Concept有限重放

Sliding-Window Bounded Replay

An optimization technique for inference state recovery scenarios that replays only the tokens within the most recent window, reusing previously cached intermediate states to reduce redundant computation.

Timeline (last 90 days)

Oct 23

DeepSeek-Infer supports limited replay with a sliding window, replaying only a small recent stretch of tokens when restoring state to reduce redundant computation.

Unverified50%

All Facts (1)

Unverified

DeepSeek-Infer supports limited replay with a sliding window, replaying only a small recent stretch of tokens when restoring state to reduce redundant computation.

50%

Source Articles

Full Analysis of DeepSeek's Open-Source Inference Framework DeepInfer: Inside LLM Inference OptimizationOct 23