feat(harness): add Action Recognition training experiment

- PerCategoryPoolStrategy: filter test pool by task_types - RunConfig: add run_holdout_eval toggle (default true) - load_config: fix YAML task_types list-to-tuple conversion - Runner: conditionally skip _holdout_four_way when disabled - CLI: add --no-run-holdout-eval flag - New config/train_action_recognition.yaml (3 epochs, per_category) - New scripts/train_action_recognition.sh (baseline + seed + train)
2026-07-14 00:58:54 -04:00
parent 37d4519905
commit dec7346da3
12 changed files with 1423 additions and 52 deletions
@@ -1368,7 +1368,10 @@ class Runner:
            momentum_updated_task_types=momentum_task_types,
            best_val_acc=state.best_val_acc,
        )
-        await self._holdout_four_way(epoch, pools, state, eval_skills_version, eval_prompts_version)
+        if self._config.run_holdout_eval:
+            await self._holdout_four_way(
+                epoch, pools, state, eval_skills_version, eval_prompts_version
+            )

        # Phase 10: gate 阶梯刷新
        self._refresh_gate_ladder(