Akira Naruse
Akira Naruse
Close #6566. This relatively simple patch can address this problem. However, I think this means that numpy.matmul takes that into account when `dtype` is specified to determine the computation type,...
This is a submission of the NVIDIA team for single GPU in T3 track. After the built index upload is complete, we will update instructions on how to download the...
Close https://github.com/rapidsai/raft/issues/2217 Improve connectivity of CAGRA graphs using approximate degree constrained MST (Minimum Spanning Tree).
It has been empirically shown that the graph indexes generated by CAGRA for search provide comparable search accuracy to other libraries, but reachability from any node to all nodes is...
The current implementation of `refine_host` supports thread parallelism in OpenMP, but the degree of parallelism is the batch size, i.e., the number of queries. Therefore, when the batch size is...
Since CAGRA improves search accuracy by merging forward and reverse graphs, reverse graph is created as part of the graph optimization process. Currently, GPU is used to create the reverse...
CAGRA takes the initial knn graph as input and optimizes it to create a search graph. Several types of processing are performed in the graph optimization, the most memory-intensive of...
This PR addresses the problem of duplicate search results that occur when multiple identical vectors exist in a vector dataset. A relatively straightforward method is to obtain more search results...
CAGRA can guarantee the connectivity of search graphs through a process called MST optimization. This MST optimization uses a GPU for speed, but if the device memory capacity is insufficient,...