Marjorie think that 1, 2, and "grabbing the scripts" is ~1 day of work for everything.
1) What numbers are reported in the papers that you read for each of these data sets? Start a table (Where something was reported, what metric, the number. Also, start collecting what the metrics are.
2) Is there leaderboard that reports these numbers together? If so, add recent instances to the table with a note about the leaderboard
3) Moving to our own evaluation, grab the scripts that were used to measure and figure out how to run them.
Marjorie think that 1, 2, and "grabbing the scripts" is ~1 day of work for everything.
1) What numbers are reported in the papers that you read for each of these data sets? Start a table (Where something was reported, what metric, the number. Also, start collecting what the metrics are.
2) Is there leaderboard that reports these numbers together? If so, add recent instances to the table with a note about the leaderboard
3) Moving to our own evaluation, grab the scripts that were used to measure and figure out how to run them.
I shared the overleaf link with the team that had draft tables gathered from the source papers.
I created grabbing the scripts as a separate task. https://github.com/orgs/isi-vista/projects/12/views/1?pane=issue&itemId=20356745