what we expected in the order we expected instead...
Create a proof of concept where data gets served as Arrow records by the backend, stream loaded in the frontend via Arrow, and then served via a virtual data grid
This is not yet a minimal reproducible example, but here goes: - 79 runs, 2261 metrics + 112 params per run, 50 values per metrics - go to Runs explorer...
Our default port is 5000, which can be taken on recent versions of macOS when AirPlay Receiver is enabled. We should change our default port to something much higher and...