JDSS/mem0
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

[docs] Template moulding in docs/platform and index improvement (#3663)

Browse Source
This commit is contained in:
Parth Sharma
2025-10-25 14:05:18 -07:00
committed by GitHub
parent ac9598a67f
commit 61faf71064
15 changed files with 426 additions and 1724 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/open-source/configuration.mdx
+134 -297
View File
@@ -1,336 +1,173 @@
---
title: "Configuration"
description: "Configure Mem0 with custom LLMs, vector stores, embedders, and rerankers for production deployments"
title: "Configure the OSS Stack"
description: "Wire up Mem0 OSS with your preferred LLM, vector store, embedder, and reranker."
icon: "sliders"
iconType: "solid"
---
Mem0 is highly configurable, allowing you to customize every component of your memory system. Choose from **51+ supported providers** across LLMs, vector databases, embedders, and rerankers.
# Configure Mem0 OSS Components
## Quick Start
<Info>
**Prerequisites**
- Python 3.10+ with `pip` available
- Running vector database (e.g., Qdrant, Postgres + pgvector) or access credentials for a managed store
- API keys for your chosen LLM, embedder, and reranker providers
</Info>
The simplest setup uses OpenAI defaults:
<Tip>
Start from the <Link href="/open-source/python-quickstart">Python quickstart</Link> if you still need the base CLI and repository.
</Tip>
```python
import os
from mem0 import Memory
## Install dependencies
os.environ["OPENAI_API_KEY"] = "your-api-key"
m = Memory()
<Tabs>
<Tab title="Python">
<Steps>
<Step title="Install Mem0 OSS">
```bash
pip install mem0ai
```
</Step>
<Step title="Add provider SDKs (example: Qdrant + OpenAI)">
```bash
pip install qdrant-client openai
```
</Step>
</Steps>
</Tab>
<Tab title="Docker Compose">
<Steps>
<Step title="Clone the repo and copy the compose file">
```bash
git clone https://github.com/mem0ai/mem0.git
cd mem0/examples/docker-compose
```
</Step>
<Step title="Install dependencies for local overrides">
```bash
pip install -r requirements.txt
```
</Step>
</Steps>
</Tab>
</Tabs>
For production or custom setups, configure specific components:
## Define your configuration
<Tabs>
<Tab title="Python">
<Steps>
<Step title="Create a configuration dictionary">
```python
from mem0 import Memory
config = {
"vector_store": {
"provider": "qdrant",
"config": {
"host": "localhost",
"port": 6333
}
"config": {"host": "localhost", "port": 6333},
},
"llm": {
"provider": "openai",
"config": {
"model": "gpt-4.1-nano-2025-04-14",
"temperature": 0.1
}
}
}
m = Memory.from_config(config)
```
## Configuration Components
Mem0 has four configurable components. Click any to see all supported providers and detailed configuration options.
<CardGroup cols={2}>
<Card title="LLMs" icon="message-bot" href="/components/llms/overview">
**17 providers** including OpenAI, Anthropic, Ollama, Groq, and more
Configure the language model for memory extraction and processing
</Card>
<Card title="Vector Databases" icon="hard-drive" href="/components/vectordbs/overview">
**25+ databases** including Qdrant, Chroma, Pinecone, Weaviate, and more
Choose where to store and retrieve memory embeddings
</Card>
<Card title="Embedding Models" icon="cube" href="/components/embedders/overview">
**9 providers** including OpenAI, HuggingFace, Ollama, and more
Select the model to convert memories into vector embeddings
</Card>
<Card title="Rerankers" icon="ranking-star" href="/components/rerankers/overview">
**4 models** including Cohere, Zero Entropy, and LLM-based
Improve search relevance by re-scoring retrieved memories
</Card>
</CardGroup>
## Configuration Recipes
### Production Setup with Qdrant
For production deployments, use a dedicated vector store:
<Steps>
<Step title="Start Qdrant">
```bash
docker pull qdrant/qdrant
docker run -p 6333:6333 -p 6334:6334 \
-v $(pwd)/qdrant_storage:/qdrant/storage:z \
qdrant/qdrant
````
</Step>
<Step title="Configure Mem0">
```python
import os
from mem0 import Memory
os.environ["OPENAI_API_KEY"] = "your-api-key"
config = {
"vector_store": {
"provider": "qdrant",
"config": {
"host": "localhost",
"port": 6333,
}
}
}
m = Memory.from_config(config)
````
</Step>
</Steps>
### Fully Local Setup
Run Mem0 completely offline with Ollama (no external APIs):
<Card title="Local Setup with Ollama" icon="server" href="/examples/mem0-with-ollama">
Step-by-step guide to run Mem0 with local LLM and embeddings
</Card>
### Multi-Cloud Setup
Mix providers from different clouds:
```python
config = {
"llm": {
"provider": "azure_openai",
"config": {
"api_key": "azure-key",
"deployment_name": "gpt-4.1-nano-2025-04-14"
}
},
"vector_store": {
"provider": "pinecone",
"config": {
"api_key": "pinecone-key",
"index_name": "mem0"
}
"config": {"model": "gpt-4.1-mini", "temperature": 0.1},
},
"embedder": {
"provider": "vertexai",
"config": {
"model": "textembedding-gecko@003"
}
}
}
```
### Graph Memory Setup
Enable relationship tracking with Neo4j:
```python
config = {
"graph_store": {
"provider": "neo4j",
"config": {
"url": "neo4j+s://your-instance.databases.neo4j.io",
"username": "neo4j",
"password": "your-password"
}
}
}
m = Memory.from_config(config)
```
<Card title="Graph Memory Guide" icon="diagram-project" href="/open-source/features/graph-memory">
Learn how to use graph memory for relationship-based retrieval
</Card>
## Advanced Configuration
### Custom Prompts
Override default prompts for memory processing:
```python
config = {
"custom_fact_extraction_prompt": """
Extract key facts from the conversation.
Focus on: preferences, decisions, and context.
Output as a single sentence.
""",
"custom_update_memory_prompt": """
Update the existing memory with new information.
Preserve important context from the old memory.
"""
}
```
<CardGroup cols={2}>
<Card title="Custom Fact Extraction" icon="sparkles" href="/open-source/features/custom-fact-extraction-prompt">
Customize how memories are extracted from conversations
</Card>
<Card title="Custom Memory Updates" icon="pen" href="/open-source/features/custom-update-memory-prompt">
Control how existing memories are modified
</Card>
</CardGroup>
### Reranking for Better Search
Add reranking to improve search relevance:
```python
config = {
"rerank": {
"config": {"model": "textembedding-gecko@003"},
},
"reranker": {
"provider": "cohere",
"config": {
"model": "rerank-english-v3.0",
"top_k": 5
}
}
"config": {"model": "rerank-english-v3.0"},
},
}
memory = Memory.from_config(config)
```
</Step>
<Step title="Store secrets as environment variables">
```bash
export QDRANT_API_KEY="..."
export OPENAI_API_KEY="..."
export COHERE_API_KEY="..."
```
</Step>
</Steps>
</Tab>
<Tab title="config.yaml">
<Steps>
<Step title="Create a `config.yaml` file">
```yaml
vector_store:
provider: qdrant
config:
host: localhost
port: 6333
<Card title="Reranker-Enhanced Search" icon="arrow-up-arrow-down" href="/open-source/features/reranker-search">
Learn how reranking improves memory search accuracy
</Card>
llm:
provider: azure_openai
config:
api_key: ${AZURE_OPENAI_KEY}
deployment_name: gpt-4.1-mini
### History Database
Configure where operation history is stored:
embedder:
provider: ollama
config:
model: nomic-embed-text
reranker:
provider: zero_entropy
config:
api_key: ${ZERO_ENTROPY_KEY}
```
</Step>
<Step title="Load the config file at runtime">
```python
config = {
"history_db_path": "/custom/path/to/history.db"
}
```
from mem0 import Memory
## All Configuration Options
memory = Memory.from_config_file("config.yaml")
```
</Step>
</Steps>
</Tab>
</Tabs>
<Info icon="check">
Run `memory.add(["Remember my favorite cafe in Tokyo."], user_id="alex")` and then `memory.search("favorite cafe", user_id="alex")`. You should see the Qdrant collection populate and the reranker mark the memory as a top hit.
</Info>
## Tune component settings
<AccordionGroup>
<Accordion title="LLM Configuration">
| Parameter | Description | Provider |
|-----------------------|-----------------------------------------------|-------------------|
| `provider` | LLM provider (e.g., "openai", "anthropic") | All |
| `model` | Model to use | All |
| `temperature` | Temperature of the model (0.0-2.0) | All |
| `api_key` | API key to use | Most |
| `max_tokens` | Maximum tokens to generate | All |
| `top_p` | Nucleus sampling threshold | All |
| `top_k` | Top-k sampling parameter | Some |
| `ollama_base_url` | Base URL for Ollama API | Ollama |
| `openai_base_url` | Base URL for OpenAI API | OpenAI |
| `azure_kwargs` | Azure-specific initialization args | Azure OpenAI |
**See all 17 LLM providers:** [LLMs Overview](/components/llms/overview)
</Accordion>
<Accordion title="Vector Store Configuration">
Common parameters (provider-specific options vary):
| Parameter | Description | Example |
| ----------------- | -------------------------- | ----------- |
| `provider` | Vector store provider | "qdrant" |
| `host` | Host address | "localhost" |
| `port` | Port number | 6333 |
| `collection_name` | Collection/index name | "memories" |
| `api_key` | API key (for cloud stores) | "your-key" |
**See all 25+ vector stores:** [Vector Databases Overview](/components/vectordbs/overview)
</Accordion>
<Accordion title="Embedder Configuration">
| Parameter | Description | Default |
|-------------|---------------------------------|------------------------------|
| `provider` | Embedding provider | "openai" |
| `model` | Embedding model to use | "text-embedding-3-small" |
| `api_key` | API key for embedding service | None |
**See all 9 embedder providers:** [Embedders Overview](/components/embedders/overview)
</Accordion>
<Accordion title="Reranker Configuration">
| Parameter | Description | Example |
|-------------|---------------------------------|------------------------------|
| `provider` | Reranker provider | "cohere" |
| `model` | Reranker model to use | "rerank-english-v3.0" |
| `top_k` | Number of results to return | 5 |
| `api_key` | API key for reranker service | "your-key" |
**See all reranker options:** [Rerankers Overview](/components/rerankers/overview)
</Accordion>
<Accordion title="Graph Store Configuration">
| Parameter | Description | Example |
|-------------|---------------------------------|------------------------------|
| `provider` | Graph store provider | "neo4j" |
| `url` | Connection URL | "neo4j+s://..." |
| `username` | Authentication username | "neo4j" |
| `password` | Authentication password | "your-password" |
**Learn more:** [Graph Memory Overview](/open-source/features/graph-memory)
</Accordion>
<Accordion title="General Configuration">
| Parameter | Description | Default |
|------------------|--------------------------------------|----------------------------|
| `history_db_path` | Path to the history database | "{mem0_dir}/history.db" |
| `custom_fact_extraction_prompt` | Custom prompt for memory extraction | None |
| `custom_update_memory_prompt` | Custom prompt for memory updates | None |
</Accordion>
<Accordion title="Vector store collections">
Name collections explicitly in production (`collection_name`) to isolate tenants and enable per-tenant retention policies.
</Accordion>
<Accordion title="LLM extraction temperature">
Keep extraction temperatures ≤0.2 so advanced memories stay deterministic. Raise it only when you see missing facts.
</Accordion>
<Accordion title="Reranker depth">
Limit `top_k` to 10–20 results; sending more adds latency without meaningful gains.
</Accordion>
</AccordionGroup>
## Next Steps
<Warning>
Mixing managed and self-hosted components? Make sure every outbound provider call happens through a secure network path. Managed rerankers often require outbound internet even if your vector store is on-prem.
</Warning>
<CardGroup cols={3}>
<Card title="Python Quickstart" icon="python" href="/open-source/python-quickstart">
Get started with the Python SDK
</Card>
## Quick recovery
<Card title="Self-Hosting Features" icon="server" href="/open-source/features/overview">
Explore OSS-specific capabilities
</Card>
- Qdrant connection errors → confirm port `6333` is exposed and API key (if set) matches.
- Empty search results → verify the embedder model name; a mismatch causes dimension errors.
- `Unknown reranker` → update the SDK (`pip install --upgrade mem0ai`) to load the latest provider registry.
<Card title="Examples" icon="book" href="/examples">
See configuration examples in action
</Card>
<CardGroup cols={2}>
<Card
title="Pick Providers"
description="Review the LLM, vector store, embedder, and reranker catalogs."
icon="sitemap"
href="/components/llms/overview"
/>
<Card
title="Deploy with Docker Compose"
description="Follow the end-to-end OSS deployment walkthrough."
icon="server"
href="/examples/mem0-with-ollama"
/>
</CardGroup>