TBNilles/LEANN
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Initial commit

Browse Source
This commit is contained in:
yichuan520030910320
2025-06-30 09:05:05 +00:00
commit 46f6cc100b
1231 changed files with 278432 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

81
packages/leann-backend-diskann/third_party/DiskANN/.github/workflows/build-python-pdoc.yml vendored Normal file
View File

@@ -0,0 +1,81 @@
name: DiskANN Build PDoc Documentation
on: [workflow_call]
jobs:
build-reference-documentation:
permissions:
contents: write
runs-on: ubuntu-latest
steps:
- name: Checkout repository
uses: actions/checkout@v3
with:
fetch-depth: 1
- name: Set up Python 3.9
uses: actions/setup-python@v2
with:
python-version: 3.9
- name: Install python build
run: python -m pip install build
shell: bash
# Install required dependencies
- name: Prepare Linux environment
run: |
sudo scripts/dev/install-dev-deps-ubuntu.bash
shell: bash
# We need to build the wheel in order to run pdoc. pdoc does not seem to work if you just point it at
# our source directory.
- name: Building Python Wheel for documentation generation
run: python -m build --wheel --outdir documentation_dist
shell: bash
- name: "Run Reference Documentation Generation"
run: |
pip install pdoc pipdeptree
pip install documentation_dist/*.whl
echo "documentation" > dependencies_documentation.txt
pipdeptree >> dependencies_documentation.txt
pdoc -o docs/python/html diskannpy
- name: Create version environment variable
run: |
echo "DISKANN_VERSION=$(python <<EOF
from importlib.metadata import version
v = version('diskannpy')
print(v)
EOF
)" >> $GITHUB_ENV
- name: Archive documentation version artifact
uses: actions/upload-artifact@v4
with:
name: dependencies
path: |
${{ github.run_id }}-dependencies_documentation.txt
overwrite: true
- name: Archive documentation artifacts
uses: actions/upload-artifact@v4
with:
name: documentation-site
path: |
docs/python/html
# Publish to /dev if we are on the "main" branch
- name: Publish reference docs for latest development version (main branch)
uses: peaceiris/actions-gh-pages@v3
if: github.ref == 'refs/heads/main'
with:
github_token: ${{ secrets.GITHUB_TOKEN }}
publish_dir: docs/python/html
destination_dir: docs/python/dev
# Publish to /<version> if we are releasing
- name: Publish reference docs by version (main branch)
uses: peaceiris/actions-gh-pages@v3
if: github.event_name == 'release'
with:
github_token: ${{ secrets.GITHUB_TOKEN }}
publish_dir: docs/python/html
destination_dir: docs/python/${{ env.DISKANN_VERSION }}
# Publish to /latest if we are releasing
- name: Publish latest reference docs (main branch)
uses: peaceiris/actions-gh-pages@v3
if: github.event_name == 'release'
with:
github_token: ${{ secrets.GITHUB_TOKEN }}
publish_dir: docs/python/html
destination_dir: docs/python/latest

42
packages/leann-backend-diskann/third_party/DiskANN/.github/workflows/build-python.yml vendored Normal file
View File

@@ -0,0 +1,42 @@
name: DiskANN Build Python Wheel
on: [workflow_call]
jobs:
linux-build:
name: Python - Ubuntu - ${{matrix.cibw-identifier}}
strategy:
fail-fast: false
matrix:
cibw-identifier: ["cp39-manylinux_x86_64", "cp310-manylinux_x86_64", "cp311-manylinux_x86_64"]
runs-on: ubuntu-latest
defaults:
run:
shell: bash
steps:
- name: Checkout repository
uses: actions/checkout@v3
with:
fetch-depth: 1
- name: Building python wheel ${{matrix.cibw-identifier}}
uses: ./.github/actions/python-wheel
with:
cibw-identifier: ${{matrix.cibw-identifier}}
windows-build:
name: Python - Windows - ${{matrix.cibw-identifier}}
strategy:
fail-fast: false
matrix:
cibw-identifier: ["cp39-win_amd64", "cp310-win_amd64", "cp311-win_amd64"]
runs-on: windows-latest
defaults:
run:
shell: bash
steps:
- name: Checkout repository
uses: actions/checkout@v3
with:
submodules: true
fetch-depth: 1
- name: Building python wheel ${{matrix.cibw-identifier}}
uses: ./.github/actions/python-wheel
with:
cibw-identifier: ${{matrix.cibw-identifier}}

28
packages/leann-backend-diskann/third_party/DiskANN/.github/workflows/common.yml vendored Normal file
View File

@@ -0,0 +1,28 @@
name: DiskANN Common Checks
# common means common to both pr-test and push-test
on: [workflow_call]
jobs:
formatting-check:
strategy:
fail-fast: true
name: Code Formatting Test
runs-on: ubuntu-latest
steps:
- name: Checkout repository
uses: actions/checkout@v3
with:
fetch-depth: 1
- name: Checking code formatting...
uses: ./.github/actions/format-check
docker-container-build:
name: Docker Container Build
needs: [formatting-check]
runs-on: ubuntu-latest
steps:
- name: Checkout repository
uses: actions/checkout@v3
with:
fetch-depth: 1
- name: Docker build
run: |
docker build .

117
packages/leann-backend-diskann/third_party/DiskANN/.github/workflows/disk-pq.yml vendored Normal file
View File

@@ -0,0 +1,117 @@
name: Disk With PQ
on: [workflow_call]
jobs:
acceptance-tests-disk-pq:
name: Disk, PQ
strategy:
fail-fast: false
matrix:
os: [ubuntu-latest, windows-2019, windows-latest]
runs-on: ${{matrix.os}}
defaults:
run:
shell: bash
steps:
- name: Checkout repository
if: ${{ runner.os == 'Linux' }}
uses: actions/checkout@v3
with:
fetch-depth: 1
- name: Checkout repository
if: ${{ runner.os == 'Windows' }}
uses: actions/checkout@v3
with:
fetch-depth: 1
submodules: true
- name: DiskANN Build CLI Applications
uses: ./.github/actions/build
- name: Generate Data
uses: ./.github/actions/generate-random
- name: build and search disk index (one shot graph build, L2, no diskPQ) (float)
if: success() || failure()
run: |
dist/bin/build_disk_index --data_type float --dist_fn l2 --data_path data/rand_float_10D_10K_norm1.0.bin --index_path_prefix data/disk_index_l2_rand_float_10D_10K_norm1.0_diskfull_oneshot -R 16 -L 32 -B 0.00003 -M 1
dist/bin/search_disk_index --data_type float --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/disk_index_l2_rand_float_10D_10K_norm1.0_diskfull_oneshot --result_path /tmp/res --query_file data/rand_float_10D_1K_norm1.0.bin --gt_file data/l2_rand_float_10D_10K_norm1.0_10D_1K_norm1.0_gt100 --recall_at 5 -L 5 12 -W 2 --num_nodes_to_cache 10 -T 16
- name: build and search disk index (one shot graph build, cosine, no diskPQ) (float)
if: success() || failure()
run: |
dist/bin/build_disk_index --data_type float --dist_fn cosine --data_path data/rand_float_10D_10K_unnorm.bin --index_path_prefix data/disk_index_cosine_rand_float_10D_10K_unnorm_diskfull_oneshot -R 16 -L 32 -B 0.00003 -M 1
dist/bin/search_disk_index --data_type float --dist_fn cosine --fail_if_recall_below 70 --index_path_prefix data/disk_index_cosine_rand_float_10D_10K_unnorm_diskfull_oneshot --result_path /tmp/res --query_file data/rand_float_10D_1K_unnorm.bin --gt_file data/cosine_rand_float_10D_10K_unnorm_10D_1K_unnorm_gt100 --recall_at 5 -L 5 12 -W 2 --num_nodes_to_cache 10 -T 16
- name: build and search disk index (one shot graph build, L2, no diskPQ) (int8)
if: success() || failure()
run: |
dist/bin/build_disk_index --data_type int8 --dist_fn l2 --data_path data/rand_int8_10D_10K_norm50.0.bin --index_path_prefix data/disk_index_l2_rand_int8_10D_10K_norm50.0_diskfull_oneshot -R 16 -L 32 -B 0.00003 -M 1
dist/bin/search_disk_index --data_type int8 --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/disk_index_l2_rand_int8_10D_10K_norm50.0_diskfull_oneshot --result_path /tmp/res --query_file data/rand_int8_10D_1K_norm50.0.bin --gt_file data/l2_rand_int8_10D_10K_norm50.0_10D_1K_norm50.0_gt100 --recall_at 5 -L 5 12 -W 2 --num_nodes_to_cache 10 -T 16
- name: build and search disk index (one shot graph build, L2, no diskPQ) (uint8)
if: success() || failure()
run: |
dist/bin/build_disk_index --data_type uint8 --dist_fn l2 --data_path data/rand_uint8_10D_10K_norm50.0.bin --index_path_prefix data/disk_index_l2_rand_uint8_10D_10K_norm50.0_diskfull_oneshot -R 16 -L 32 -B 0.00003 -M 1
dist/bin/search_disk_index --data_type uint8 --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/disk_index_l2_rand_uint8_10D_10K_norm50.0_diskfull_oneshot --result_path /tmp/res --query_file data/rand_uint8_10D_1K_norm50.0.bin --gt_file data/l2_rand_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100 --recall_at 5 -L 5 12 -W 2 --num_nodes_to_cache 10 -T 16
- name: build and search disk index (one shot graph build, L2, no diskPQ, build with PQ distance comparisons) (float)
if: success() || failure()
run: |
dist/bin/build_disk_index --data_type float --dist_fn l2 --data_path data/rand_float_10D_10K_norm1.0.bin --index_path_prefix data/disk_index_l2_rand_float_10D_10K_norm1.0_diskfull_oneshot_buildpq5 -R 16 -L 32 -B 0.00003 -M 1 --build_PQ_bytes 5
dist/bin/search_disk_index --data_type float --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/disk_index_l2_rand_float_10D_10K_norm1.0_diskfull_oneshot_buildpq5 --result_path /tmp/res --query_file data/rand_float_10D_1K_norm1.0.bin --gt_file data/l2_rand_float_10D_10K_norm1.0_10D_1K_norm1.0_gt100 --recall_at 5 -L 5 12 -W 2 --num_nodes_to_cache 10 -T 16
- name: build and search disk index (one shot graph build, L2, no diskPQ, build with PQ distance comparisons) (int8)
if: success() || failure()
run: |
dist/bin/build_disk_index --data_type int8 --dist_fn l2 --data_path data/rand_int8_10D_10K_norm50.0.bin --index_path_prefix data/disk_index_l2_rand_int8_10D_10K_norm50.0_diskfull_oneshot_buildpq5 -R 16 -L 32 -B 0.00003 -M 1 --build_PQ_bytes 5
dist/bin/search_disk_index --data_type int8 --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/disk_index_l2_rand_int8_10D_10K_norm50.0_diskfull_oneshot_buildpq5 --result_path /tmp/res --query_file data/rand_int8_10D_1K_norm50.0.bin --gt_file data/l2_rand_int8_10D_10K_norm50.0_10D_1K_norm50.0_gt100 --recall_at 5 -L 5 12 -W 2 --num_nodes_to_cache 10 -T 16\
- name: build and search disk index (one shot graph build, L2, no diskPQ, build with PQ distance comparisons) (uint8)
if: success() || failure()
run: |
dist/bin/build_disk_index --data_type uint8 --dist_fn l2 --data_path data/rand_uint8_10D_10K_norm50.0.bin --index_path_prefix data/disk_index_l2_rand_uint8_10D_10K_norm50.0_diskfull_oneshot_buildpq5 -R 16 -L 32 -B 0.00003 -M 1 --build_PQ_bytes 5
dist/bin/search_disk_index --data_type uint8 --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/disk_index_l2_rand_uint8_10D_10K_norm50.0_diskfull_oneshot_buildpq5 --result_path /tmp/res --query_file data/rand_uint8_10D_1K_norm50.0.bin --gt_file data/l2_rand_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100 --recall_at 5 -L 5 12 -W 2 --num_nodes_to_cache 10 -T 16
- name: build and search disk index (sharded graph build, L2, no diskPQ) (float)
if: success() || failure()
run: |
dist/bin/build_disk_index --data_type float --dist_fn l2 --data_path data/rand_float_10D_10K_norm1.0.bin --index_path_prefix data/disk_index_l2_rand_float_10D_10K_norm1.0_diskfull_sharded -R 16 -L 32 -B 0.00003 -M 0.00006
dist/bin/search_disk_index --data_type float --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/disk_index_l2_rand_float_10D_10K_norm1.0_diskfull_sharded --result_path /tmp/res --query_file data/rand_float_10D_1K_norm1.0.bin --gt_file data/l2_rand_float_10D_10K_norm1.0_10D_1K_norm1.0_gt100 --recall_at 5 -L 5 12 -W 2 --num_nodes_to_cache 10 -T 16
- name: build and search disk index (sharded graph build, cosine, no diskPQ) (float)
if: success() || failure()
run: |
dist/bin/build_disk_index --data_type float --dist_fn cosine --data_path data/rand_float_10D_10K_unnorm.bin --index_path_prefix data/disk_index_cosine_rand_float_10D_10K_unnorm_diskfull_sharded -R 16 -L 32 -B 0.00003 -M 0.00006
dist/bin/search_disk_index --data_type float --dist_fn cosine --fail_if_recall_below 70 --index_path_prefix data/disk_index_cosine_rand_float_10D_10K_unnorm_diskfull_sharded --result_path /tmp/res --query_file data/rand_float_10D_1K_unnorm.bin --gt_file data/cosine_rand_float_10D_10K_unnorm_10D_1K_unnorm_gt100 --recall_at 5 -L 5 12 -W 2 --num_nodes_to_cache 10 -T 16
- name: build and search disk index (sharded graph build, L2, no diskPQ) (int8)
run: |
dist/bin/build_disk_index --data_type int8 --dist_fn l2 --data_path data/rand_int8_10D_10K_norm50.0.bin --index_path_prefix data/disk_index_l2_rand_int8_10D_10K_norm50.0_diskfull_sharded -R 16 -L 32 -B 0.00003 -M 0.00006
dist/bin/search_disk_index --data_type int8 --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/disk_index_l2_rand_int8_10D_10K_norm50.0_diskfull_sharded --result_path /tmp/res --query_file data/rand_int8_10D_1K_norm50.0.bin --gt_file data/l2_rand_int8_10D_10K_norm50.0_10D_1K_norm50.0_gt100 --recall_at 5 -L 5 12 -W 2 --num_nodes_to_cache 10 -T 16
- name: build and search disk index (sharded graph build, L2, no diskPQ) (uint8)
if: success() || failure()
run: |
dist/bin/build_disk_index --data_type uint8 --dist_fn l2 --data_path data/rand_uint8_10D_10K_norm50.0.bin --index_path_prefix data/disk_index_l2_rand_uint8_10D_10K_norm50.0_diskfull_sharded -R 16 -L 32 -B 0.00003 -M 0.00006
dist/bin/search_disk_index --data_type uint8 --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/disk_index_l2_rand_uint8_10D_10K_norm50.0_diskfull_sharded --result_path /tmp/res --query_file data/rand_uint8_10D_1K_norm50.0.bin --gt_file data/l2_rand_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100 --recall_at 5 -L 5 12 -W 2 --num_nodes_to_cache 10 -T 16
- name: build and search disk index (one shot graph build, L2, diskPQ) (float)
if: success() || failure()
run: |
dist/bin/build_disk_index --data_type float --dist_fn l2 --data_path data/rand_float_10D_10K_norm1.0.bin --index_path_prefix data/disk_index_l2_rand_float_10D_10K_norm1.0_diskpq_oneshot -R 16 -L 32 -B 0.00003 -M 1 --PQ_disk_bytes 5
dist/bin/search_disk_index --data_type float --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/disk_index_l2_rand_float_10D_10K_norm1.0_diskpq_oneshot --result_path /tmp/res --query_file data/rand_float_10D_1K_norm1.0.bin --gt_file data/l2_rand_float_10D_10K_norm1.0_10D_1K_norm1.0_gt100 --recall_at 5 -L 5 12 -W 2 --num_nodes_to_cache 10 -T 16
- name: build and search disk index (one shot graph build, L2, diskPQ) (int8)
if: success() || failure()
run: |
dist/bin/build_disk_index --data_type int8 --dist_fn l2 --data_path data/rand_int8_10D_10K_norm50.0.bin --index_path_prefix data/disk_index_l2_rand_int8_10D_10K_norm50.0_diskpq_oneshot -R 16 -L 32 -B 0.00003 -M 1 --PQ_disk_bytes 5
dist/bin/search_disk_index --data_type int8 --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/disk_index_l2_rand_int8_10D_10K_norm50.0_diskpq_oneshot --result_path /tmp/res --query_file data/rand_int8_10D_1K_norm50.0.bin --gt_file data/l2_rand_int8_10D_10K_norm50.0_10D_1K_norm50.0_gt100 --recall_at 5 -L 5 12 -W 2 --num_nodes_to_cache 10 -T 16
- name: build and search disk index (one shot graph build, L2, diskPQ) (uint8)
if: success() || failure()
run: |
dist/bin/build_disk_index --data_type uint8 --dist_fn l2 --data_path data/rand_uint8_10D_10K_norm50.0.bin --index_path_prefix data/disk_index_l2_rand_uint8_10D_10K_norm50.0_diskpq_oneshot -R 16 -L 32 -B 0.00003 -M 1 --PQ_disk_bytes 5
dist/bin/search_disk_index --data_type uint8 --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/disk_index_l2_rand_uint8_10D_10K_norm50.0_diskpq_oneshot --result_path /tmp/res --query_file data/rand_uint8_10D_1K_norm50.0.bin --gt_file data/l2_rand_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100 --recall_at 5 -L 5 12 -W 2 --num_nodes_to_cache 10 -T 16
- name: build and search disk index (sharded graph build, MIPS, diskPQ) (float)
if: success() || failure()
run: |
dist/bin/build_disk_index --data_type float --dist_fn mips --data_path data/rand_float_10D_10K_norm1.0.bin --index_path_prefix data/disk_index_mips_rand_float_10D_10K_norm1.0_diskpq_sharded -R 16 -L 32 -B 0.00003 -M 0.00006 --PQ_disk_bytes 5
dist/bin/search_disk_index --data_type float --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/disk_index_mips_rand_float_10D_10K_norm1.0_diskpq_sharded --result_path /tmp/res --query_file data/rand_float_10D_1K_norm1.0.bin --gt_file data/mips_rand_float_10D_10K_norm1.0_10D_1K_norm1.0_gt100 --recall_at 5 -L 5 12 -W 2 --num_nodes_to_cache 10 -T 16
- name: upload data and bin
uses: actions/upload-artifact@v4
with:
name: disk-pq-${{matrix.os}}
path: |
./dist/**
./data/**

102
packages/leann-backend-diskann/third_party/DiskANN/.github/workflows/dynamic-labels.yml vendored Normal file
View File

@@ -0,0 +1,102 @@
name: Dynamic-Labels
on: [workflow_call]
jobs:
acceptance-tests-dynamic:
name: Dynamic-Labels
strategy:
fail-fast: false
matrix:
os: [ubuntu-latest, windows-2019, windows-latest]
runs-on: ${{matrix.os}}
defaults:
run:
shell: bash
steps:
- name: Checkout repository
if: ${{ runner.os == 'Linux' }}
uses: actions/checkout@v3
with:
fetch-depth: 1
- name: Checkout repository
if: ${{ runner.os == 'Windows' }}
uses: actions/checkout@v3
with:
fetch-depth: 1
submodules: true
- name: DiskANN Build CLI Applications
uses: ./.github/actions/build
- name: Generate Data
uses: ./.github/actions/generate-random
- name: Generate Labels
run: |
echo "Generating synthetic labels and computing ground truth for filtered search with universal label"
dist/bin/generate_synthetic_labels --num_labels 50 --num_points 10000 --output_file data/rand_labels_50_10K.txt --distribution_type random
echo "Generating synthetic labels with a zipf distribution and computing ground truth for filtered search with universal label"
dist/bin/generate_synthetic_labels --num_labels 50 --num_points 10000 --output_file data/zipf_labels_50_10K.txt --distribution_type zipf
- name: Test a streaming index (float) with labels (Zipf distributed)
run: |
dist/bin/test_streaming_scenario --data_type float --dist_fn l2 --data_path data/rand_float_10D_10K_norm1.0.bin --universal_label 0 --label_file data/zipf_labels_50_10K.txt --index_path_prefix data/index_zipf_stream -R 64 --FilteredLbuild 200 -L 50 --alpha 1.2 --insert_threads 8 --consolidate_threads 8 --max_points_to_insert 10000 --active_window 4000 --consolidate_interval 2000 --start_point_norm 3.2 --unique_labels_supported 51
echo "Computing groundtruth with filter"
dist/bin/compute_groundtruth_for_filters --data_type float --universal_label 0 --filter_label 1 --dist_fn l2 --base_file data/index_zipf_stream.after-streaming-act4000-cons2000-max10000.data --query_file data/rand_float_10D_1K_norm1.0.bin --K 100 --gt_file data/gt100_zipf_base-act4000-cons2000-max10000_1 --label_file data/index_zipf_stream.after-streaming-act4000-cons2000-max10000_raw_labels.txt --tags_file data/index_zipf_stream.after-streaming-act4000-cons2000-max10000.tags
echo "Searching with filter"
dist/bin/search_memory_index --data_type float --dist_fn l2 --filter_label 1 --fail_if_recall_below 40 --index_path_prefix data/index_zipf_stream.after-streaming-act4000-cons2000-max10000 --result_path data/res_stream --query_file data/rand_float_10D_1K_norm1.0.bin --gt_file data/gt100_zipf_base-act4000-cons2000-max10000_1 -K 10 -L 20 40 60 80 100 150 -T 64 --dynamic true --tags 1
echo "Computing groundtruth w/o filter"
dist/bin/compute_groundtruth --data_type float --dist_fn l2 --base_file data/index_zipf_stream.after-streaming-act4000-cons2000-max10000.data --query_file data/rand_float_10D_1K_norm1.0.bin --K 100 --gt_file data/gt100_zipf_base-act4000-cons2000-max10000
echo "Searching without filter"
dist/bin/search_memory_index --data_type float --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/index_zipf_stream.after-streaming-act4000-cons2000-max10000 --result_path res_stream --query_file data/rand_float_10D_1K_norm1.0.bin --gt_file data/gt100_zipf_base-act4000-cons2000-max10000 -K 10 -L 20 40 60 80 100 -T 64
- name: Test a streaming index (float) with labels (random distributed)
run: |
dist/bin/test_streaming_scenario --data_type float --dist_fn l2 --data_path data/rand_float_10D_10K_norm1.0.bin --universal_label 0 --label_file data/rand_labels_50_10K.txt --index_path_prefix data/index_rand_stream -R 64 --FilteredLbuild 200 -L 50 --alpha 1.2 --insert_threads 8 --consolidate_threads 8 --max_points_to_insert 10000 --active_window 4000 --consolidate_interval 2000 --start_point_norm 3.2 --unique_labels_supported 51
echo "Computing groundtruth with filter"
dist/bin/compute_groundtruth_for_filters --data_type float --universal_label 0 --filter_label 1 --dist_fn l2 --base_file data/index_rand_stream.after-streaming-act4000-cons2000-max10000.data --query_file data/rand_float_10D_1K_norm1.0.bin --K 100 --gt_file data/gt100_rand_base-act4000-cons2000-max10000_1 --label_file data/index_rand_stream.after-streaming-act4000-cons2000-max10000_raw_labels.txt --tags_file data/index_rand_stream.after-streaming-act4000-cons2000-max10000.tags
echo "Searching with filter"
dist/bin/search_memory_index --data_type float --dist_fn l2 --filter_label 1 --fail_if_recall_below 40 --index_path_prefix data/index_rand_stream.after-streaming-act4000-cons2000-max10000 --result_path data/res_stream --query_file data/rand_float_10D_1K_norm1.0.bin --gt_file data/gt100_rand_base-act4000-cons2000-max10000_1 -K 10 -L 20 40 60 80 100 150 -T 64 --dynamic true --tags 1
echo "Computing groundtruth w/o filter"
dist/bin/compute_groundtruth --data_type float --dist_fn l2 --base_file data/index_rand_stream.after-streaming-act4000-cons2000-max10000.data --query_file data/rand_float_10D_1K_norm1.0.bin --K 100 --gt_file data/gt100_rand_base-act4000-cons2000-max10000
echo "Searching without filter"
dist/bin/search_memory_index --data_type float --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/index_rand_stream.after-streaming-act4000-cons2000-max10000 --result_path res_stream --query_file data/rand_float_10D_1K_norm1.0.bin --gt_file data/gt100_rand_base-act4000-cons2000-max10000 -K 10 -L 20 40 60 80 100 -T 64
- name: Test Insert Delete Consolidate (float) with labels (zipf distributed)
run: |
dist/bin/test_insert_deletes_consolidate --data_type float --dist_fn l2 --universal_label 0 --label_file data/zipf_labels_50_10K.txt --FilteredLbuild 70 --data_path data/rand_float_10D_10K_norm1.0.bin --index_path_prefix data/index_zipf_ins_del -R 64 -L 10 --alpha 1.2 --points_to_skip 0 --max_points_to_insert 7500 --beginning_index_size 0 --points_per_checkpoint 1000 --checkpoints_per_snapshot 0 --points_to_delete_from_beginning 2500 --start_deletes_after 5000 --do_concurrent true --start_point_norm 3.2 --unique_labels_supported 51
echo "Computing groundtruth with filter"
dist/bin/compute_groundtruth_for_filters --data_type float --filter_label 5 --universal_label 0 --dist_fn l2 --base_file data/index_zipf_ins_del.after-concurrent-delete-del2500-7500.data --query_file data/rand_float_10D_1K_norm1.0.bin --K 100 --gt_file data/gt100_zipf_random10D_1K_wlabel_5 --label_file data/index_zipf_ins_del.after-concurrent-delete-del2500-7500_raw_labels.txt --tags_file data/index_zipf_ins_del.after-concurrent-delete-del2500-7500.tags
echo "Searching with filter"
dist/bin/search_memory_index --data_type float --dist_fn l2 --filter_label 5 --fail_if_recall_below 10 --index_path_prefix data/index_zipf_ins_del.after-concurrent-delete-del2500-7500 --result_path data/res_zipf_stream --query_file data/rand_float_10D_1K_norm1.0.bin --gt_file data/gt100_zipf_random10D_1K_wlabel_5 -K 10 -L 20 40 60 80 100 150 -T 64 --dynamic true --tags 1
echo "Computing groundtruth w/o filter"
dist/bin/compute_groundtruth --data_type float --dist_fn l2 --base_file data/index_zipf_ins_del.after-concurrent-delete-del2500-7500.data --query_file data/rand_float_10D_1K_norm1.0.bin --K 100 --gt_file data/gt100_zipf_random10D_1K
echo "Searching without filter"
dist/bin/search_memory_index --data_type float --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/index_zipf_ins_del.after-concurrent-delete-del2500-7500 --result_path res_stream --query_file data/rand_float_10D_1K_norm1.0.bin --gt_file data/gt100_zipf_random10D_1K -K 10 -L 20 40 60 80 100 -T 64
- name: Test Insert Delete Consolidate (float) with labels (random distributed)
run: |
dist/bin/test_insert_deletes_consolidate --data_type float --dist_fn l2 --universal_label 0 --label_file data/rand_labels_50_10K.txt --FilteredLbuild 70 --data_path data/rand_float_10D_10K_norm1.0.bin --index_path_prefix data/index_rand_ins_del -R 64 -L 10 --alpha 1.2 --points_to_skip 0 --max_points_to_insert 7500 --beginning_index_size 0 --points_per_checkpoint 1000 --checkpoints_per_snapshot 0 --points_to_delete_from_beginning 2500 --start_deletes_after 5000 --do_concurrent true --start_point_norm 3.2 --unique_labels_supported 51
echo "Computing groundtruth with filter"
dist/bin/compute_groundtruth_for_filters --data_type float --filter_label 5 --universal_label 0 --dist_fn l2 --base_file data/index_rand_ins_del.after-concurrent-delete-del2500-7500.data --query_file data/rand_float_10D_1K_norm1.0.bin --K 100 --gt_file data/gt100_rand_random10D_1K_wlabel_5 --label_file data/index_rand_ins_del.after-concurrent-delete-del2500-7500_raw_labels.txt --tags_file data/index_rand_ins_del.after-concurrent-delete-del2500-7500.tags
echo "Searching with filter"
dist/bin/search_memory_index --data_type float --dist_fn l2 --filter_label 5 --fail_if_recall_below 40 --index_path_prefix data/index_rand_ins_del.after-concurrent-delete-del2500-7500 --result_path data/res_rand_stream --query_file data/rand_float_10D_1K_norm1.0.bin --gt_file data/gt100_rand_random10D_1K_wlabel_5 -K 10 -L 20 40 60 80 100 150 -T 64 --dynamic true --tags 1
echo "Computing groundtruth w/o filter"
dist/bin/compute_groundtruth --data_type float --dist_fn l2 --base_file data/index_rand_ins_del.after-concurrent-delete-del2500-7500.data --query_file data/rand_float_10D_1K_norm1.0.bin --K 100 --gt_file data/gt100_rand_random10D_1K
echo "Searching without filter"
dist/bin/search_memory_index --data_type float --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/index_rand_ins_del.after-concurrent-delete-del2500-7500 --result_path res_stream --query_file data/rand_float_10D_1K_norm1.0.bin --gt_file data/gt100_rand_random10D_1K -K 10 -L 20 40 60 80 100 -T 64
- name: upload data and bin
uses: actions/upload-artifact@v4
with:
name: dynamic-labels-${{matrix.os}}
path: |
./dist/**
./data/**

75
packages/leann-backend-diskann/third_party/DiskANN/.github/workflows/dynamic.yml vendored Normal file
View File

@@ -0,0 +1,75 @@
name: Dynamic
on: [workflow_call]
jobs:
acceptance-tests-dynamic:
name: Dynamic
strategy:
fail-fast: false
matrix:
os: [ubuntu-latest, windows-2019, windows-latest]
runs-on: ${{matrix.os}}
defaults:
run:
shell: bash
steps:
- name: Checkout repository
if: ${{ runner.os == 'Linux' }}
uses: actions/checkout@v3
with:
fetch-depth: 1
- name: Checkout repository
if: ${{ runner.os == 'Windows' }}
uses: actions/checkout@v3
with:
fetch-depth: 1
submodules: true
- name: DiskANN Build CLI Applications
uses: ./.github/actions/build
- name: Generate Data
uses: ./.github/actions/generate-random
- name: test a streaming index (float)
run: |
dist/bin/test_streaming_scenario --data_type float --dist_fn l2 --data_path data/rand_float_10D_10K_norm1.0.bin --index_path_prefix data/index_stream -R 64 -L 600 --alpha 1.2 --insert_threads 4 --consolidate_threads 4 --max_points_to_insert 10000 --active_window 4000 --consolidate_interval 2000 --start_point_norm 3.2
dist/bin/compute_groundtruth --data_type float --dist_fn l2 --base_file data/index_stream.after-streaming-act4000-cons2000-max10000.data --query_file data/rand_float_10D_1K_norm1.0.bin --K 100 --gt_file data/gt100_base-act4000-cons2000-max10000 --tags_file data/index_stream.after-streaming-act4000-cons2000-max10000.tags
dist/bin/search_memory_index --data_type float --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/index_stream.after-streaming-act4000-cons2000-max10000 --result_path data/res_stream --query_file data/rand_float_10D_1K_norm1.0.bin --gt_file data/gt100_base-act4000-cons2000-max10000 -K 10 -L 20 40 60 80 100 -T 64 --dynamic true --tags 1
- name: test a streaming index (int8)
if: success() || failure()
run: |
dist/bin/test_streaming_scenario --data_type int8 --dist_fn l2 --data_path data/rand_int8_10D_10K_norm50.0.bin --index_path_prefix data/index_stream -R 64 -L 600 --alpha 1.2 --insert_threads 4 --consolidate_threads 4 --max_points_to_insert 10000 --active_window 4000 --consolidate_interval 2000 --start_point_norm 200
dist/bin/compute_groundtruth --data_type int8 --dist_fn l2 --base_file data/index_stream.after-streaming-act4000-cons2000-max10000.data --query_file data/rand_int8_10D_1K_norm50.0.bin --K 100 --gt_file data/gt100_base-act4000-cons2000-max10000 --tags_file data/index_stream.after-streaming-act4000-cons2000-max10000.tags
dist/bin/search_memory_index --data_type int8 --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/index_stream.after-streaming-act4000-cons2000-max10000 --result_path res_stream --query_file data/rand_int8_10D_1K_norm50.0.bin --gt_file data/gt100_base-act4000-cons2000-max10000 -K 10 -L 20 40 60 80 100 -T 64 --dynamic true --tags 1
- name: test a streaming index
if: success() || failure()
run: |
dist/bin/test_streaming_scenario --data_type uint8 --dist_fn l2 --data_path data/rand_uint8_10D_10K_norm50.0.bin --index_path_prefix data/index_stream -R 64 -L 600 --alpha 1.2 --insert_threads 4 --consolidate_threads 4 --max_points_to_insert 10000 --active_window 4000 --consolidate_interval 2000 --start_point_norm 200
dist/bin/compute_groundtruth --data_type uint8 --dist_fn l2 --base_file data/index_stream.after-streaming-act4000-cons2000-max10000.data --query_file data/rand_uint8_10D_1K_norm50.0.bin --K 100 --gt_file data/gt100_base-act4000-cons2000-max10000 --tags_file data/index_stream.after-streaming-act4000-cons2000-max10000.tags
dist/bin/search_memory_index --data_type uint8 --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/index_stream.after-streaming-act4000-cons2000-max10000 --result_path data/res_stream --query_file data/rand_uint8_10D_1K_norm50.0.bin --gt_file data/gt100_base-act4000-cons2000-max10000 -K 10 -L 20 40 60 80 100 -T 64 --dynamic true --tags 1
- name: build and search an incremental index (float)
if: success() || failure()
run: |
dist/bin/test_insert_deletes_consolidate --data_type float --dist_fn l2 --data_path data/rand_float_10D_10K_norm1.0.bin --index_path_prefix data/index_ins_del -R 64 -L 300 --alpha 1.2 -T 8 --points_to_skip 0 --max_points_to_insert 7500 --beginning_index_size 0 --points_per_checkpoint 1000 --checkpoints_per_snapshot 0 --points_to_delete_from_beginning 2500 --start_deletes_after 5000 --do_concurrent true --start_point_norm 3.2;
dist/bin/compute_groundtruth --data_type float --dist_fn l2 --base_file data/index_ins_del.after-concurrent-delete-del2500-7500.data --query_file data/rand_float_10D_1K_norm1.0.bin --K 100 --gt_file data/gt100_random10D_1K-conc-2500-7500 --tags_file data/index_ins_del.after-concurrent-delete-del2500-7500.tags
dist/bin/search_memory_index --data_type float --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/index_ins_del.after-concurrent-delete-del2500-7500 --result_path data/res_ins_del --query_file data/rand_float_10D_1K_norm1.0.bin --gt_file data/gt100_random10D_1K-conc-2500-7500 -K 10 -L 20 40 60 80 100 -T 8 --dynamic true --tags 1
- name: build and search an incremental index (int8)
if: success() || failure()
run: |
dist/bin/test_insert_deletes_consolidate --data_type int8 --dist_fn l2 --data_path data/rand_int8_10D_10K_norm50.0.bin --index_path_prefix data/index_ins_del -R 64 -L 300 --alpha 1.2 -T 8 --points_to_skip 0 --max_points_to_insert 7500 --beginning_index_size 0 --points_per_checkpoint 1000 --checkpoints_per_snapshot 0 --points_to_delete_from_beginning 2500 --start_deletes_after 5000 --do_concurrent true --start_point_norm 200
dist/bin/compute_groundtruth --data_type int8 --dist_fn l2 --base_file data/index_ins_del.after-concurrent-delete-del2500-7500.data --query_file data/rand_int8_10D_1K_norm50.0.bin --K 100 --gt_file data/gt100_random10D_1K-conc-2500-7500 --tags_file data/index_ins_del.after-concurrent-delete-del2500-7500.tags
dist/bin/search_memory_index --data_type int8 --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/index_ins_del.after-concurrent-delete-del2500-7500 --result_path data/res_ins_del --query_file data/rand_int8_10D_1K_norm50.0.bin --gt_file data/gt100_random10D_1K-conc-2500-7500 -K 10 -L 20 40 60 80 100 -T 8 --dynamic true --tags 1
- name: build and search an incremental index (uint8)
if: success() || failure()
run: |
dist/bin/test_insert_deletes_consolidate --data_type uint8 --dist_fn l2 --data_path data/rand_uint8_10D_10K_norm50.0.bin --index_path_prefix data/index_ins_del -R 64 -L 300 --alpha 1.2 -T 8 --points_to_skip 0 --max_points_to_insert 7500 --beginning_index_size 0 --points_per_checkpoint 1000 --checkpoints_per_snapshot 0 --points_to_delete_from_beginning 2500 --start_deletes_after 5000 --do_concurrent true --start_point_norm 200
dist/bin/compute_groundtruth --data_type uint8 --dist_fn l2 --base_file data/index_ins_del.after-concurrent-delete-del2500-7500.data --query_file data/rand_uint8_10D_1K_norm50.0.bin --K 100 --gt_file data/gt100_random10D_10K-conc-2500-7500 --tags_file data/index_ins_del.after-concurrent-delete-del2500-7500.tags
dist/bin/search_memory_index --data_type uint8 --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/index_ins_del.after-concurrent-delete-del2500-7500 --result_path data/res_ins_del --query_file data/rand_uint8_10D_1K_norm50.0.bin --gt_file data/gt100_random10D_10K-conc-2500-7500 -K 10 -L 20 40 60 80 100 -T 8 --dynamic true --tags 1
- name: upload data and bin
uses: actions/upload-artifact@v4
with:
name: dynamic-${{matrix.os}}
path: |
./dist/**
./data/**

81
packages/leann-backend-diskann/third_party/DiskANN/.github/workflows/in-mem-no-pq.yml vendored Normal file
View File

@@ -0,0 +1,81 @@
name: In-Memory Without PQ
on: [workflow_call]
jobs:
acceptance-tests-mem-no-pq:
name: In-Mem, Without PQ
strategy:
fail-fast: false
matrix:
os: [ubuntu-latest, windows-2019, windows-latest]
runs-on: ${{matrix.os}}
defaults:
run:
shell: bash
steps:
- name: Checkout repository
if: ${{ runner.os == 'Linux' }}
uses: actions/checkout@v3
with:
fetch-depth: 1
- name: Checkout repository
if: ${{ runner.os == 'Windows' }}
uses: actions/checkout@v3
with:
fetch-depth: 1
submodules: true
- name: DiskANN Build CLI Applications
uses: ./.github/actions/build
- name: Generate Data
uses: ./.github/actions/generate-random
- name: build and search in-memory index with L2 metrics (float)
if: success() || failure()
run: |
dist/bin/build_memory_index --data_type float --dist_fn l2 --data_path data/rand_float_10D_10K_norm1.0.bin --index_path_prefix data/index_l2_rand_float_10D_10K_norm1.0
dist/bin/search_memory_index --data_type float --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/index_l2_rand_float_10D_10K_norm1.0 --query_file data/rand_float_10D_1K_norm1.0.bin --recall_at 10 --result_path temp --gt_file data/l2_rand_float_10D_10K_norm1.0_10D_1K_norm1.0_gt100 -L 16 32
- name: build and search in-memory index with L2 metrics (int8)
if: success() || failure()
run: |
dist/bin/build_memory_index --data_type int8 --dist_fn l2 --data_path data/rand_int8_10D_10K_norm50.0.bin --index_path_prefix data/index_l2_rand_int8_10D_10K_norm50.0
dist/bin/search_memory_index --data_type int8 --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/index_l2_rand_int8_10D_10K_norm50.0 --query_file data/rand_int8_10D_1K_norm50.0.bin --recall_at 10 --result_path temp --gt_file data/l2_rand_int8_10D_10K_norm50.0_10D_1K_norm50.0_gt100 -L 16 32
- name: build and search in-memory index with L2 metrics (uint8)
if: success() || failure()
run: |
dist/bin/build_memory_index --data_type uint8 --dist_fn l2 --data_path data/rand_uint8_10D_10K_norm50.0.bin --index_path_prefix data/index_l2_rand_uint8_10D_10K_norm50.0
dist/bin/search_memory_index --data_type uint8 --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/index_l2_rand_uint8_10D_10K_norm50.0 --query_file data/rand_uint8_10D_1K_norm50.0.bin --recall_at 10 --result_path temp --gt_file data/l2_rand_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100 -L 16 32
- name: Searching with fast_l2 distance function (float)
if: runner.os != 'Windows' && (success() || failure())
run: |
dist/bin/search_memory_index --data_type float --dist_fn fast_l2 --fail_if_recall_below 70 --index_path_prefix data/index_l2_rand_float_10D_10K_norm1.0 --query_file data/rand_float_10D_1K_norm1.0.bin --recall_at 10 --result_path temp --gt_file data/l2_rand_float_10D_10K_norm1.0_10D_1K_norm1.0_gt100 -L 16 32
- name: build and search in-memory index with MIPS metric (float)
if: success() || failure()
run: |
dist/bin/build_memory_index --data_type float --dist_fn mips --data_path data/rand_float_10D_10K_norm1.0.bin --index_path_prefix data/index_mips_rand_float_10D_10K_norm1.0
dist/bin/search_memory_index --data_type float --dist_fn mips --fail_if_recall_below 70 --index_path_prefix data/index_l2_rand_float_10D_10K_norm1.0 --query_file data/rand_float_10D_1K_norm1.0.bin --recall_at 10 --result_path temp --gt_file data/mips_rand_float_10D_10K_norm1.0_10D_1K_norm1.0_gt100 -L 16 32
- name: build and search in-memory index with cosine metric (float)
if: success() || failure()
run: |
dist/bin/build_memory_index --data_type float --dist_fn cosine --data_path data/rand_float_10D_10K_norm1.0.bin --index_path_prefix data/index_cosine_rand_float_10D_10K_norm1.0
dist/bin/search_memory_index --data_type float --dist_fn cosine --fail_if_recall_below 70 --index_path_prefix data/index_l2_rand_float_10D_10K_norm1.0 --query_file data/rand_float_10D_1K_norm1.0.bin --recall_at 10 --result_path temp --gt_file data/cosine_rand_float_10D_10K_norm1.0_10D_1K_norm1.0_gt100 -L 16 32
- name: build and search in-memory index with cosine metric (int8)
if: success() || failure()
run: |
dist/bin/build_memory_index --data_type int8 --dist_fn cosine --data_path data/rand_int8_10D_10K_norm50.0.bin --index_path_prefix data/index_cosine_rand_int8_10D_10K_norm50.0
dist/bin/search_memory_index --data_type int8 --dist_fn cosine --fail_if_recall_below 70 --index_path_prefix data/index_l2_rand_int8_10D_10K_norm50.0 --query_file data/rand_int8_10D_1K_norm50.0.bin --recall_at 10 --result_path temp --gt_file data/cosine_rand_int8_10D_10K_norm50.0_10D_1K_norm50.0_gt100 -L 16 32
- name: build and search in-memory index with cosine metric
if: success() || failure()
run: |
dist/bin/build_memory_index --data_type uint8 --dist_fn cosine --data_path data/rand_uint8_10D_10K_norm50.0.bin --index_path_prefix data/index_cosine_rand_uint8_10D_10K_norm50.0
dist/bin/search_memory_index --data_type uint8 --dist_fn cosine --fail_if_recall_below 70 --index_path_prefix data/index_l2_rand_uint8_10D_10K_norm50.0 --query_file data/rand_uint8_10D_1K_norm50.0.bin --recall_at 10 --result_path temp --gt_file data/cosine_rand_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100 -L 16 32
- name: upload data and bin
uses: actions/upload-artifact@v4
with:
name: in-memory-no-pq-${{matrix.os}}
path: |
./dist/**
./data/**

56
packages/leann-backend-diskann/third_party/DiskANN/.github/workflows/in-mem-pq.yml vendored Normal file
View File

@@ -0,0 +1,56 @@
name: In-Memory With PQ
on: [workflow_call]
jobs:
acceptance-tests-mem-pq:
name: In-Mem, PQ
strategy:
fail-fast: false
matrix:
os: [ubuntu-latest, windows-2019, windows-latest]
runs-on: ${{matrix.os}}
defaults:
run:
shell: bash
steps:
- name: Checkout repository
if: ${{ runner.os == 'Linux' }}
uses: actions/checkout@v3
with:
fetch-depth: 1
- name: Checkout repository
if: ${{ runner.os == 'Windows' }}
uses: actions/checkout@v3
with:
fetch-depth: 1
submodules: true
- name: DiskANN Build CLI Applications
uses: ./.github/actions/build
- name: Generate Data
uses: ./.github/actions/generate-random
- name: build and search in-memory index with L2 metric with PQ based distance comparisons (float)
if: success() || failure()
run: |
dist/bin/build_memory_index --data_type float --dist_fn l2 --data_path data/rand_float_10D_10K_norm1.0.bin --index_path_prefix data/index_l2_rand_float_10D_10K_norm1.0_buildpq5 --build_PQ_bytes 5
dist/bin/search_memory_index --data_type float --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/index_l2_rand_float_10D_10K_norm1.0_buildpq5 --query_file data/rand_float_10D_1K_norm1.0.bin --recall_at 10 --result_path temp --gt_file data/l2_rand_float_10D_10K_norm1.0_10D_1K_norm1.0_gt100 -L 16 32
- name: build and search in-memory index with L2 metrics with PQ base distance comparisons (int8)
if: success() || failure()
run: |
dist/bin/build_memory_index --data_type int8 --dist_fn l2 --data_path data/rand_int8_10D_10K_norm50.0.bin --index_path_prefix data/index_l2_rand_int8_10D_10K_norm50.0_buildpq5 --build_PQ_bytes 5
dist/bin/search_memory_index --data_type int8 --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/index_l2_rand_int8_10D_10K_norm50.0_buildpq5 --query_file data/rand_int8_10D_1K_norm50.0.bin --recall_at 10 --result_path temp --gt_file data/l2_rand_int8_10D_10K_norm50.0_10D_1K_norm50.0_gt100 -L 16 32
- name: build and search in-memory index with L2 metrics with PQ base distance comparisons (uint8)
if: success() || failure()
run: |
dist/bin/build_memory_index --data_type uint8 --dist_fn l2 --data_path data/rand_uint8_10D_10K_norm50.0.bin --index_path_prefix data/index_l2_rand_uint8_10D_10K_norm50.0_buildpq5 --build_PQ_bytes 5
dist/bin/search_memory_index --data_type uint8 --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/index_l2_rand_uint8_10D_10K_norm50.0_buildpq5 --query_file data/rand_uint8_10D_1K_norm50.0.bin --recall_at 10 --result_path temp --gt_file data/l2_rand_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100 -L 16 32
- name: upload data and bin
uses: actions/upload-artifact@v4
with:
name: in-memory-pq-${{matrix.os}}
path: |
./dist/**
./data/**

120
packages/leann-backend-diskann/third_party/DiskANN/.github/workflows/labels.yml vendored Normal file
View File

@@ -0,0 +1,120 @@
name: Labels
on: [workflow_call]
jobs:
acceptance-tests-labels:
name: Labels
strategy:
fail-fast: false
matrix:
os: [ubuntu-latest, windows-2019, windows-latest]
runs-on: ${{matrix.os}}
defaults:
run:
shell: bash
steps:
- name: Checkout repository
if: ${{ runner.os == 'Linux' }}
uses: actions/checkout@v3
with:
fetch-depth: 1
- name: Checkout repository
if: ${{ runner.os == 'Windows' }}
uses: actions/checkout@v3
with:
fetch-depth: 1
submodules: true
- name: DiskANN Build CLI Applications
uses: ./.github/actions/build
- name: Generate Data
uses: ./.github/actions/generate-random
- name: Generate Labels
run: |
echo "Generating synthetic labels and computing ground truth for filtered search with universal label"
dist/bin/generate_synthetic_labels --num_labels 50 --num_points 10000 --output_file data/rand_labels_50_10K.txt --distribution_type random
dist/bin/compute_groundtruth_for_filters --data_type uint8 --dist_fn l2 --universal_label 0 --filter_label 10 --base_file data/rand_uint8_10D_10K_norm50.0.bin --query_file data/rand_uint8_10D_1K_norm50.0.bin --label_file data/rand_labels_50_10K.txt --gt_file data/l2_rand_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100_wlabel --K 100
dist/bin/compute_groundtruth_for_filters --data_type uint8 --dist_fn mips --universal_label 0 --filter_label 10 --base_file data/rand_uint8_10D_10K_norm50.0.bin --query_file data/rand_uint8_10D_1K_norm50.0.bin --label_file data/rand_labels_50_10K.txt --gt_file data/mips_rand_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100_wlabel --K 100
dist/bin/compute_groundtruth_for_filters --data_type uint8 --dist_fn cosine --universal_label 0 --filter_label 10 --base_file data/rand_uint8_10D_10K_norm50.0.bin --query_file data/rand_uint8_10D_1K_norm50.0.bin --label_file data/rand_labels_50_10K.txt --gt_file data/cosine_rand_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100_wlabel --K 100
echo "Generating synthetic labels with a zipf distribution and computing ground truth for filtered search with universal label"
dist/bin/generate_synthetic_labels --num_labels 50 --num_points 10000 --output_file data/zipf_labels_50_10K.txt --distribution_type zipf
dist/bin/compute_groundtruth_for_filters --data_type uint8 --dist_fn l2 --universal_label 0 --filter_label 5 --base_file data/rand_uint8_10D_10K_norm50.0.bin --query_file data/rand_uint8_10D_1K_norm50.0.bin --label_file data/zipf_labels_50_10K.txt --gt_file data/l2_zipf_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100_wlabel --K 100
dist/bin/compute_groundtruth_for_filters --data_type uint8 --dist_fn mips --universal_label 0 --filter_label 5 --base_file data/rand_uint8_10D_10K_norm50.0.bin --query_file data/rand_uint8_10D_1K_norm50.0.bin --label_file data/zipf_labels_50_10K.txt --gt_file data/mips_zipf_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100_wlabel --K 100
dist/bin/compute_groundtruth_for_filters --data_type uint8 --dist_fn cosine --universal_label 0 --filter_label 5 --base_file data/rand_uint8_10D_10K_norm50.0.bin --query_file data/rand_uint8_10D_1K_norm50.0.bin --label_file data/zipf_labels_50_10K.txt --gt_file data/cosine_zipf_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100_wlabel --K 100
echo "Generating synthetic labels and computing ground truth for filtered search without a universal label"
dist/bin/compute_groundtruth_for_filters --data_type uint8 --dist_fn l2 --filter_label 5 --base_file data/rand_uint8_10D_10K_norm50.0.bin --query_file data/rand_uint8_10D_1K_norm50.0.bin --label_file data/zipf_labels_50_10K.txt --gt_file data/l2_zipf_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100_wlabel_nouniversal --K 100
dist/bin/generate_synthetic_labels --num_labels 10 --num_points 1000 --output_file data/query_labels_1K.txt --distribution_type one_per_point
dist/bin/compute_groundtruth_for_filters --data_type uint8 --dist_fn l2 --universal_label 0 --filter_label_file data/query_labels_1K.txt --base_file data/rand_uint8_10D_10K_norm50.0.bin --query_file data/rand_uint8_10D_1K_norm50.0.bin --label_file data/zipf_labels_50_10K.txt --gt_file data/combined_l2_zipf_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100_wlabel --K 100
- name: build and search in-memory index with labels using L2 and Cosine metrics (random distributed labels)
if: success() || failure()
run: |
dist/bin/build_memory_index --data_type uint8 --dist_fn l2 --FilteredLbuild 90 --universal_label 0 --data_path data/rand_uint8_10D_10K_norm50.0.bin --label_file data/rand_labels_50_10K.txt --index_path_prefix data/index_l2_rand_uint8_10D_10K_norm50_wlabel
dist/bin/build_memory_index --data_type uint8 --dist_fn cosine --FilteredLbuild 90 --universal_label 0 --data_path data/rand_uint8_10D_10K_norm50.0.bin --label_file data/rand_labels_50_10K.txt --index_path_prefix data/index_cosine_rand_uint8_10D_10K_norm50_wlabel
dist/bin/search_memory_index --data_type uint8 --dist_fn l2 --filter_label 10 --fail_if_recall_below 70 --index_path_prefix data/index_l2_rand_uint8_10D_10K_norm50_wlabel --query_file data/rand_uint8_10D_1K_norm50.0.bin --recall_at 10 --result_path temp --gt_file data/l2_rand_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100_wlabel -L 16 32
dist/bin/search_memory_index --data_type uint8 --dist_fn cosine --filter_label 10 --fail_if_recall_below 70 --index_path_prefix data/index_cosine_rand_uint8_10D_10K_norm50_wlabel --query_file data/rand_uint8_10D_1K_norm50.0.bin --recall_at 10 --result_path temp --gt_file data/cosine_rand_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100_wlabel -L 16 32
echo "Searching without filters"
dist/bin/search_memory_index --data_type uint8 --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/index_l2_rand_uint8_10D_10K_norm50_wlabel --query_file data/rand_uint8_10D_1K_norm50.0.bin --recall_at 10 --result_path temp --gt_file data/l2_rand_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100 -L 32 64
dist/bin/search_memory_index --data_type uint8 --dist_fn cosine --fail_if_recall_below 70 --index_path_prefix data/index_cosine_rand_uint8_10D_10K_norm50_wlabel --query_file data/rand_uint8_10D_1K_norm50.0.bin --recall_at 10 --result_path temp --gt_file data/cosine_rand_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100 -L 32 64
- name: build and search disk index with labels using L2 and Cosine metrics (random distributed labels)
if: success() || failure()
run: |
dist/bin/build_disk_index --data_type uint8 --dist_fn l2 --universal_label 0 --FilteredLbuild 90 --data_path data/rand_uint8_10D_10K_norm50.0.bin --label_file data/rand_labels_50_10K.txt --index_path_prefix data/disk_index_l2_rand_uint8_10D_10K_norm50_wlabel -R 32 -L 5 -B 0.00003 -M 1
dist/bin/search_disk_index --data_type uint8 --dist_fn l2 --filter_label 10 --fail_if_recall_below 50 --index_path_prefix data/disk_index_l2_rand_uint8_10D_10K_norm50_wlabel --result_path temp --query_file data/rand_uint8_10D_1K_norm50.0.bin --gt_file data/l2_rand_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100_wlabel --recall_at 5 -L 5 12 -W 2 --num_nodes_to_cache 10 -T 16
- name: build and search in-memory index with labels using L2 and Cosine metrics (zipf distributed labels)
if: success() || failure()
run: |
dist/bin/build_memory_index --data_type uint8 --dist_fn l2 --FilteredLbuild 90 --universal_label 0 --data_path data/rand_uint8_10D_10K_norm50.0.bin --label_file data/zipf_labels_50_10K.txt --index_path_prefix data/index_l2_zipf_uint8_10D_10K_norm50_wlabel
dist/bin/build_memory_index --data_type uint8 --dist_fn cosine --FilteredLbuild 90 --universal_label 0 --data_path data/rand_uint8_10D_10K_norm50.0.bin --label_file data/zipf_labels_50_10K.txt --index_path_prefix data/index_cosine_zipf_uint8_10D_10K_norm50_wlabel
dist/bin/search_memory_index --data_type uint8 --dist_fn l2 --filter_label 5 --fail_if_recall_below 70 --index_path_prefix data/index_l2_zipf_uint8_10D_10K_norm50_wlabel --query_file data/rand_uint8_10D_1K_norm50.0.bin --recall_at 10 --result_path temp --gt_file data/l2_zipf_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100_wlabel -L 16 32
dist/bin/search_memory_index --data_type uint8 --dist_fn cosine --filter_label 5 --fail_if_recall_below 70 --index_path_prefix data/index_cosine_zipf_uint8_10D_10K_norm50_wlabel --query_file data/rand_uint8_10D_1K_norm50.0.bin --recall_at 10 --result_path temp --gt_file data/cosine_zipf_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100_wlabel -L 16 32
echo "Searching without filters"
dist/bin/compute_groundtruth --data_type uint8 --dist_fn l2 --base_file data/rand_uint8_10D_10K_norm50.0.bin --query_file data/rand_uint8_10D_1K_norm50.0.bin --gt_file data/l2_zipf_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100 --K 100
dist/bin/compute_groundtruth --data_type uint8 --dist_fn cosine --base_file data/rand_uint8_10D_10K_norm50.0.bin --query_file data/rand_uint8_10D_1K_norm50.0.bin --gt_file data/cosine_zipf_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100 --K 100
dist/bin/search_memory_index --data_type uint8 --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/index_l2_zipf_uint8_10D_10K_norm50_wlabel --query_file data/rand_uint8_10D_1K_norm50.0.bin --recall_at 10 --result_path temp --gt_file data/l2_zipf_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100 -L 32 64
dist/bin/search_memory_index --data_type uint8 --dist_fn cosine --fail_if_recall_below 70 --index_path_prefix data/index_cosine_zipf_uint8_10D_10K_norm50_wlabel --query_file data/rand_uint8_10D_1K_norm50.0.bin --recall_at 10 --result_path temp --gt_file data/cosine_zipf_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100 -L 32 64
- name: build and search disk index with labels using L2 and Cosine metrics (zipf distributed labels)
if: success() || failure()
run: |
dist/bin/build_disk_index --data_type uint8 --dist_fn l2 --universal_label 0 --FilteredLbuild 90 --data_path data/rand_uint8_10D_10K_norm50.0.bin --label_file data/zipf_labels_50_10K.txt --index_path_prefix data/disk_index_l2_zipf_uint8_10D_10K_norm50_wlabel -R 32 -L 5 -B 0.00003 -M 1
dist/bin/search_disk_index --data_type uint8 --dist_fn l2 --filter_label 5 --fail_if_recall_below 50 --index_path_prefix data/disk_index_l2_zipf_uint8_10D_10K_norm50_wlabel --result_path temp --query_file data/rand_uint8_10D_1K_norm50.0.bin --gt_file data/l2_zipf_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100_wlabel --recall_at 5 -L 5 12 -W 2 --num_nodes_to_cache 10 -T 16
- name : build and search in-memory and disk index (without universal label, zipf distributed)
if: success() || failure()
run: |
dist/bin/build_memory_index --data_type uint8 --dist_fn l2 --FilteredLbuild 90 --data_path data/rand_uint8_10D_10K_norm50.0.bin --label_file data/zipf_labels_50_10K.txt --index_path_prefix data/index_l2_zipf_uint8_10D_10K_norm50_wlabel_nouniversal
dist/bin/build_disk_index --data_type uint8 --dist_fn l2 --FilteredLbuild 90 --data_path data/rand_uint8_10D_10K_norm50.0.bin --label_file data/zipf_labels_50_10K.txt --index_path_prefix data/disk_index_l2_zipf_uint8_10D_10K_norm50_wlabel_nouniversal -R 32 -L 5 -B 0.00003 -M 1
dist/bin/search_memory_index --data_type uint8 --dist_fn l2 --filter_label 5 --fail_if_recall_below 70 --index_path_prefix data/index_l2_zipf_uint8_10D_10K_norm50_wlabel_nouniversal --query_file data/rand_uint8_10D_1K_norm50.0.bin --recall_at 10 --result_path temp --gt_file data/l2_zipf_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100_wlabel_nouniversal -L 16 32
dist/bin/search_disk_index --data_type uint8 --dist_fn l2 --filter_label 5 --index_path_prefix data/disk_index_l2_zipf_uint8_10D_10K_norm50_wlabel_nouniversal --result_path temp --query_file data/rand_uint8_10D_1K_norm50.0.bin --gt_file data/l2_zipf_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100_wlabel_nouniversal --recall_at 5 -L 5 12 -W 2 --num_nodes_to_cache 10 -T 16
- name: Generate combined GT for each query with a separate label and search
if: success() || failure()
run: |
dist/bin/build_memory_index --data_type uint8 --dist_fn l2 --FilteredLbuild 90 --universal_label 0 --data_path data/rand_uint8_10D_10K_norm50.0.bin --label_file data/zipf_labels_50_10K.txt --index_path_prefix data/index_l2_zipf_uint8_10D_10K_norm50_wlabel
dist/bin/search_memory_index --data_type uint8 --dist_fn l2 --query_filters_file data/query_labels_1K.txt --fail_if_recall_below 70 --index_path_prefix data/index_l2_zipf_uint8_10D_10K_norm50_wlabel --query_file data/rand_uint8_10D_1K_norm50.0.bin --recall_at 10 --result_path temp --gt_file data/combined_l2_zipf_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100_wlabel -L 16 32
- name: build and search in-memory index with pq_dist of 5 with 10 dimensions
if: success() || failure()
run: |
dist/bin/build_memory_index --data_type uint8 --dist_fn l2 --FilteredLbuild 90 --universal_label 0 --data_path data/rand_uint8_10D_10K_norm50.0.bin --label_file data/rand_labels_50_10K.txt --index_path_prefix data/index_l2_rand_uint8_10D_10K_norm50_wlabel --build_PQ_bytes 5
dist/bin/search_memory_index --data_type uint8 --dist_fn l2 --filter_label 10 --fail_if_recall_below 70 --index_path_prefix data/index_l2_rand_uint8_10D_10K_norm50_wlabel --query_file data/rand_uint8_10D_1K_norm50.0.bin --recall_at 10 --result_path temp --gt_file data/l2_rand_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100_wlabel -L 16 32
- name: Build and search stitched vamana with random and zipf distributed labels
if: success() || failure()
run: |
dist/bin/build_stitched_index --num_threads 48 --data_type uint8 --data_path data/rand_uint8_10D_10K_norm50.0.bin --label_file data/rand_labels_50_10K.txt -R 32 -L 100 --alpha 1.2 --stitched_R 64 --index_path_prefix data/stit_rand_32_100_64_new --universal_label 0
dist/bin/build_stitched_index --num_threads 48 --data_type uint8 --data_path data/rand_uint8_10D_10K_norm50.0.bin --label_file data/zipf_labels_50_10K.txt -R 32 -L 100 --alpha 1.2 --stitched_R 64 --index_path_prefix data/stit_zipf_32_100_64_new --universal_label 0
dist/bin/search_memory_index --num_threads 48 --data_type uint8 --dist_fn l2 --filter_label 10 --index_path_prefix data/stit_rand_32_100_64_new --query_file data/rand_uint8_10D_1K_norm50.0.bin --result_path data/rand_stit_96_10_90_new --gt_file data/l2_rand_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100_wlabel -K 10 -L 16 32 150
dist/bin/search_memory_index --num_threads 48 --data_type uint8 --dist_fn l2 --filter_label 5 --index_path_prefix data/stit_zipf_32_100_64_new --query_file data/rand_uint8_10D_1K_norm50.0.bin --result_path data/zipf_stit_96_10_90_new --gt_file data/l2_zipf_uint8_10D_10K_norm50.0_10D_1K_norm50.0_gt100_wlabel -K 10 -L 16 32 150
- name: upload data and bin
if: success() || failure()
uses: actions/upload-artifact@v4
with:
name: labels-${{matrix.os}}
path: |
./dist/**
./data/**

60
packages/leann-backend-diskann/third_party/DiskANN/.github/workflows/multi-sector-disk-pq.yml vendored Normal file
View File

@@ -0,0 +1,60 @@
name: Disk With PQ
on: [workflow_call]
jobs:
acceptance-tests-disk-pq:
name: Disk, PQ
strategy:
fail-fast: false
matrix:
os: [ubuntu-latest, windows-2019, windows-latest]
runs-on: ${{matrix.os}}
defaults:
run:
shell: bash
steps:
- name: Checkout repository
if: ${{ runner.os == 'Linux' }}
uses: actions/checkout@v3
with:
fetch-depth: 1
- name: Checkout repository
if: ${{ runner.os == 'Windows' }}
uses: actions/checkout@v3
with:
fetch-depth: 1
submodules: true
- name: DiskANN Build CLI Applications
uses: ./.github/actions/build
- name: Generate Data
uses: ./.github/actions/generate-high-dim-random
- name: build and search disk index (1020D, one shot graph build, L2, no diskPQ) (float)
if: success() || failure()
run: |
dist/bin/build_disk_index --data_type float --dist_fn l2 --data_path data/rand_float_1020D_5K_norm1.0.bin --index_path_prefix data/disk_index_l2_rand_float_1020D_5K_norm1.0_diskfull_oneshot -R 32 -L 500 -B 0.003 -M 1
dist/bin/search_disk_index --data_type float --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/disk_index_l2_rand_float_1020D_5K_norm1.0_diskfull_oneshot --result_path /tmp/res --query_file data/rand_float_1020D_1K_norm1.0.bin --gt_file data/l2_rand_float_1020D_5K_norm1.0_1020D_1K_norm1.0_gt100 --recall_at 5 -L 250 -W 2 --num_nodes_to_cache 100 -T 16
#- name: build and search disk index (1024D, one shot graph build, L2, no diskPQ) (float)
# if: success() || failure()
# run: |
# dist/bin/build_disk_index --data_type float --dist_fn l2 --data_path data/rand_float_1024D_5K_norm1.0.bin --index_path_prefix data/disk_index_l2_rand_float_1024D_5K_norm1.0_diskfull_oneshot -R 32 -L 500 -B 0.003 -M 1
# dist/bin/search_disk_index --data_type float --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/disk_index_l2_rand_float_1024D_5K_norm1.0_diskfull_oneshot --result_path /tmp/res --query_file data/rand_float_1024D_1K_norm1.0.bin --gt_file data/l2_rand_float_1024D_5K_norm1.0_1024D_1K_norm1.0_gt100 --recall_at 5 -L 250 -W 2 --num_nodes_to_cache 100 -T 16
- name: build and search disk index (1536D, one shot graph build, L2, no diskPQ) (float)
if: success() || failure()
run: |
dist/bin/build_disk_index --data_type float --dist_fn l2 --data_path data/rand_float_1536D_5K_norm1.0.bin --index_path_prefix data/disk_index_l2_rand_float_1536D_5K_norm1.0_diskfull_oneshot -R 32 -L 500 -B 0.003 -M 1
dist/bin/search_disk_index --data_type float --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/disk_index_l2_rand_float_1536D_5K_norm1.0_diskfull_oneshot --result_path /tmp/res --query_file data/rand_float_1536D_1K_norm1.0.bin --gt_file data/l2_rand_float_1536D_5K_norm1.0_1536D_1K_norm1.0_gt100 --recall_at 5 -L 250 -W 2 --num_nodes_to_cache 100 -T 16
- name: build and search disk index (4096D, one shot graph build, L2, no diskPQ) (int8)
if: success() || failure()
run: |
dist/bin/build_disk_index --data_type int8 --dist_fn l2 --data_path data/rand_int8_4096D_5K_norm1.0.bin --index_path_prefix data/disk_index_l2_rand_int8_4096D_5K_norm1.0_diskfull_oneshot -R 32 -L 500 -B 0.003 -M 1
dist/bin/search_disk_index --data_type int8 --dist_fn l2 --fail_if_recall_below 70 --index_path_prefix data/disk_index_l2_rand_int8_4096D_5K_norm1.0_diskfull_oneshot --result_path /tmp/res --query_file data/rand_int8_4096D_1K_norm1.0.bin --gt_file data/l2_rand_int8_4096D_5K_norm1.0_4096D_1K_norm1.0_gt100 --recall_at 5 -L 250 -W 2 --num_nodes_to_cache 100 -T 16
- name: upload data and bin
uses: actions/upload-artifact@v4
with:
name: multi-sector-disk-pq-${{matrix.os}}
path: |
./dist/**
./data/**

26
packages/leann-backend-diskann/third_party/DiskANN/.github/workflows/perf.yml vendored Normal file
View File

@@ -0,0 +1,26 @@
name: DiskANN Nightly Performance Metrics
on:
schedule:
- cron: "41 14 * * *" # 14:41 UTC, 7:41 PDT, 8:41 PST, 08:11 IST
jobs:
perf-test:
name: Run Perf Test from main
runs-on: ubuntu-latest
steps:
- name: Checkout repository
uses: actions/checkout@v3
with:
fetch-depth: 1
- name: Build Perf Container
run: |
docker build --build-arg GIT_COMMIT_ISH="$GITHUB_SHA" -t perf -f scripts/perf/Dockerfile scripts
- name: Performance Tests
run: |
mkdir metrics
docker run -v ./metrics:/app/logs perf &> ./metrics/combined_stdouterr.log
- name: Upload Metrics Logs
uses: actions/upload-artifact@v4
with:
name: metrics-${{matrix.os}}
path: |
./metrics/**

35
packages/leann-backend-diskann/third_party/DiskANN/.github/workflows/pr-test.yml vendored Normal file
View File

@@ -0,0 +1,35 @@
name: DiskANN Pull Request Build and Test
on: [pull_request]
jobs:
common:
strategy:
fail-fast: true
name: DiskANN Common Build Checks
uses: ./.github/workflows/common.yml
unit-tests:
name: Unit tests
uses: ./.github/workflows/unit-tests.yml
in-mem-pq:
name: In-Memory with PQ
uses: ./.github/workflows/in-mem-pq.yml
in-mem-no-pq:
name: In-Memory without PQ
uses: ./.github/workflows/in-mem-no-pq.yml
disk-pq:
name: Disk with PQ
uses: ./.github/workflows/disk-pq.yml
multi-sector-disk-pq:
name: Multi-sector Disk with PQ
uses: ./.github/workflows/multi-sector-disk-pq.yml
labels:
name: Labels
uses: ./.github/workflows/labels.yml
dynamic:
name: Dynamic
uses: ./.github/workflows/dynamic.yml
dynamic-labels:
name: Dynamic Labels
uses: ./.github/workflows/dynamic-labels.yml
python:
name: Python
uses: ./.github/workflows/build-python.yml

50
packages/leann-backend-diskann/third_party/DiskANN/.github/workflows/push-test.yml vendored Normal file
View File

@@ -0,0 +1,50 @@
name: DiskANN Push Build
on: [push]
jobs:
common:
strategy:
fail-fast: true
name: DiskANN Common Build Checks
uses: ./.github/workflows/common.yml
build-documentation:
permissions:
contents: write
strategy:
fail-fast: true
name: DiskANN Build Documentation
uses: ./.github/workflows/build-python-pdoc.yml
build:
strategy:
fail-fast: false
matrix:
os: [ ubuntu-latest, windows-2019, windows-latest ]
name: Build for ${{matrix.os}}
runs-on: ${{matrix.os}}
defaults:
run:
shell: bash
steps:
- name: Checkout repository
if: ${{ runner.os == 'Linux' }}
uses: actions/checkout@v3
with:
fetch-depth: 1
- name: Checkout repository
if: ${{ runner.os == 'Windows' }}
uses: actions/checkout@v3
with:
fetch-depth: 1
submodules: true
- name: Build diskannpy dependency tree
run: |
pip install diskannpy pipdeptree
echo "dependencies" > dependencies_${{ matrix.os }}.txt
pipdeptree >> dependencies_${{ matrix.os }}.txt
- name: Archive diskannpy dependencies artifact
uses: actions/upload-artifact@v4
with:
name: dependencies_${{ matrix.os }}
path: |
dependencies_${{ matrix.os }}.txt
- name: DiskANN Build CLI Applications
uses: ./.github/actions/build

43
packages/leann-backend-diskann/third_party/DiskANN/.github/workflows/python-release.yml vendored Normal file
View File

@@ -0,0 +1,43 @@
name: Build and Release Python Wheels
on:
release:
types: [published]
jobs:
python-release-wheels:
name: Python
uses: ./.github/workflows/build-python.yml
build-documentation:
strategy:
fail-fast: true
name: DiskANN Build Documentation
uses: ./.github/workflows/build-python-pdoc.yml
release:
permissions:
contents: write
runs-on: ubuntu-latest
needs: python-release-wheels
steps:
- uses: actions/download-artifact@v3
with:
name: wheels
path: dist/
- name: Generate SHA256 files for each wheel
run: |
sha256sum dist/*.whl > checksums.txt
cat checksums.txt
- uses: actions/setup-python@v3
- name: Install twine
run: python -m pip install twine
- name: Publish with twine
env:
TWINE_USERNAME: ${{ secrets.PYPI_USERNAME }}
TWINE_PASSWORD: ${{ secrets.PYPI_PASSWORD }}
run: |
twine upload dist/*.whl
- name: Update release with SHA256 and Artifacts
uses: softprops/action-gh-release@v1
with:
token: ${{ secrets.GITHUB_TOKEN }}
files: |
dist/*.whl
checksums.txt

32
packages/leann-backend-diskann/third_party/DiskANN/.github/workflows/unit-tests.yml vendored Normal file
View File

@@ -0,0 +1,32 @@
name: Unit Tests
on: [workflow_call]
jobs:
acceptance-tests-labels:
name: Unit Tests
strategy:
fail-fast: false
matrix:
os: [ubuntu-latest, windows-2019, windows-latest]
runs-on: ${{matrix.os}}
defaults:
run:
shell: bash
steps:
- name: Checkout repository
if: ${{ runner.os == 'Linux' }}
uses: actions/checkout@v3
with:
fetch-depth: 1
- name: Checkout repository
if: ${{ runner.os == 'Windows' }}
uses: actions/checkout@v3
with:
fetch-depth: 1
submodules: true
- name: DiskANN Build CLI Applications
uses: ./.github/actions/build
- name: Run Unit Tests
run: |
cd build
ctest -C Release