面向生成式推荐的 CUDA Graph 推理引擎,支持宽 beam search 和 FP8 服务,解决推荐系统延迟和吞吐瓶颈。85 stars 反映 AI 基础设施关注。商业化可做推荐服务或开源标准,但竞品是 NVIDIA Merlin 和自研系统,风险是生态锁定。
FlashRec is a CUDA-graph engine for generative recommendation: wide beam search (3–5 SID steps, n=50–512+) over a trie-constrained catalog, in-process FP8 serving, and ranked beams on /v1/chat/complet…
★85