README
¶
✨ Features
- Easy-to-use bindings for Faiss, enabling fast and efficient similarity search of high-dimensional vectors.
- Idiomatic Go bindings for Faiss, allowing seamless integration of Faiss' state-of-the-art vector indexing and search capabilities into Go applications.
- (COMING SOON) Support's GPU acceleration for blazingly fast vector operations.
📦 Installation
go get github.com/Eigen-DB/eigen-db/libs/faissgo
Documentation
¶
Index ¶
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
func GetLastError ¶
func GetLastError() error
Types ¶
type Index ¶
type Index interface {
Train(vecsFlat []float32) error
Add(vecsFlat []float32) error
AddWithIds(vecsFlat []float32, ids []int64) error
RemoveIds(n int64, ids []int64) error
Search(queryVectsFlat []float32, k int64) ([]int64, []float32, error)
Reconstruct(id int64) ([]float32, error) // can be used to fetch a vector for any type of index (i think)
ReconstructN(id int64, n int64) ([]float32, error) // Reconstruct vectors id to id + n - 1
WriteToDisk(path string) error
LoadFromDisk(path string) error
IsTrained() bool
NTotal() int64
Free()
}
func IndexFactory ¶
func IndexFactory(dim int, desc string, metric MetricType) (Index, error)
desc: https://github.com/facebookresearch/faiss/wiki/The-index-factory
type MetricType ¶
type MetricType int
const ( MetricInnerProduct MetricType = C.METRIC_INNER_PRODUCT MetricL2 MetricType = C.METRIC_L2 MetricL1 MetricType = C.METRIC_L1 MetricLinf MetricType = C.METRIC_Linf MetricLp MetricType = C.METRIC_Lp MetricCanberra MetricType = C.METRIC_Canberra MetricBrayCurtis MetricType = C.METRIC_BrayCurtis MetricJensenShannon MetricType = C.METRIC_JensenShannon )
Source Files
Click to show internal directories.
Click to hide internal directories.