jinliyl
44a5217fcd
Merge pull request #101 from agentscope-ai/dev_0205
...
refactor(core): update config parsing and memory management system
2026-02-06 17:48:05 +08:00
jinliyl
f6ca7a733a
Merge branch 'main' into dev_0205
2026-02-06 17:47:44 +08:00
jinli.yl
c8bbb54de1
feat(utils): add cosine similarity functions and integrate into vector operations
2026-02-06 17:45:59 +08:00
jinli.yl
bc5e87db5f
refactor(vector_store): simplify search method signature and remove threshold parameter
2026-02-06 16:53:57 +08:00
jinli.yl
02d5154e1c
style(logo_utils): center align logo display in console output
2026-02-06 16:00:37 +08:00
jinli.yl
6e880c5d7e
refactor(config): rename FileWatchConfig to FileWatcherConfig
2026-02-06 15:56:00 +08:00
jinli.yl
b0b800fbd8
feat(vector-store): add threshold-based filtering to vector search
2026-02-06 15:52:09 +08:00
jinli.yl
9e50e71b52
feat(core): add memory store and file watcher support to application
2026-02-06 15:35:09 +08:00
jinli.yl
1dd81f9c25
refactor(core): update config parsing and memory management system
2026-02-06 15:02:41 +08:00
jinliyl
90cd9ddea8
Merge pull request #100 from agentscope-ai/dev_0205
...
Dev 0205
2026-02-06 12:17:15 +08:00
jinli.yl
88674c52f9
fix(memory): resolve enum value access and tool processing issues
2026-02-06 12:16:28 +08:00
jinli.yl
c5438c52fc
feat(memory): implement file watcher with delta and full sync strategies
2026-02-06 01:58:28 +08:00
jinliyl
ce7fffbd58
Merge pull request #99 from agentscope-ai/fix_0204
...
fix(agent): implement memory targets filtering in base memory agent
2026-02-04 20:47:02 +08:00
jinli.yl
8c05f96fb6
feat(agent): implement memory targets filtering in base memory agent
...
- Add memory_targets parameter to filter memory target type mapping
- Modify memory_target_type_mapping property to use filtered targets when available
- Update REME class to collect memory targets from user, task, and tool names
- Pass collected memory targets to agent initialization
- Support both string and list inputs for user, task, and tool names
- Maintain
2026-02-04 17:51:28 +08:00
jinli.yl
857cd52a8e
feat(memory): add memory management and file system tools
2026-02-03 22:29:45 +08:00
jinliyl
1aca24b804
Merge pull request #97 from nitwtog/main
...
修正halumem/eval_reme.py评估函数, 添加v2版本用于适配readhistoryv2
2026-02-03 15:56:57 +08:00
方应
6e26006c9f
feat(core): 添加 v2 版本的个人总结器和检索器支持
...
- 在版本控制逻辑中加入 v2 版本的支持判断
- 激活了之前被注释掉的 ReadHistory 工具并整合到检索器中
- 调整了 ReMeSummarizer 和 ReMeRetriever 的版本兼容性检查
2026-02-03 14:55:26 +08:00
方应
5ffb85a10e
fix(halumem): 修复评估模型名称变量引用错误
...
- 将reme_model_name更正为eval_model_name以匹配正确的模型配置
2026-02-03 14:39:08 +08:00
方应
d83551c2cf
feat(benchmark): 更新ReMe评估配置参数
...
- 将最大并发数从2调整为1，批量大小从20调整为40
- 算法版本从halumem更新为v1
- 添加了新的reme_model_name参数用于模型名称配置
- 在MemoryProcessor中使用reme_model_name替代eval_model_name
- 主函数中添加batch_size参数传递
- 命令行解析器中添加batch_size参数选项
- 在ReMe中将ReadHistory替换为ReadHistoryV2并配置新参数
2026-02-03 14:31:04 +08:00
jinliyl
0fb898ca3c
Merge pull request #95 from agentscope-ai/dev_0202
...
feat(memory): add ReadHistoryV2 tool with enhanced history retrieval …
2026-02-02 17:18:42 +08:00
jinli.yl
4159cfc51c
feat(memory): add ReadHistoryV2 tool with enhanced history retrieval capabilities
2026-02-02 17:14:16 +08:00
jinli.yl
cb4b6b2800
fix(workflow): correct output file extension in translation workflow
2026-02-01 15:32:39 +08:00
jinli.yl
b6d3ce69a7
fix(translate_ts): correct translation output handling
2026-02-01 15:32:23 +08:00
jinli.yl
2233aa8403
feat(workflow): add translate_ts operation and rename test module to gallery
2026-02-01 15:25:49 +08:00
jinliyl
ad8b889461
Merge pull request #93 from agentscope-ai/dev_0131
...
feat(cmd_service): initialize service context on command service startup
2026-01-31 16:41:40 +08:00
jinli.yl
ce653e1f11
fix(cmd): update cmd flow error message and integrate service context
2026-01-31 16:39:29 +08:00
jinli.yl
ec4fd33fbe
refactor(history): update tool call schema description
2026-01-31 11:27:21 +08:00
jinli.yl
93861942ca
docs(memory): update personal memory retriever and summarizer documentation
2026-01-31 11:23:03 +08:00
jinli.yl
b5fe827a36
refactor(reme): convert _resolve_memory_target to static method
2026-01-31 03:47:25 +08:00
jinli.yl
6a7396732b
feat(reme): add comprehensive memory management API
2026-01-31 03:40:38 +08:00
jinli.yl
c7ba031cea
feat(benchmark): add enable_thinking_params configuration option
2026-01-31 02:50:37 +08:00
jinli.yl
5a5e06a43d
feat(memory): add draft memory functionality and update memory/profile management tools
2026-01-31 02:36:10 +08:00
jinli.yl
9a0d5058a2
refactor(memory): reorganize memory module imports and update personal retriever
2026-01-31 02:05:08 +08:00
jinli.yl
5a4077e68c
fix(halumem): correct async context manager implementation
2026-01-31 01:25:23 +08:00
jinli.yl
0163cb25ce
feat(cmd_service): initialize service context on command service startup
2026-01-31 01:17:00 +08:00
jinli.yl
3680571c94
refactor(core): simplify component registration and improve application lifecycle management
2026-01-31 01:05:31 +08:00
jinliyl
777e08ecc9
Merge pull request #91 from nitwtog/main
...
优化agentic memory效果
2026-01-30 17:35:58 +08:00
方应
c934c8c6d2
config: 更新默认配置中的模型设置
...
- 将默认模型从 qwen-flash 更改为 qwen3-30b-a3b-instruct-2507
- 移除已注释的模型配置选项
- 删除配置文件末尾的空行
- 移除未使用的 flow 配置段落
2026-01-30 17:17:45 +08:00
方应
61071ab654
feat(benchmark): 添加ReMe模型名称配置选项
...
- 在EvalConfig中新增reme_model_name参数，默认值为qwen-flash
- 修改ReMe实例化方式，传入配置的模型名称参数
- 更新函数签名以支持reme_model_name参数传递
- 调整命令行参数解析，区分reme_model_name和eval_model_name
- 将默认reme模型从硬编码改为可配置参数
2026-01-30 17:17:06 +08:00
方应
07905a0358
feat(memory): 添加用户配置文件增删功能并优化内存管理
...
- 新增 AddProfile 工具类用于添加用户配置文件
- 新增 DeleteProfile 工具类用于删除指定ID的配置文件
- 从 __init__.py 中移除已废弃的 UpdateProfileFilterOlder 工具
- 在 reme.py 中注释掉 AddProfile 和 DeleteProfile 的导入
- 从 reme.py 的工具列表中移除 UpdateProfileFilterOlder 相关代码
- 优化 ES 和 Qdrant 向量存储客户端初始化的日志记录
- 更新个人记忆检索器和摘要器的 YAML 配置文件格式
- 修复 personal_halumem_summarizer.py 中的参数传递格式问题
- 在 UpdateProfile 工具中添加按 memory_target 分组的功能
2026-01-30 17:03:58 +08:00
方应
74238ba9bd
feat(memory): 添加个人记忆检索和总结功能
...
- 实现 PersonalHalumemRetriever 类用于向量搜索检索个人记忆
- 实现 PersonalHalumemSummarizer 类用于两阶段个人记忆处理
- 配置个人记忆检索和总结的系统提示词和用户消息模板
- 支持多阶段检索策略包括意图分解和深度追踪
- 实现记忆评估和原始来源追溯功能
- 添加记忆存储范围过滤和配置管理规则
2026-01-30 16:47:10 +08:00
方应
381162bf63
Merge remote-tracking branch 'origin/main'
2026-01-30 16:42:36 +08:00
方应
0f90926781
Merge remote-tracking branch 'origin/main'
...
# Conflicts:
# reme/reme.py
2026-01-30 16:33:32 +08:00
Zhouwk
ba812cdd6a
Merge branch 'agentscope-ai:main' into main
2026-01-30 16:31:07 +08:00
方应
a0aec042c9
remove(bench): 删除旧的基准测试分析脚本
...
- 移除 HaluMem 数据集统计分析脚本 (analyze_dataset_stats.py)
- 移除评估结果分析脚本 (analyze_results.py)
- 移除人工回路问答统计计算脚本 (compute_qa_stats.py)
- 清理重复的人工回路2问答统计脚本
- 删除相关的数据分析和结果统计功能模块
2026-01-30 16:30:28 +08:00
方应
053c537845
feat(agent): 添加 Halumem 版本的记忆检索器和摘要器
...
- 添加 PersonalHalumemRetriever 和 PersonalHalumemSummarizer 类
- 在 memory 模块中注册新的检索器和摘要器
- 添加 UpdateProfileFilterOlder、DeleteProfile 和 AddProfile 工具
- 将 AddDraftAndRetrieveSimilarMemory 重命名为 AddAndRetrieveSimilarMemory
- 修改配置文件中的默认模型名称为 qwen-flash
- 在 benchmark 中添加 Halumem 评估支持和实时更新功能
- 降低 ProfileHandler 的最大容量限制并添加重复节点过滤逻辑
- 在 ReMe 中添加 halumem 版本的记忆代理配置
2026-01-30 16:30:18 +08:00
jinliyl
5017a46f51
Merge pull request #88 from agentscope-ai/dev_0129
...
refactor(core): implement lazy initialization for Elasticsearch and Q…
2026-01-29 16:28:23 +08:00
jinli.yl
0ea51a9378
refactor(core): implement lazy initialization for Elasticsearch and Qdrant clients
2026-01-29 16:27:02 +08:00
jinliyl
11d3af302b
Merge pull request #87 from agentscope-ai/dev_0129
...
refactor(memory): update memory registration and configuration handling
2026-01-29 15:17:22 +08:00
jinli.yl
c581dd38b7
feat(vector-store): add async collection reset functionality
2026-01-29 15:10:58 +08:00
