danvics/haiku.rag
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Clarify refine_factor, set default to 30

Browse source
This commit is contained in:
Yiorgis Gozadinos 2025-11-21 15:53:24 +02:00
parent 2f54ee7d59
commit 6d8ab2575d
No known key found for this signature in database
3 changed files with 7 additions and 6 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
CHANGELOG.md
View file

@ -7,7 +7,7 @@
- Creates IVF_PQ indexes - Creates IVF_PQ indexes
- Requires minimum 256 chunks (LanceDB training data requirement) - Requires minimum 256 chunks (LanceDB training data requirement)
- New `search.vector_index_metric` config option: `cosine` (default), `l2`, or `dot` - New `search.vector_index_metric` config option: `cosine` (default), `l2`, or `dot`
- New `search.vector_refine_factor` config option (default: 10) for accuracy/speed tradeoff - New `search.vector_refine_factor` config option (default: 30) for accuracy/speed tradeoff
- Indexes not created automatically during ingestion to avoid performance degradation - Indexes not created automatically during ingestion to avoid performance degradation
- Manual rebuilding required after adding significant new data - Manual rebuilding required after adding significant new data
- **Enhanced Info Command**: `haiku-rag info` now shows storage sizes and vector index statistics - **Enhanced Info Command**: `haiku-rag info` now shows storage sizes and vector index statistics

7
docs/configuration.md
View file

@ -88,7 +88,7 @@ research:
search: search:
vector_index_metric: cosine # cosine, l2, or dot vector_index_metric: cosine # cosine, l2, or dot
vector_refine_factor: 10 vector_refine_factor: 30
agui: agui:
host: "0.0.0.0" host: "0.0.0.0"
@ -732,14 +732,15 @@ Configure vector indexing behavior for efficient similarity search:
```yaml ```yaml
search: search:
vector_index_metric: cosine # cosine, l2, or dot vector_index_metric: cosine # cosine, l2, or dot
vector_refine_factor: 10 # Re-ranking factor for accuracy vector_refine_factor: 30 # Re-ranking factor for accuracy
``` ```
- **vector_index_metric**: Distance metric for vector similarity: - **vector_index_metric**: Distance metric for vector similarity:
- `cosine`: Cosine similarity (default, best for most embeddings) - `cosine`: Cosine similarity (default, best for most embeddings)
- `l2`: Euclidean distance - `l2`: Euclidean distance
- `dot`: Dot product similarity - `dot`: Dot product similarity
- **vector_refine_factor**: Retrieve `refine_factor * limit` candidates and re-rank in memory for better accuracy. Higher values increase accuracy but slow down queries. Default: 10 - **vector_refine_factor**: Improves accuracy when using a vector index by retrieving `refine_factor * limit` candidates (using approximate search) and re-ranking them with exact distances. Higher values increase accuracy but slow down queries. Default: 30
- **Only applies with a vector index** - has no effect on brute-force search, which already returns exact results
!!! note !!! note
Vector indexes are only necessary for large datasets with over 100,000 chunks. For smaller datasets, LanceDB's brute-force kNN search provides exact results with good performance. Only create an index if you notice search performance degradation on large datasets. Vector indexes are only necessary for large datasets with over 100,000 chunks. For smaller datasets, LanceDB's brute-force kNN search provides exact results with good performance. Only create an index if you notice search performance degradation on large datasets.

2
haiku_rag_slim/haiku/rag/config/models.py
View file

@ -83,7 +83,7 @@ class ProcessingConfig(BaseModel):
class SearchConfig(BaseModel): class SearchConfig(BaseModel):
vector_index_metric: Literal["cosine", "l2", "dot"] = "cosine" vector_index_metric: Literal["cosine", "l2", "dot"] = "cosine"
vector_refine_factor: int = 10 vector_refine_factor: int = 30
class OllamaConfig(BaseModel): class OllamaConfig(BaseModel):