Skip to content

feat: add configurable LLM settings for mem0_memory tool #221

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

Merged
merged 1 commit into from
Aug 20, 2025
Merged

feat: add configurable LLM settings for mem0_memory tool #221

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
9 changes: 8 additions & 1 deletion README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -759,12 +759,19 @@ The Mem0 Memory Tool supports three different backend configurations:
| OPENSEARCH_HOST | OpenSearch Host URL | None | OpenSearch |
| AWS_REGION | AWS Region for OpenSearch | us-west-2 | OpenSearch |
| DEV | Enable development mode (bypasses confirmations) | false | All modes |
| MEM0_LLM_PROVIDER | LLM provider for memory processing | aws_bedrock | All modes |
| MEM0_LLM_MODEL | LLM model for memory processing | anthropic.claude-3-5-haiku-20241022-v1:0 | All modes |
| MEM0_LLM_TEMPERATURE | LLM temperature (0.0-2.0) | 0.1 | All modes |
| MEM0_LLM_MAX_TOKENS | LLM maximum tokens | 2000 | All modes |
| MEM0_EMBEDDER_PROVIDER | Embedder provider for vector embeddings | aws_bedrock | All modes |
| MEM0_EMBEDDER_MODEL | Embedder model for vector embeddings | amazon.titan-embed-text-v2:0 | All modes |


**Note**:
- If `MEM0_API_KEY` is set, the tool will use the Mem0 Platform
- If `OPENSEARCH_HOST` is set, the tool will use OpenSearch
- If neither is set, the tool will default to FAISS (requires `faiss-cpu` package)

- LLM configuration applies to all backend modes and allows customization of the language model used for memory processing
#### Memory Tool

| Environment Variable | Description | Default |
Expand Down
13 changes: 8 additions & 5 deletions src/strands_tools/mem0_memory.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -150,13 +150,16 @@ class Mem0ServiceClient:
"""Client for interacting with Mem0 service."""

DEFAULT_CONFIG = {
"embedder": {"provider": "aws_bedrock", "config": {"model": "amazon.titan-embed-text-v2:0"}},
"embedder": {
"provider": os.environ.get("MEM0_EMBEDDER_PROVIDER", "aws_bedrock"),
"config": {"model": os.environ.get("MEM0_EMBEDDER_MODEL", "amazon.titan-embed-text-v2:0")},
},
"llm": {
"provider": "aws_bedrock",
"provider": os.environ.get("MEM0_LLM_PROVIDER", "aws_bedrock"),
"config": {
"model": "anthropic.claude-3-5-haiku-20241022-v1:0",
"temperature": 0.1,
"max_tokens": 2000,
"model": os.environ.get("MEM0_LLM_MODEL", "anthropic.claude-3-5-haiku-20241022-v1:0"),
"temperature": float(os.environ.get("MEM0_LLM_TEMPERATURE", 0.1)),
"max_tokens": int(os.environ.get("MEM0_LLM_MAX_TOKENS", 2000)),
},
},
"vector_store": {
Expand Down
13 changes: 12 additions & 1 deletion tests/test_mem0.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -54,7 +54,18 @@ def extract_result_text(result):
return str(result)


@patch.dict(os.environ, {"OPENSEARCH_HOST": "test.opensearch.amazonaws.com"})
@patch.dict(
os.environ,
{
"MEM0_LLM_PROVIDER": "openai",
"MEM0_LLM_MODEL": "gpt-4o",
"MEM0_LLM_TEMPERATURE": "0.2",
"MEM0_LLM_MAX_TOKENS": "4000",
"MEM0_EMBEDDER_PROVIDER": "openai",
"MEM0_EMBEDDER_MODEL": "text-embedding-3-large",
"OPENSEARCH_HOST": "test.opensearch.amazonaws.com",
},
)
@patch("strands_tools.mem0_memory.Mem0Memory")
@patch("strands_tools.mem0_memory.boto3.Session")
def test_store_memory(mock_boto3_session, mock_mem0_memory, mock_tool):
Expand Down
Loading