fix(grpo): eliminate zero-advantage collapse from stale rewards

fix(grpo): harden rollout reward injection on cloud runtime

fix(pm_ops_trainer): use direct closure capture instead of _self_ref

fix(pm_ops_trainer): use direct closure capture instead of _self_ref

fix(pm_ops_trainer): capture rewards via rollout_func wrapper, not inputs

feat(training): PMOpsGRPOTrainer — override _calculate_rewards to inject rollout rewards directly