• matthijs's avatar
    * moved most FAISS_ASSERT calls to C++ exceptions, and adjusted · aa1585cb
    matthijs authored
      memory allocation to avoid mem leaks
    
    * added an IndexIVFScalarQuantizer type that offers an
      intermediate compression between IVFFlat and IVFPQ
    
    * support removal of indices in IndexIDMap / IndexFlat combination
    
    * various fixes in GPU code
    aa1585cb
bench_scalar_quantizer.py 3.33 KB