Skip to content

Diskann Benchmarking Wrapper #260

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Merged
merged 169 commits into from
Feb 27, 2025
Merged

Diskann Benchmarking Wrapper #260

merged 169 commits into from
Feb 27, 2025

Conversation

tarang-jain
Copy link
Contributor

@tarang-jain tarang-jain commented Jul 29, 2024 •
edited
Loading

Brings DiskANN into cuvs-bench

  • Build and search in-memory DiskANN index
  • Build and search SSD DiskANN index
  • Build a cuvs Vamana index on GPU and serialize it in DiskANN format. Search on CPU using in-memory DiskANN search API.

@tarang-jain tarang-jain added feature request New feature or request non-breaking Introduces a non-breaking change labels Jul 29, 2024
@tarang-jain tarang-jain self-assigned this Jul 29, 2024
@github-actions github-actions bot added the Python label Aug 3, 2024
tarang-jain and others added 12 commits August 5, 2024 11:10
@cjnolet cjnolet requested a review from a team as a code owner February 21, 2025 16:14
@cjnolet
Copy link
Member

cjnolet commented Feb 21, 2025

/ok to test

@tarang-jain
Copy link
Contributor Author

/ok to test

@tarang-jain
Copy link
Contributor Author

/rerun tests

@tarang-jain
Copy link
Contributor Author

/ok to test

@tarang-jain
Copy link
Contributor Author

/ok to test

@tarang-jain
Copy link
Contributor Author

/ok to test

jakirkham
jakirkham reviewed Feb 27, 2025
View reviewed changes
Copy link
Member

@jakirkham jakirkham left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks Tarang! 🙏

Had a couple questions below

Note: There are other occurrences like those commented on. Just picked one of each to simplify the discussion

tarang-jain and others added 2 commits February 27, 2025 12:14
@tarang-jain @jakirkham
Update conda/recipes/cuvs-bench-cpu/meta.yaml
2ef2e35 
Remove libaio from run environment (meta.yaml)

Co-authored-by: jakirkham <[email protected]>
@tarang-jain @jakirkham
Update conda/recipes/cuvs-bench-cpu/meta.yaml
18e2de4 
"rm devel packages from run environment"

Co-authored-by: jakirkham <[email protected]>
@cjnolet
Copy link
Member

cjnolet commented Feb 27, 2025

/ok to test

@tarang-jain
Copy link
Contributor Author

/ok to test

@cjnolet cjnolet removed request for a team and jameslamb February 27, 2025 16:46
@cjnolet
Copy link
Member

cjnolet commented Feb 27, 2025

/merge

@rapids-bot rapids-bot bot merged commit ce321a9 into rapidsai:branch-25.04 Feb 27, 2025
61 checks passed
This was referenced Mar 5, 2025
Closed
Merged
rapids-bot bot pushed a commit that referenced this pull request Mar 6, 2025
@jameslamb
cuvs-bench-cpu: avoid 'mkl' dependency on aarch64 (#750)
f14266c 
Fixes rapidsai/docker#739

#260 introduced a runtime dependency on `mkl` for the `cuvs-bench-cpu` conda package. There are not aarch64 packages for `mkl` on conda-forge, so this makes `cuvs-bench-cpu` impossible to install on aarch64.

This fixes that, by applying the same "only add on x86_64" guard used for `mkl` everywhere else in this project, e.g. like this:

https://github.com/rapidsai/cuvs/blob/89b03493b487910d2125fde6680590adde8e2a95/conda/recipes/cuvs-bench-cpu/meta.yaml#L51

It also proposes adding import tests to the `cuvs-bench-cpu` conda recipe, so issues like this can be caught in CI in the future.

## Notes for Reviewers

I searched for references like this

```shell
git grep -i mkl
```

### How I tested this

Saw lines like this in `conda-python-build` logs:

```text
BUILD START: ['cuvs-bench-cpu-25.04.00a96-py312_250305_g94340bc_96.conda']
...
TEST START: /tmp/conda-bld-output/linux-aarch64/cuvs-bench-cpu-25.04.00a96-py312_250305_g94340bc_96.conda
...
import: 'cuvs_bench'
import: 'cuvs_bench.generate_groundtruth'
import: 'cuvs_bench.get_dataset'
import: 'cuvs_bench.plot'
import: 'cuvs_bench.run'
import: 'cuvs_bench.run.data_export'
import: 'cuvs_bench.split_groundtruth'
import: 'cuvs_bench'
import: 'cuvs_bench.generate_groundtruth'
import: 'cuvs_bench.get_dataset'
import: 'cuvs_bench.plot'
import: 'cuvs_bench.run'
import: 'cuvs_bench.run.data_export'
import: 'cuvs_bench.split_groundtruth'
...
TEST END: /tmp/conda-bld-output/linux-aarch64/cuvs-bench-cpu-25.04.00a96-py312_250305_g94340bc_96.conda
```

([build link](https://github.com/rapidsai/cuvs/actions/runs/13687659537/job/38276160494?pr=750#step:9:5961))

Authors:
  - James Lamb (https://github.com/jameslamb)

Approvers:
  - Gil Forsyth (https://github.com/gforsyth)
  - Divye Gala (https://github.com/divyegala)

URL: #750
jiangyinzuo pushed a commit to jiangyinzuo/cuvs that referenced this pull request Mar 27, 2025
@jameslamb @jiangyinzuo
cuvs-bench-cpu: avoid 'mkl' dependency on aarch64 (rapidsai#750)
cedd565 
Fixes rapidsai/docker#739

rapidsai#260 introduced a runtime dependency on `mkl` for the `cuvs-bench-cpu` conda package. There are not aarch64 packages for `mkl` on conda-forge, so this makes `cuvs-bench-cpu` impossible to install on aarch64.

This fixes that, by applying the same "only add on x86_64" guard used for `mkl` everywhere else in this project, e.g. like this:

https://github.com/rapidsai/cuvs/blob/89b03493b487910d2125fde6680590adde8e2a95/conda/recipes/cuvs-bench-cpu/meta.yaml#L51

It also proposes adding import tests to the `cuvs-bench-cpu` conda recipe, so issues like this can be caught in CI in the future.

## Notes for Reviewers

I searched for references like this

```shell
git grep -i mkl
```

### How I tested this

Saw lines like this in `conda-python-build` logs:

```text
BUILD START: ['cuvs-bench-cpu-25.04.00a96-py312_250305_g94340bc_96.conda']
...
TEST START: /tmp/conda-bld-output/linux-aarch64/cuvs-bench-cpu-25.04.00a96-py312_250305_g94340bc_96.conda
...
import: 'cuvs_bench'
import: 'cuvs_bench.generate_groundtruth'
import: 'cuvs_bench.get_dataset'
import: 'cuvs_bench.plot'
import: 'cuvs_bench.run'
import: 'cuvs_bench.run.data_export'
import: 'cuvs_bench.split_groundtruth'
import: 'cuvs_bench'
import: 'cuvs_bench.generate_groundtruth'
import: 'cuvs_bench.get_dataset'
import: 'cuvs_bench.plot'
import: 'cuvs_bench.run'
import: 'cuvs_bench.run.data_export'
import: 'cuvs_bench.split_groundtruth'
...
TEST END: /tmp/conda-bld-output/linux-aarch64/cuvs-bench-cpu-25.04.00a96-py312_250305_g94340bc_96.conda
```

([build link](https://github.com/rapidsai/cuvs/actions/runs/13687659537/job/38276160494?pr=750#step:9:5961))

Authors:
  - James Lamb (https://github.com/jameslamb)

Approvers:
  - Gil Forsyth (https://github.com/gforsyth)
  - Divye Gala (https://github.com/divyegala)

URL: rapidsai#750
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@jakirkham jakirkham jakirkham left review comments

@bdice bdice bdice left review comments

@achirkin achirkin achirkin approved these changes

@cjnolet cjnolet cjnolet approved these changes

@gforsyth gforsyth gforsyth approved these changes

@bkarsin bkarsin bkarsin approved these changes

Assignees

@tarang-jain tarang-jain

Labels
benchmarking CMake cpp feature request New feature or request non-breaking Introduces a non-breaking change Python
Projects
Vector Search, ML, & Data Mining Release Board
Status: Done
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
8 participants
@tarang-jain @cjnolet @gforsyth @jakirkham @bdice @bkarsin @achirkin @dantegd