Posted 2023-12-18Updated 2026-03-11Artificial Intelligencea few seconds read (About 47 words)Inference Basic 导言 RL 涉及到 推理,推理的流程细节不是很明晰。 warmup,计算kvcache chunked prefill,降低prefill的显存 Inference Basichttp://icarus.shaojiemike.top/2023/12/18/Work/Artificial Intelligence/Inference/InferenceBasic/AuthorShaojie TanPosted on2023-12-18Updated on2026-03-11Licensed under#fun
2026-02-05The Mechanics of RL: How Inference Sampling Shapes the Probability LandscapeArtificial Intelligence