Skip to content

fix: general_text retrieval stability + get_* compatibility #1163

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
rjstratminds wants to merge 3 commits into
base: main
Choose a base branch
from
Open

fix: general_text retrieval stability + get_* compatibility #1163

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
66d4798
backup: local memos customizations before upstream sync 20260305-002125
rjstratminds Mar 5, 2026
5c40f67
fix(api): support general_text get_all signatures in memory handlers
rjstratminds Mar 5, 2026
19f0b10
fix(api): stabilize get_* for general_text with paged qdrant reads
rjstratminds Mar 5, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
187 changes: 120 additions & 67 deletions src/memos/api/config.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -437,6 +437,7 @@ def get_embedder_config() -> dict[str, Any]:
"provider": os.getenv("MOS_EMBEDDER_PROVIDER", "openai"),
"api_key": os.getenv("MOS_EMBEDDER_API_KEY", "sk-xxxx"),
"model_name_or_path": os.getenv("MOS_EMBEDDER_MODEL", "text-embedding-3-large"),
"embedding_dims": int(os.getenv("EMBEDDING_DIMENSION", "1024")),
"headers_extra": json.loads(os.getenv("MOS_EMBEDDER_HEADERS_EXTRA", "{}")),
"base_url": os.getenv("MOS_EMBEDDER_API_BASE", "http://openai.com"),
"backup_client": os.getenv("MOS_EMBEDDER_BACKUP_CLIENT", "false").lower()
Expand Down Expand Up @@ -984,43 +985,71 @@ def create_user_config(user_name: str, user_id: str) -> tuple["MOSConfig", "Gene
graph_db_backend = os.getenv(
"GRAPH_DB_BACKEND", os.getenv("NEO4J_BACKEND", "neo4j-community")
).lower()
if graph_db_backend in graph_db_backend_map:
text_mem_type = os.getenv("MOS_TEXT_MEM_TYPE", "tree_text").lower()

if text_mem_type == "general_text":
text_mem_cfg = {
"backend": "general_text",
"config": {
"extractor_llm": {"backend": "openai", "config": openai_config},
"vector_db": {
"backend": "qdrant",
"config": {
"collection_name": os.getenv("MOS_GENERAL_TEXT_COLLECTION", "general_text_mem"),
"vector_dimension": int(os.getenv("EMBEDDING_DIMENSION", 3072)),
"distance_metric": "cosine",
"host": os.getenv("QDRANT_HOST", "localhost"),
"port": int(os.getenv("QDRANT_PORT", "6333")),
"path": os.getenv("QDRANT_PATH"),
"url": os.getenv("QDRANT_URL"),
"api_key": os.getenv("QDRANT_API_KEY"),
},
},
"embedder": APIConfig.get_embedder_config(),
},
}
elif graph_db_backend in graph_db_backend_map:
text_mem_cfg = {
"backend": "tree_text",
"config": {
"extractor_llm": {"backend": "openai", "config": openai_config},
"dispatcher_llm": {"backend": "openai", "config": openai_config},
"graph_db": {
"backend": graph_db_backend,
"config": graph_db_backend_map[graph_db_backend],
},
"embedder": APIConfig.get_embedder_config(),
"internet_retriever": internet_config,
"reranker": APIConfig.get_reranker_config(),
"reorganize": os.getenv("MOS_ENABLE_REORGANIZE", "false").lower()
== "true",
"memory_size": {
"WorkingMemory": int(os.getenv("NEBULAR_WORKING_MEMORY", 20)),
"LongTermMemory": int(os.getenv("NEBULAR_LONGTERM_MEMORY", 1e6)),
"UserMemory": int(os.getenv("NEBULAR_USER_MEMORY", 1e6)),
},
"search_strategy": {
"fast_graph": bool(os.getenv("FAST_GRAPH", "false") == "true"),
"bm25": bool(os.getenv("BM25_CALL", "false") == "true"),
"cot": bool(os.getenv("VEC_COT_CALL", "false") == "true"),
"fulltext": bool(os.getenv("FULLTEXT_CALL", "false") == "true"),
},
"include_embedding": bool(
os.getenv("INCLUDE_EMBEDDING", "false") == "true"
),
},
}
else:
raise ValueError(f"Invalid graph db backend: {graph_db_backend}")

if text_mem_cfg:
# Create MemCube config

default_cube_config = GeneralMemCubeConfig.model_validate(
{
"user_id": user_id,
"cube_id": f"{user_name}_default_cube",
"text_mem": {
"backend": "tree_text",
"config": {
"extractor_llm": {"backend": "openai", "config": openai_config},
"dispatcher_llm": {"backend": "openai", "config": openai_config},
"graph_db": {
"backend": graph_db_backend,
"config": graph_db_backend_map[graph_db_backend],
},
"embedder": APIConfig.get_embedder_config(),
"internet_retriever": internet_config,
"reranker": APIConfig.get_reranker_config(),
"reorganize": os.getenv("MOS_ENABLE_REORGANIZE", "false").lower()
== "true",
"memory_size": {
"WorkingMemory": int(os.getenv("NEBULAR_WORKING_MEMORY", 20)),
"LongTermMemory": int(os.getenv("NEBULAR_LONGTERM_MEMORY", 1e6)),
"UserMemory": int(os.getenv("NEBULAR_USER_MEMORY", 1e6)),
},
"search_strategy": {
"fast_graph": bool(os.getenv("FAST_GRAPH", "false") == "true"),
"bm25": bool(os.getenv("BM25_CALL", "false") == "true"),
"cot": bool(os.getenv("VEC_COT_CALL", "false") == "true"),
"fulltext": bool(os.getenv("FULLTEXT_CALL", "false") == "true"),
},
"include_embedding": bool(
os.getenv("INCLUDE_EMBEDDING", "false") == "true"
),
},
},
"text_mem": text_mem_cfg,
"act_mem": {}
if os.getenv("ENABLE_ACTIVATION_MEMORY", "false").lower() == "false"
else APIConfig.get_activation_vllm_config(),
Expand All @@ -1030,8 +1059,6 @@ def create_user_config(user_name: str, user_id: str) -> tuple["MOSConfig", "Gene
else APIConfig.get_preference_memory_config(),
}
)
else:
raise ValueError(f"Invalid Neo4j backend: {graph_db_backend}")
default_mem_cube = GeneralMemCube(default_cube_config)
return default_config, default_mem_cube

Expand Down Expand Up @@ -1069,42 +1096,70 @@ def get_default_cube_config() -> "GeneralMemCubeConfig | None":
graph_db_backend = os.getenv(
"GRAPH_DB_BACKEND", os.getenv("NEO4J_BACKEND", "neo4j-community")
).lower()
if graph_db_backend in graph_db_backend_map:
text_mem_type = os.getenv("MOS_TEXT_MEM_TYPE", "tree_text").lower()

if text_mem_type == "general_text":
text_mem_cfg = {
"backend": "general_text",
"config": {
"extractor_llm": {"backend": "openai", "config": openai_config},
"vector_db": {
"backend": "qdrant",
"config": {
"collection_name": os.getenv("MOS_GENERAL_TEXT_COLLECTION", "general_text_mem"),
"vector_dimension": int(os.getenv("EMBEDDING_DIMENSION", 3072)),
"distance_metric": "cosine",
"host": os.getenv("QDRANT_HOST", "localhost"),
"port": int(os.getenv("QDRANT_PORT", "6333")),
"path": os.getenv("QDRANT_PATH"),
"url": os.getenv("QDRANT_URL"),
"api_key": os.getenv("QDRANT_API_KEY"),
},
},
"embedder": APIConfig.get_embedder_config(),
},
}
elif graph_db_backend in graph_db_backend_map:
text_mem_cfg = {
"backend": "tree_text",
"config": {
"extractor_llm": {"backend": "openai", "config": openai_config},
"dispatcher_llm": {"backend": "openai", "config": openai_config},
"graph_db": {
"backend": graph_db_backend,
"config": graph_db_backend_map[graph_db_backend],
},
"embedder": APIConfig.get_embedder_config(),
"reranker": APIConfig.get_reranker_config(),
"reorganize": os.getenv("MOS_ENABLE_REORGANIZE", "false").lower()
== "true",
"internet_retriever": internet_config,
"memory_size": {
"WorkingMemory": int(os.getenv("NEBULAR_WORKING_MEMORY", 20)),
"LongTermMemory": int(os.getenv("NEBULAR_LONGTERM_MEMORY", 1e6)),
"UserMemory": int(os.getenv("NEBULAR_USER_MEMORY", 1e6)),
},
"search_strategy": {
"fast_graph": bool(os.getenv("FAST_GRAPH", "false") == "true"),
"bm25": bool(os.getenv("BM25_CALL", "false") == "true"),
"cot": bool(os.getenv("VEC_COT_CALL", "false") == "true"),
"fulltext": bool(os.getenv("FULLTEXT_CALL", "false") == "true"),
},
"mode": os.getenv("ASYNC_MODE", "sync"),
"include_embedding": bool(
os.getenv("INCLUDE_EMBEDDING", "false") == "true"
),
},
}
else:
raise ValueError(f"Invalid graph db backend: {graph_db_backend}")

if text_mem_cfg:
return GeneralMemCubeConfig.model_validate(
{
"user_id": "default",
"cube_id": "default_cube",
"text_mem": {
"backend": "tree_text",
"config": {
"extractor_llm": {"backend": "openai", "config": openai_config},
"dispatcher_llm": {"backend": "openai", "config": openai_config},
"graph_db": {
"backend": graph_db_backend,
"config": graph_db_backend_map[graph_db_backend],
},
"embedder": APIConfig.get_embedder_config(),
"reranker": APIConfig.get_reranker_config(),
"reorganize": os.getenv("MOS_ENABLE_REORGANIZE", "false").lower()
== "true",
"internet_retriever": internet_config,
"memory_size": {
"WorkingMemory": int(os.getenv("NEBULAR_WORKING_MEMORY", 20)),
"LongTermMemory": int(os.getenv("NEBULAR_LONGTERM_MEMORY", 1e6)),
"UserMemory": int(os.getenv("NEBULAR_USER_MEMORY", 1e6)),
},
"search_strategy": {
"fast_graph": bool(os.getenv("FAST_GRAPH", "false") == "true"),
"bm25": bool(os.getenv("BM25_CALL", "false") == "true"),
"cot": bool(os.getenv("VEC_COT_CALL", "false") == "true"),
"fulltext": bool(os.getenv("FULLTEXT_CALL", "false") == "true"),
},
"mode": os.getenv("ASYNC_MODE", "sync"),
"include_embedding": bool(
os.getenv("INCLUDE_EMBEDDING", "false") == "true"
),
},
},
"text_mem": text_mem_cfg,
"act_mem": {}
if os.getenv("ENABLE_ACTIVATION_MEMORY", "false").lower() == "false"
else APIConfig.get_activation_vllm_config(),
Expand All @@ -1114,5 +1169,3 @@ def get_default_cube_config() -> "GeneralMemCubeConfig | None":
else APIConfig.get_preference_memory_config(),
}
)
else:
raise ValueError(f"Invalid Neo4j backend: {graph_db_backend}")
8 changes: 5 additions & 3 deletions src/memos/api/handlers/add_handler.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -33,9 +33,11 @@ def __init__(self, dependencies: HandlerDependencies):
dependencies: HandlerDependencies instance
"""
super().__init__(dependencies)
self._validate_dependencies(
"naive_mem_cube", "mem_reader", "mem_scheduler", "feedback_server"
)
required = ["naive_mem_cube", "mem_reader", "mem_scheduler"]
text_mem = getattr(getattr(dependencies, "naive_mem_cube", None), "text_mem", None)
if hasattr(text_mem, "get_searcher"):
required.append("feedback_server")
self._validate_dependencies(*required)

def handle_add_memories(self, add_req: APIADDRequest) -> MemoryResponse:
"""
Expand Down
Loading