This article dissects Hermes Agent’s self‑evolution mechanism, explaining how stable facts are stored in memory, reusable procedures become skills, and rollout trajectories are turned into training data through background review, context compression, and OPD‑based token‑level distillation.
The talk explains how Vivo leverages a Kubernetes‑based solution that combines Calico and RoCEv2 to migrate offline training workloads from single‑NIC to multi‑NIC, integrating loss‑less RDMA, planning topology and IP allocation, and employing Volcano, SpiderPool, Macvlan, and Multus CNI for efficient container networking.
To eliminate data fragmentation, feature inconsistencies, and multi‑language implementation challenges, Bilibili built a unified real‑time advertising feature platform that aligns offline, hourly, and online pipelines via a shared C++ library and JNI, boosting CTR prediction accuracy, cutting training costs, and increasing ad revenue by over 1 %.
The article details how Youku Tudou redesigned its video recommendation architecture, moving ranking from offline to online processing, outlining the comparative architecture, benefits, challenges, feature handling, offline evaluation methods, and weight‑fusion techniques that enabled a successful launch after two months of development.
