...
| Code Block | ||||
|---|---|---|---|---|
| ||||
namespace IndexEnum {
const char* INDEX_SCANN = "SCANN";
} |
Design Details
As part of third party library, we will have to add a new library for ScaNN. This directory here: https://github.com/google-research/google-research/tree/master/scann/scann/scann_ops contains APIs for build/search etc. Also, ScaNN uses Bazel as it's dependency management while we would need to use CMake.One the library has been added,
We will create new wrapper classes called IndexScann.h and IndexScann.cpp in knowhere.
| Code Block | ||||
|---|---|---|---|---|
| ||||
class IndexScann : public VecIndex {
public:
IndexScann() {
index_type_ = IndexEnum::INDEX_SCANN;
}
BinarySet
Serialize(const Config& config) override;
void
Load(const BinarySet& index_binary) override;
void
Train(const DatasetPtr& dataset_ptr, const Config& config) override;
void
AddWithoutIds(const DatasetPtr&, const Config&) override;
DatasetPtr
Query(const DatasetPtr& dataset_ptr, const Config& config, const faiss::BitsetView bitset) override;
int64_t
Count() override;
int64_t
Dim() override;
void
UpdateIndexSize() override;
} |
IndexScann.cpp would be the implementation of IndexScann.h.
Compatibility, Deprecation, and Migration Plan
...