implementation detail · filed under post-training
Decoupling agent rollout into sandbox and worker container
Splits agent rollout execution between an agent sandbox and a worker container for RL across diverse scaffolds.
- source
- 1
- model
- 1
- lab adopt it
- 1
- strongest
- used
How sources treat it
One count per evidence span, weakest treatment to strongest.
used 2
Documented in
Evidence
2 spans quoted from the sources, strongest treatment first.
For RL training across diverse scaffolds, we decouple agent rollout execution into an agent sandbox and a worker container.
usedunclearin DeepSeek-V4.1-FlashDeepSeek
For RL training across diverse scaffolds, we decouple agent rollout execution into an agent sandbox and a worker container.
usedsoftware implementationin DeepSeek Elastic Compute (DSec)DeepSeek
Filed alongside
Other methods under post-training :: rollout & RL infrastructure.
Asynchronous reinforcement learningPartial rolloutAsynchronous RL frameworks for large-scale agent scaffolds and environment orchestrationToken-in-token-out (TITO)Asynchronous reinforcement learning infrastructureCo-located RL trainingData SchedulerDecoupled control plane and data planeDeficit-corrected schedulingLarge-scale asynchronous RL in synthesized tasksOne-step off-policy asynchronous reinforcement learningPredictive Rollout DispatchSample-grained garbage collectionSeamless Rollout EngineSLIMEToken-granularity persistence of rollout statesToken-level interruptionTool ManagerToolboxAdaptive Rollout ConcurrencyAdaptive Rollout SchedulingAgent LoopAgent-centric rollout executionAsynchronous Agent RL algorithms