Tamas Bela Feher
Tamas Bela Feher
Fixed by https://github.com/rapidsai/cuml/pull/5937
Also we need to handle distance epilogue for NN Descent #2364
> Why would the first dimension change within the graph? Some networks do reshape operations that change the first dim. It is done internally to express some operations more conveniently,...
> can tf-trt dynamic_batch support the situation of being divided into multiple trt subgraphs now? Yes, dynamic shape mode supports graphs that have multiple TRT subgraphs. There is a known...
- Library size in a recent build is 875 MB https://downloads.rapids.ai/ci/cuvs/pull-request/254/91f48f8/cuda11_x86_64.compile_lib.html - Library size after this PR 1.2 GB https://downloads.rapids.ai/ci/cuvs/pull-request/215/04c2745/cuda12_x86_64.compile_lib.html
To clarify: this issue aims to add another overload to `search` that supports [neighbors]( https://github.com/rapidsai/cuvs/blob/7d144cf1285f7113cfbeed68dda5362efb8f2657/cpp/include/cuvs/neighbors/cagra.hpp#L971) argument as ```c++ raft::device_matrix_view neighbors, ``` Simplest way to achive this to add a simple...
The kernels are already compiled and included in `libcuvs.so`. The additional instantiations of the API entry points shall have negligible size. I will confirm this once CI finishes. It was...
Cross referencing FAISS issue https://github.com/facebookresearch/faiss/issues/3621
It seems, that there is a difference in terminology what do we mean by "synchronous". `async_scope`, or `tf.config.experimental.set_synchronous_execution` use a different meaning, compared to what we use in CUDA terminology....