能否提供嵌入模型和重排模型的本地化方案？

[luojiaaoo]

建议compose通过profile属性提供接入DeepInfra或者本地化（比如ollama/vllm/openAI兼容）的方案选择

[chuanruihu]

您好， 我们新版本已经支持了vllm 启动embedidng / rerank 模型

[luojiaaoo]

谢谢你因为有你，温暖了四级

[cyfyifanchen]

Closing this as part of the EverOS 1.0 issue triage. This issue references the pre-1.0 API or retired infrastructure such as /api/v1/memories, /api/v3/agentic/*, MongoDB, Elasticsearch, Milvus, Redis, Kafka, longjob, or old memory type names. EverOS 1.0 uses POST /api/v1/memory/{add,flush,search,get} with Markdown + SQLite + LanceDB. Migration notes are tracked in PR #258: #258. If the same behavior still occurs on current main with the 1.0 API, please open a fresh issue with a 1.0 repro.

[cyfyifanchen]

Post-triage verification: keeping this closed because EverOS 1.0 removed the old compose/Milvus-centered deployment model and now supports configurable embedding/rerank providers directly. Embedding uses an OpenAI-compatible endpoint via EVEROS_EMBEDDING__MODEL / API_KEY / BASE_URL, and rerank supports provider=deepinfra or provider=vllm through EVEROS_RERANK__PROVIDER / MODEL / API_KEY / BASE_URL. If a specific local provider still fails on current main, that should be filed as a new 1.0 provider bug with the exact endpoint shape.