feat(cli): Support adk eval --num_runs N as AgentEvaluator - #6622
Closed
copybara-service[bot] wants to merge 1 commit into
Closed
feat(cli): Support adk eval --num_runs N as AgentEvaluator#6622copybara-service[bot] wants to merge 1 commit into
adk eval --num_runs N as AgentEvaluator#6622copybara-service[bot] wants to merge 1 commit into
Conversation
copybara-service
Bot
force-pushed
the
copybara/959841339
branch
from
August 7, 2026 18:44
2229c83 to
39c8cfa
Compare
Merge #4411 **Problem:** `adk eval` did not support repeated runs, so users had to use external loops to reduce nondeterminism. Also, when adding multi-run support, reporting needs to stay consistent between summary and detailed output. **Solution:** Added `--num_runs` to `adk eval` (default: `1`, min: `1`) and wired it into inference execution so each eval set can be run multiple times in a single CLI invocation. For reporting: - Summary is aggregated per eval case (`eval_set_id` + `eval_id`) across runs. - Detailed output (`--print_detailed_results`) keeps per-run details (each run's `EvalCaseResult`) to avoid mismatches between aggregated scores and invocation-level debugging info. Also added unit tests for: - Aggregation behavior across runs. - CLI validation for invalid `--num_runs` values. - Multi-run summary aggregation and detailed per-run printing behavior. Related: #4410 Co-authored-by: Haran Rajkumar <haranrk@google.com> PiperOrigin-RevId: 959841339
copybara-service
Bot
force-pushed
the
copybara/959841339
branch
from
August 7, 2026 18:45
39c8cfa to
c9d5ed6
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
feat(cli): Support
adk eval --num_runs Nas AgentEvaluatorMerge #4411
Problem:
adk evaldid not support repeated runs, so users had to use external loops toreduce nondeterminism. Also, when adding multi-run support, reporting needs to
stay consistent between summary and detailed output.
Solution:
Added
--num_runstoadk eval(default:1, min:1) and wired it intoinference execution so each eval set can be run multiple times in a single CLI
invocation.
For reporting:
eval_set_id+eval_id) across runs.--print_detailed_results) keeps per-run details (each run'sEvalCaseResult) to avoid mismatches between aggregated scores andinvocation-level debugging info.
Also added unit tests for:
--num_runsvalues.Related: #4410
Co-authored-by: Haran Rajkumar haranrk@google.com