Paul Huliganga 650f893d2d fix: complete LR override — must patch lr_schedule, not just param_groups ... PPO.load() bakes lr_schedule=FloatSchedule(saved_lr) into the model. train() calls _update_learning_rate() which reads lr_schedule, not model.learning_rate. So even with param_groups patched, the first gradient step reverts the optimizer to the saved LR. Complete 3-part fix in create_or_load_model(): model.learning_rate = lr # attribute model.lr_schedule = get_schedule_fn(lr) # prevents train() reverting for pg in optimizer.param_groups: pg['lr'] = lr # immediate effect Also: - SEED_PARAMS: second seed now uses LR=0.001 (was 0.000225) so GP starts with real LR diversity instead of two identical seeds - tests/test_end_to_end.py: 13 new tests covering the full LR override path including a live learn() call; would have caught both bugs - Phase 3 results re-cleared (seed trial 1 ran with half-fix) - 96 tests total, all passing Agent: pi Tests: 96 passed Tests-Added: 13 TypeScript: N/A		2026-04-14 21:27:43 -04:00
..
__init__.py	feat: Wave 1 complete — real PPO training, model save, GP+UCB autoresearch, 37 tests passing	2026-04-13 10:03:15 -04:00
test_autoresearch_controller.py	fix: reward v4 — full sim bypass kills circular driving at root	2026-04-13 20:56:32 -04:00
test_behavioral_wrappers.py	feat: Phase 3 — behavioral control, enhanced evaluator, 53 tests	2026-04-14 09:28:43 -04:00
test_discretize_action.py	feat: Wave 1 complete — real PPO training, model save, GP+UCB autoresearch, 37 tests passing	2026-04-13 10:03:15 -04:00
test_end_to_end.py	fix: complete LR override — must patch lr_schedule, not just param_groups	2026-04-14 21:27:43 -04:00
test_reward_wrapper.py	fix: reward v4 — full sim bypass kills circular driving at root	2026-04-13 20:56:32 -04:00
test_runner_integration.py	feat: Wave 1 complete — real PPO training, model save, GP+UCB autoresearch, 37 tests passing	2026-04-13 10:03:15 -04:00
test_wave3.py	fix: remove Warren from test set — indoor carpet, broken done condition	2026-04-14 13:47:28 -04:00