Cheqi/merge gpu code #159
Merged
Cheqi/merge gpu code #159
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
* try to add an end-to-end tutorial for building vector search online s… (#116) * try to add an end-to-end tutorial for building vector search online service. * update the tutorial and getting start page * fix space and new line issues * fix gcc5 compiling issue (#120) * fix gcc5 compiling issue * fix gcc5 compiling issue
* Small changes before new code... * Fix cmake to work on dtcluster * Debugging and testing * Fixed several bugs, added unit test for internal testing, just need to fix parallel RNG pruning. * Initial GPU refinement code * Fixed bug with refinement * Fixed bug with RNG shrinking function * Stable GPU refinement debugged. * Cleaned up code a bit * Removed specific paths for local build * Initial implementation of "strict" initial RNG construction * Fixed strict RNG construction and added separate options for loose and strict construction * Initial code of new strict RNG construction that reaches refinement accuracy target. * Fix / improve int8 cosine using dp4a intrinsic * Debugging batched execution of RNG graph build * before merge * More work debugging TPT issues for large datasets * Polishing, fixing default values, and removing some scaffolding * before merge * Added more useful error handling and debug info Co-authored-by: Ben Karsin <bkarsin@dtlogin.cm.cluster> Co-authored-by: Ben Karsin <bkarsin@dt03.eth.cluster> Co-authored-by: Ben Karsin <bkarsin@prm-dgx-17.nvidia.com> Co-authored-by: Ben Karsin <bkarsin@prm-dgx-01.nvidia.com> Co-authored-by: Ben Karsin <bkarsin@prom.nvidia.com> Co-authored-by: Ben Karsin <bkarsin@prm-dgx-23.nvidia.com> Co-authored-by: Ben Karsin <bkarsin@prm-dgx-05.nvidia.com> Co-authored-by: Ben Karsin <bkarsin@prm-dgx-04.nvidia.com> Co-authored-by: Ben Karsin <bkarsin@prm-dgx-12.nvidia.com> Co-authored-by: Ben Karsin <bkarsin@prm-dgx-15.nvidia.com>
…PTAG into cheqi/merge_gpu_code
…run on hardware with compute capability < 61 (#153) Co-authored-by: Ben Karsin <bkarsin@dtlogin.cm.cluster>
…PTAG into cheqi/merge_gpu_code
* Added non-intrinsic version of cosine distance so it can be compiled/run on hardware with compute capability < 61 * Add signed int8 support Co-authored-by: Ben Karsin <bkarsin@dtlogin.cm.cluster> Co-authored-by: Ben Karsin <bkarsin@prom.nvidia.com>
SkyInTheSea
reviewed
Aug 10, 2020
| { | ||
| std::lock_guard<std::mutex> lock(m_dataUpdateLock[node]); | ||
|
|
||
| SizeType* nodes = m_pNeighborhoodGraph[node]; |
SkyInTheSea
approved these changes
Aug 10, 2020
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.