my-project-public

repository

loading code, commits, and activity

repositories

loading repo index

#1	---
#2	title: Configurations
#3	---
#4
#5	## How to define configurations?
#6
#7	<Tabs>
#8	<Tab title="Python">
#9	The `config` is defined as a Python dictionary with two main keys:
#10	- `llm`: Specifies the llm provider and its configuration
#11	- `provider`: The name of the llm (e.g., "openai", "groq")
#12	- `config`: A nested dictionary containing provider-specific settings
#13	</Tab>
#14	<Tab title="TypeScript">
#15	The `config` is defined as a TypeScript object with these keys:
#16	- `llm`: Specifies the LLM provider and its configuration (required)
#17	- `provider`: The name of the LLM (e.g., "openai", "groq")
#18	- `config`: A nested object containing provider-specific settings
#19	- `embedder`: Specifies the embedder provider and its configuration (optional)
#20	- `vectorStore`: Specifies the vector store provider and its configuration (optional)
#21	- `historyDbPath`: Path to the history database file (optional)
#22	</Tab>
#23	</Tabs>
#24
#25	### Config Values Precedence
#26
#27	Config values are applied in the following order of precedence (from highest to lowest):
#28
#29	1. Values explicitly set in the `config` object/dictionary
#30	2. Environment variables (e.g., `OPENAI_API_KEY`, `OPENAI_BASE_URL`)
#31	3. Default values defined in the LLM implementation
#32
#33	This means that values specified in the `config` will override corresponding environment variables, which in turn override default values.
#34
#35	## How to Use Config
#36
#37	Here's a general example of how to use the config with Mem0:
#38
#39	<CodeGroup>
#40	```python Python
#41	import os
#42	from mem0 import Memory
#43
#44	os.environ["OPENAI_API_KEY"] = "sk-xx" # for embedder
#45
#46	config = {
#47	"llm": {
#48	"provider": "your_chosen_provider",
#49	"config": {
#50	# Provider-specific settings go here
#51	}
#52	}
#53	}
#54
#55	m = Memory.from_config(config)
#56	m.add("Your text here", user_id="user", metadata={"category": "example"})
#57
#58	```
#59
#60	```typescript TypeScript
#61	import { Memory } from 'mem0ai/oss';
#62
#63	// Minimal configuration with just the LLM settings
#64	const config = {
#65	llm: {
#66	provider: 'your_chosen_provider',
#67	config: {
#68	// Provider-specific settings go here
#69	}
#70	}
#71	};
#72
#73	const memory = new Memory(config);
#74	await memory.add("Your text here", { userId: "user123", metadata: { category: "example" } });
#75	```
#76
#77	</CodeGroup>
#78
#79	## Why is Config Needed?
#80
#81	Config is essential for:
#82	1. Specifying which LLM to use.
#83	2. Providing necessary connection details (e.g., model, api_key, temperature).
#84	3. Ensuring proper initialization and connection to your chosen LLM.
#85
#86	## Master List of All Params in Config
#87
#88	Here's a comprehensive list of all parameters that can be used across different LLMs:
#89
#90	<Tabs>
#91	<Tab title="Python">
#92	\| Parameter \| Description \| Provider \|
#93	\|----------------------\|-----------------------------------------------\|-------------------\|
#94	\| `model` \| Embedding model to use \| All \|
#95	\| `temperature` \| Temperature of the model \| All \|
#96	\| `api_key` \| API key to use \| All \|
#97	\| `max_tokens` \| Tokens to generate \| All \|
#98	\| `top_p` \| Probability threshold for nucleus sampling \| All \|
#99	\| `top_k` \| Number of highest probability tokens to keep \| All \|
#100	\| `http_client_proxies`\| Allow proxy server settings \| AzureOpenAI \|
#101	\| `models` \| List of models \| Openrouter \|
#102	\| `route` \| Routing strategy \| Openrouter \|
#103	\| `openrouter_base_url`\| Base URL for Openrouter API \| Openrouter \|
#104	\| `site_url` \| Site URL \| Openrouter \|
#105	\| `app_name` \| Application name \| Openrouter \|
#106	\| `ollama_base_url` \| Base URL for Ollama API \| Ollama \|
#107	\| `openai_base_url` \| Base URL for OpenAI API \| OpenAI \|
#108	\| `azure_kwargs` \| Azure LLM args for initialization \| AzureOpenAI \|
#109	\| `deepseek_base_url` \| Base URL for DeepSeek API \| DeepSeek \|
#110	\| `xai_base_url` \| Base URL for XAI API \| XAI \|
#111	\| `sarvam_base_url` \| Base URL for Sarvam API \| Sarvam \|
#112	\| `reasoning_effort` \| Reasoning level (low, medium, high) \| Sarvam \|
#113	\| `frequency_penalty` \| Penalize frequent tokens (-2.0 to 2.0) \| Sarvam \|
#114	\| `presence_penalty` \| Penalize existing tokens (-2.0 to 2.0) \| Sarvam \|
#115	\| `seed` \| Seed for deterministic sampling \| Sarvam \|
#116	\| `stop` \| Stop sequences (max 4) \| Sarvam \|
#117	\| `lmstudio_base_url` \| Base URL for LM Studio API \| LM Studio \|
#118	\| `response_callback` \| LLM response callback function \| OpenAI \|
#119	</Tab>
#120	<Tab title="TypeScript">
#121	\| Parameter \| Description \| Provider \|
#122	\|----------------------\|-----------------------------------------------\|-------------------\|
#123	\| `model` \| Embedding model to use \| All \|
#124	\| `temperature` \| Temperature of the model \| All \|
#125	\| `apiKey` \| API key to use \| All \|
#126	\| `maxTokens` \| Tokens to generate \| All \|
#127	\| `topP` \| Probability threshold for nucleus sampling \| All \|
#128	\| `topK` \| Number of highest probability tokens to keep \| All \|
#129	\| `openaiBaseUrl` \| Base URL for OpenAI API \| OpenAI \|
#130	</Tab>
#131	</Tabs>
#132
#133	## Supported LLMs
#134
#135	For detailed information on configuring specific LLMs, please visit the [LLMs](./models) section. There you'll find information for each supported LLM with provider-specific usage examples and configuration details.
#136

z6Mkq5mY3JWtxoxUobWcfNHm7AkRubgSWEZTkBVqZXJviFZ5/my-project-public