Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
Summary:
• Memory & Retrieval Core: async add (plain & preference), Preference Memory pipeline, Reranker strategy suite, BM25 for TreeTextMemory, MemReader structural parsing.
• Scheduler & Observability: API scheduler modularization (schema/utils/analyzer), Redis ORM for history sync & mixture search, metrics + request logs, Nacos-based dynamic config.
• Data & Infra: PolarDB graph backend with pool/timeout & fixes, unified graph factory (Nebula/Neo4j/PolarDB), Milvus interface & item optimizations, enhanced logging.
• Evaluation: PrefEval standardization; LoCoMo/LongMemEval/PersonaMem pipeline upgrades; new utilities (e.g., mirix_utils.py).
• Stability & Fixes: query scheduling, message schema, Tree search inputs, self-input prompts, SQLite list users, PolarDB value mapping; pool/timeout tuning, usage data removal, graph-call toggle.
• Compatibility: adopt new async path/schemas/metrics; align preference fields; migrate tests to Redis ORM; configure PolarDB pool/timeout before rollout.
• 记忆与检索内核:新增 async add(明文/偏好)、偏好记忆全链路、Reranker 策略集、TreeTextMemory 引入 BM25、MemReader 结构化解析。
• 调度与可观测性:API 调度模块化(schema/utils/analyzer)、Redis ORM(历史同步与混合搜索)、metrics 指标与请求日志、Nacos 动态配置。
• 数据与基础设施:PolarDB 图后端(连接池/超时与修复)、图工厂统一(Nebula/Neo4j/PolarDB)、Milvus 接口与数据项优化、日志链路增强。
• 评估体系:PrefEval 字段标准化;LoCoMo/LongMemEval/PersonaMem 评测升级;新增工具(如 mirix_utils.py)。
• 稳定性与修复:修复查询调度、消息 schema、树检索输入、自输入提示、SQLite 用户列表、PolarDB 值映射;连接池/超时优化、移除 usage 数据、图调用开关。
• 兼容性:适配异步路径/新 schema/metrics;偏好字段按新规范;测试迁移至 Redis ORM;启用 PolarDB 前配置连接池与超时并压测。
Fix: #424 #426 #423 #443 #384 #406 #445
• Scheduler/Query: fixed query-schedule edge cases (#424).
• Schemas: corrected message schema inconsistencies (#426).
• Search I/O: fixed TreeTextMemory searcher input mismatch (#423).
• Prompts: fixed self-input prompt error (#443).
• Storage: fixed SQLite list-users error (#384).
• Graph/DB: corrected PolarDB value/graph issues (#406, #445).
• 调度/查询:修复查询调度边界问题(https://github.com/MemTensor/MemOS/pull/424)。
• Schema:修正消息 schema 不一致(https://github.com/MemTensor/MemOS/pull/426)。
• 检索 I/O:修复 Tree 搜索器输入不匹配(https://github.com/MemTensor/MemOS/pull/423)。
• 提示:修复自输入提示错误(https://github.com/MemTensor/MemOS/pull/443)。
• 存储:修复 SQLite 用户列表错误(https://github.com/MemTensor/MemOS/pull/384)。
• 图/数据库:修正 PolarDB 值与图相关问题(https://github.com/MemTensor/MemOS/pull/406、https://github.com/MemTensor/MemOS/pull/445)。
Docs Issue/PR: (docs-issue-or-pr-link)
Reviewer: @(reviewer)
Checklist: