Skip to content

v1.78.0

Choose a tag to compare

@github-actions github-actions released this 21 Oct 16:11
· 24 commits to master since this release

New Features

  • 51bc2e3 - gooddata-eval: record when a run started and finished
  • f11d849 - gooddata-eval: record per-run latency and report percentiles
  • a113ff8 - gooddata-eval: restyle HTML report, show latency percentiles
  • c78d0fb - gooddata-eval: split latency by single-shot vs multi-turn
  • a391a5d - gooddata-eval: label item latency, multi-select kind filter
  • aee4e86 - gooddata-eval: filters on top drive the whole HTML report
  • b82e66a - gooddata-eval: two filter tiers, multi-select status and run
  • 504bf1a - gooddata-eval: add a trimmed mean to the latency distributions
  • cfb5e8b - gooddata-eval: redesign the HTML report as a readable document
  • f9cab90 - gooddata-eval: add a By model view and outcomes table to the HTML report
  • f272509 - gooddata-eval: show plain model positions in the quality-against-speed chart
  • 81387ec - gooddata-eval: open the HTML report by date and order its runs by start time
  • bf52f72 - gooddata-eval: update HTML report to include run date and name in outcomes table
  • 869498f - add started_at timestamp to EvalReport and related tests
  • 5223bd4 - gooddata-eval: render markdown titles as bold in reasoning sections
  • 66c2b26 - gooddata-eval: add legend for sparkline dots to clarify comparisons in reports

Bug Fixes

  • 37fadc3 - gooddata-eval: stop alert eval failing correct runs (commit by @myhoai)
  • 057c263 - gooddata-eval: name the builder skill flag the backend defines
  • 8ba9a37 - gooddata-eval: count timed-out turns and fix HTML change labels
  • 5935e7d - gooddata-eval: draw the HTML report sparkline dots as true circles
  • 417b74e - gooddata-eval: do not label a first-seen skipped test case New
  • acf6b6e - gooddata-eval: address review on latency and redacted reports
  • ac60411 - gooddata-sdk: keep isHiddenFromKda on declarative metrics (commit by @tychtjan)

Tests

  • 291b39d - gooddata-eval: annotate the run start/finish tests and fake controller

Chores