Closed JensRoland closed 3 months ago
I haven't had time to adjust the tests for all the changes, so they probably don't work at all right now. I've only used test_rerun_trajectories.py to run regressions on the changes. But because of changes I did in the prompts to optimize for Clause 3.5 Sonnet, those also need to be rewritten. My plan right now is to rewrite test_rerun_trajectories.py so that I can at least test the entire flows and check regressions. But I don't have time to fix it right now.
Running the tests (after fixing #19 ) results in:
And:
And:
Maybe I am simply invoking the tests wrong?