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

docs(rerankers): align reranker config docs with v3 SDK source

Browse Source
This commit is contained in:
kartik-mem0
2026-06-25 11:49:24 +05:30
parent ac296f7534
commit 5345ed5e0b
9 changed files with 55 additions and 64 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/components/rerankers/models/cohere.mdx
+2 -2
View File
@@ -74,7 +74,7 @@ os.environ["COHERE_API_KEY"] = "your-api-key"
config = {
"vector_store": {"provider": "chroma"},
"llm": {"provider": "openai", "config": {"model": "gpt-4o-mini"}},
"rerank": {
"reranker": {
"provider": "cohere",
"config": {
"model": "rerank-english-v3.0",
@@ -110,7 +110,7 @@ For multilingual applications, use the multilingual model:
```python Python
config = {
"rerank": {
"reranker": {
"provider": "cohere",
"config": {
"model": "rerank-multilingual-v3.0",
docs/components/rerankers/models/huggingface.mdx
+11 -17
View File
@@ -35,7 +35,7 @@ m = Memory.from_config(config)
| `device` | str | "cpu" | Device to run model on ("cpu", "cuda", "mps") |
| `batch_size` | int | 32 | Batch size for processing |
| `max_length` | int | 512 | Maximum input sequence length |
| `trust_remote_code` | bool | False | Allow remote code execution |
| `normalize` | bool | True | Whether to normalize scores to [0, 1] range |
### Advanced Configuration
@@ -48,10 +48,7 @@ config = {
"device": "cuda",
"batch_size": 16,
"max_length": 512,
"trust_remote_code": False,
"model_kwargs": {
"torch_dtype": "float16"
}
"normalize": True
}
}
}
@@ -154,7 +151,7 @@ m.add("I enjoy reading science fiction books", user_id="alice")
# Search with reranking
results = m.search(
"What outdoor activities do I enjoy?",
user_id="alice",
filters={"user_id": "alice"},
rerank=True
)
@@ -208,10 +205,7 @@ config = {
"model": "BAAI/bge-reranker-base",
"device": "cpu",
"batch_size": 8, # Smaller batch size
"max_length": 256, # Shorter sequences
"model_kwargs": {
"torch_dtype": "float16" # Half precision
}
"max_length": 256 # Shorter sequences
}
}
}
@@ -232,7 +226,7 @@ config = {
try:
results = m.search(
"test query",
user_id="alice",
filters={"user_id": "alice"},
rerank=True
)
except Exception as e:
@@ -240,7 +234,7 @@ except Exception as e:
# Fall back to vector search only
results = m.search(
"test query",
user_id="alice",
filters={"user_id": "alice"},
rerank=False
)
```
@@ -251,13 +245,13 @@ except Exception as e:
```python
# Use a private model from Hugging Face
# Set HF_TOKEN environment variable for private model access
config = {
"reranker": {
"provider": "huggingface",
"config": {
"model": "your-org/custom-reranker",
"device": "cuda",
"use_auth_token": "your-hf-token"
"device": "cuda"
}
}
}
@@ -311,13 +305,13 @@ config = {
import os
os.environ["TRANSFORMERS_CACHE"] = "/path/to/cache"
# Or use offline mode
# Or use offline mode by setting the environment variable
# TRANSFORMERS_OFFLINE=1 before running your script
config = {
"reranker": {
"provider": "huggingface",
"config": {
"model": "BAAI/bge-reranker-base",
"local_files_only": True
"model": "BAAI/bge-reranker-base"
}
}
}
docs/components/rerankers/models/llm.mdx
+5 -5
View File
@@ -40,7 +40,7 @@ config = {
}
},
"reranker": {
"provider": "llm",
"provider": "llm_reranker",
"config": {
"model": "gpt-4o-mini",
"provider": "openai",
@@ -90,7 +90,7 @@ config = {
"vector_store": {"provider": "chroma"},
"llm": {"provider": "openai", "config": {"model": "gpt-4o-mini"}},
"reranker": {
"provider": "llm",
"provider": "llm_reranker",
"config": {
"model": "gpt-4o-mini",
"provider": "openai",
@@ -150,7 +150,7 @@ Score from 0.0 to 1.0. Provide only the numerical score."""
config = {
"reranker": {
"provider": "llm",
"provider": "llm_reranker",
"config": {
"model": "gpt-4o-mini",
"provider": "openai",
@@ -169,7 +169,7 @@ Use different LLM providers for reranking:
# Using Anthropic Claude
anthropic_config = {
"reranker": {
"provider": "llm",
"provider": "llm_reranker",
"config": {
"model": "claude-3-haiku-20240307",
"provider": "anthropic",
@@ -181,7 +181,7 @@ anthropic_config = {
# Using local Ollama model
ollama_config = {
"reranker": {
"provider": "llm",
"provider": "llm_reranker",
"config": {
"model": "llama2:7b",
"provider": "ollama",
docs/components/rerankers/models/llm_reranker.mdx
+1 -1
View File
@@ -268,7 +268,7 @@ m.add("I'm vegetarian", user_id="alice")
# Search with LLM reranking
results = m.search(
"What foods should I avoid?",
user_id="alice",
filters={"user_id": "alice"},
rerank=True
)
docs/components/rerankers/models/sentence_transformer.mdx
+4 -4
View File
@@ -39,7 +39,7 @@ config = {
"model": "gpt-4o-mini"
}
},
"rerank": {
"reranker": {
"provider": "sentence_transformer",
"config": {
"model": "cross-encoder/ms-marco-MiniLM-L-6-v2",
@@ -60,7 +60,7 @@ For better performance, use GPU acceleration:
```python Python
config = {
"rerank": {
"reranker": {
"provider": "sentence_transformer",
"config": {
"model": "cross-encoder/ms-marco-MiniLM-L-6-v2",
@@ -80,7 +80,7 @@ from mem0 import Memory
config = {
"vector_store": {"provider": "chroma"},
"llm": {"provider": "openai", "config": {"model": "gpt-4o-mini"}},
"rerank": {
"reranker": {
"provider": "sentence_transformer",
"config": {
"model": "cross-encoder/ms-marco-MiniLM-L-6-v2",
@@ -117,7 +117,7 @@ You can use any HuggingFace cross-encoder model:
```python Python
# Using a different model
config = {
"rerank": {
"reranker": {
"provider": "sentence_transformer",
"config": {
"model": "cross-encoder/stsb-distilroberta-base",
docs/components/rerankers/models/zero_entropy.mdx
+2 -2
View File
@@ -37,7 +37,7 @@ config = {
"model": "gpt-4o-mini"
}
},
"rerank": {
"reranker": {
"provider": "zero_entropy",
"config": {
"model": "zerank-1", # or "zerank-1-small"
@@ -71,7 +71,7 @@ os.environ["ZERO_ENTROPY_API_KEY"] = "your-api-key"
config = {
"vector_store": {"provider": "chroma"},
"llm": {"provider": "openai", "config": {"model": "gpt-4o-mini"}},
"rerank": {"provider": "zero_entropy", "config": {"model": "zerank-1"}}
"reranker": {"provider": "zero_entropy", "config": {"model": "zerank-1"}}
}
memory = Memory.from_config(config)