-
Notifications
You must be signed in to change notification settings - Fork 24
vector indexing gt modified #45
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.
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
Open
patrickkon
wants to merge
4
commits into
main
Choose a base branch
from
vector-indexing-gt
base: main
Could not load branches
Branch not found: {{ refName }}
Loading
Could not load tags
Nothing to show
Loading
Are you sure you want to change the base?
Some commits from the old base branch may be removed from the timeline,
and old review comments may become outdated.
Open
Changes from all commits
Commits
File filter
Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
There are no files selected for viewing
6 changes: 2 additions & 4 deletions
6
benchmark/experimentation_bench/vector_index/ground_truth/q11.txt
This file contains hidden or 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
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -1,6 +1,4 @@ | ||
| Setup: check that np.random is used to construct the dataset, and that it is somehow passed and used as the dataset in the HNSW index. | ||
| Variables: check that mean, variance and dataset size are all varied individually and tested to see their effects. | ||
| Expected result: | ||
| 1. **Effect on Increasing Data Size**: The HNSW index generally maintains high recall as data size increases, with occasional slight performance drops in certain cases. | ||
| 1. **Effect on Increasing Data Size**: Increasing data size set leads to a slightly decrease in recall. | ||
| 2. **Effect on Increasing Mean**: The recall is not significantly affected by changes in the mean, showing robustness. | ||
| 3. **Effect on Increasing Variance**: The recall remains mostly stable, with some rare instances of decreased performance. | ||
| 3. **Effect on Increasing Variance**: Higher variance leads to a decline in recall. | ||
2 changes: 1 addition & 1 deletion
2
benchmark/experimentation_bench/vector_index/ground_truth/q2.txt
This file contains hidden or 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
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -1 +1 @@ | ||
| Using higher \( M \) values improves recall with manageable increases in query latency, making it a favorable trade-off for applications prioritizing accuracy. | ||
| Using higher \( M \) values improves recall with manageable increases in query latency and memory usage, making it a favorable trade-off for applications prioritizing accuracy. |
This file contains hidden or 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
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1 @@ | ||
| Memory usage remains same as efSearch varies. With M=16, setting efSearch to 64 or higher achieves at least 90% recall while minimizing memory usage. | ||
yiliazhuxyxy marked this conversation as resolved.
Show resolved
Hide resolved
|
||
5 changes: 1 addition & 4 deletions
5
benchmark/experimentation_bench/vector_index/ground_truth/q7.txt
This file contains hidden or 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
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -1,4 +1 @@ | ||
| Check that the script installs a package for memory measurement or uses some utility that can legitimately measure memory used. | ||
|
|
||
| HNSW should be better, something like this: | ||
| The HNSW index type achieves the required recall rate of at least 96% and does so with significantly less memory usage compared to the IVF index type. | ||
| The IVF index type achieves the required recall rate of at least 96% and does so with significantly less memory usage compared to the HNSW index type. |
This file contains hidden or 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
This file contains hidden or 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
141 changes: 141 additions & 0 deletions
141
benchmark/experimentation_bench/vector_index/raw_data/q11_data.txt
This file contains hidden or 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
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,141 @@ | ||
| All experiment based on fixed input argument: 10 32 40 | ||
|
|
||
| Effect on Increasing Data Size: | ||
|
|
||
| nb = 50000 | ||
| efSearch 16 bounded queue True 0.002 ms per query, R@1 0.8257, missing rate 0.0000 | ||
| efSearch 16 bounded queue False 0.001 ms per query, R@1 0.8257, missing rate 0.0000 | ||
| efSearch 32 bounded queue True 0.002 ms per query, R@1 0.9368, missing rate 0.0000 | ||
| efSearch 32 bounded queue False 0.002 ms per query, R@1 0.9368, missing rate 0.0000 | ||
| efSearch 64 bounded queue True 0.004 ms per query, R@1 0.9832, missing rate 0.0000 | ||
| efSearch 64 bounded queue False 0.004 ms per query, R@1 0.9832, missing rate 0.0000 | ||
| efSearch 128 bounded queue True 0.008 ms per query, R@1 0.9964, missing rate 0.0000 | ||
| efSearch 128 bounded queue False 0.007 ms per query, R@1 0.9964, missing rate 0.0000 | ||
| efSearch 256 bounded queue True 0.013 ms per query, R@1 0.9995, missing rate 0.0000 | ||
| efSearch 256 bounded queue False 0.013 ms per query, R@1 0.9995, missing rate 0.0000 | ||
| nb = 75000 | ||
| efSearch 16 bounded queue True 0.002 ms per query, R@1 0.8250, missing rate 0.0000 | ||
| efSearch 16 bounded queue False 0.002 ms per query, R@1 0.8250, missing rate 0.0000 | ||
| efSearch 32 bounded queue True 0.002 ms per query, R@1 0.9322, missing rate 0.0000 | ||
| efSearch 32 bounded queue False 0.002 ms per query, R@1 0.9322, missing rate 0.0000 | ||
| efSearch 64 bounded queue True 0.004 ms per query, R@1 0.9837, missing rate 0.0000 | ||
| efSearch 64 bounded queue False 0.004 ms per query, R@1 0.9837, missing rate 0.0000 | ||
| efSearch 128 bounded queue True 0.007 ms per query, R@1 0.9977, missing rate 0.0000 | ||
| efSearch 128 bounded queue False 0.007 ms per query, R@1 0.9977, missing rate 0.0000 | ||
| efSearch 256 bounded queue True 0.014 ms per query, R@1 0.9995, missing rate 0.0000 | ||
| efSearch 256 bounded queue False 0.013 ms per query, R@1 0.9995, missing rate 0.0000 | ||
| nb = 100000 | ||
| efSearch 16 bounded queue True 0.002 ms per query, R@1 0.8326, missing rate 0.0000 | ||
| efSearch 16 bounded queue False 0.002 ms per query, R@1 0.8326, missing rate 0.0000 | ||
| efSearch 32 bounded queue True 0.003 ms per query, R@1 0.9371, missing rate 0.0000 | ||
| efSearch 32 bounded queue False 0.002 ms per query, R@1 0.9371, missing rate 0.0000 | ||
| efSearch 64 bounded queue True 0.004 ms per query, R@1 0.9853, missing rate 0.0000 | ||
| efSearch 64 bounded queue False 0.004 ms per query, R@1 0.9853, missing rate 0.0000 | ||
| efSearch 128 bounded queue True 0.007 ms per query, R@1 0.9970, missing rate 0.0000 | ||
| efSearch 128 bounded queue False 0.007 ms per query, R@1 0.9970, missing rate 0.0000 | ||
| efSearch 256 bounded queue True 0.013 ms per query, R@1 0.9996, missing rate 0.0000 | ||
| efSearch 256 bounded queue False 0.012 ms per query, R@1 0.9996, missing rate 0.0000 | ||
| nb = 125000 | ||
| efSearch 16 bounded queue True 0.001 ms per query, R@1 0.8247, missing rate 0.0000 | ||
| efSearch 16 bounded queue False 0.001 ms per query, R@1 0.8247, missing rate 0.0000 | ||
| efSearch 32 bounded queue True 0.002 ms per query, R@1 0.9336, missing rate 0.0000 | ||
| efSearch 32 bounded queue False 0.002 ms per query, R@1 0.9336, missing rate 0.0000 | ||
| efSearch 64 bounded queue True 0.004 ms per query, R@1 0.9815, missing rate 0.0000 | ||
| efSearch 64 bounded queue False 0.004 ms per query, R@1 0.9815, missing rate 0.0000 | ||
| efSearch 128 bounded queue True 0.007 ms per query, R@1 0.9965, missing rate 0.0000 | ||
| efSearch 128 bounded queue False 0.007 ms per query, R@1 0.9965, missing rate 0.0000 | ||
| efSearch 256 bounded queue True 0.014 ms per query, R@1 0.9995, missing rate 0.0000 | ||
| efSearch 256 bounded queue False 0.013 ms per query, R@1 0.9995, missing rate 0.0000 | ||
|
|
||
| Effect on Increasing Mean: | ||
| Mean= 0 | ||
| efSearch 16 bounded queue True 0.001 ms per query, R@1 0.8297, missing rate 0.0000 | ||
| efSearch 16 bounded queue False 0.002 ms per query, R@1 0.8297, missing rate 0.0000 | ||
| efSearch 32 bounded queue True 0.002 ms per query, R@1 0.9364, missing rate 0.0000 | ||
| efSearch 32 bounded queue False 0.002 ms per query, R@1 0.9364, missing rate 0.0000 | ||
| efSearch 64 bounded queue True 0.004 ms per query, R@1 0.9852, missing rate 0.0000 | ||
| efSearch 64 bounded queue False 0.005 ms per query, R@1 0.9852, missing rate 0.0000 | ||
| efSearch 128 bounded queue True 0.008 ms per query, R@1 0.9971, missing rate 0.0000 | ||
| efSearch 128 bounded queue False 0.010 ms per query, R@1 0.9971, missing rate 0.0000 | ||
| efSearch 256 bounded queue True 0.017 ms per query, R@1 0.9999, missing rate 0.0000 | ||
| efSearch 256 bounded queue False 0.014 ms per query, R@1 0.9999, missing rate 0.0000 | ||
| Mean= 1 | ||
| efSearch 16 bounded queue True 0.002 ms per query, R@1 0.8332, missing rate 0.0000 | ||
| efSearch 16 bounded queue False 0.002 ms per query, R@1 0.8332, missing rate 0.0000 | ||
| efSearch 32 bounded queue True 0.003 ms per query, R@1 0.9367, missing rate 0.0000 | ||
| efSearch 32 bounded queue False 0.002 ms per query, R@1 0.9367, missing rate 0.0000 | ||
| efSearch 64 bounded queue True 0.004 ms per query, R@1 0.9865, missing rate 0.0000 | ||
| efSearch 64 bounded queue False 0.004 ms per query, R@1 0.9865, missing rate 0.0000 | ||
| efSearch 128 bounded queue True 0.008 ms per query, R@1 0.9976, missing rate 0.0000 | ||
| efSearch 128 bounded queue False 0.007 ms per query, R@1 0.9976, missing rate 0.0000 | ||
| efSearch 256 bounded queue True 0.014 ms per query, R@1 0.9998, missing rate 0.0000 | ||
| efSearch 256 bounded queue False 0.013 ms per query, R@1 0.9998, missing rate 0.0000 | ||
| Mean= 2 | ||
| efSearch 16 bounded queue True 0.001 ms per query, R@1 0.8292, missing rate 0.0000 | ||
| efSearch 16 bounded queue False 0.001 ms per query, R@1 0.8292, missing rate 0.0000 | ||
| efSearch 32 bounded queue True 0.002 ms per query, R@1 0.9364, missing rate 0.0000 | ||
| efSearch 32 bounded queue False 0.002 ms per query, R@1 0.9364, missing rate 0.0000 | ||
| efSearch 64 bounded queue True 0.004 ms per query, R@1 0.9848, missing rate 0.0000 | ||
| efSearch 64 bounded queue False 0.004 ms per query, R@1 0.9848, missing rate 0.0000 | ||
| efSearch 128 bounded queue True 0.007 ms per query, R@1 0.9972, missing rate 0.0000 | ||
| efSearch 128 bounded queue False 0.007 ms per query, R@1 0.9972, missing rate 0.0000 | ||
| efSearch 256 bounded queue True 0.013 ms per query, R@1 0.9998, missing rate 0.0000 | ||
| efSearch 256 bounded queue False 0.013 ms per query, R@1 0.9998, missing rate 0.0000 | ||
| Mean= 3 | ||
| efSearch 16 bounded queue True 0.001 ms per query, R@1 0.8276, missing rate 0.0000 | ||
| efSearch 16 bounded queue False 0.001 ms per query, R@1 0.8276, missing rate 0.0000 | ||
| efSearch 32 bounded queue True 0.003 ms per query, R@1 0.9359, missing rate 0.0000 | ||
| efSearch 32 bounded queue False 0.002 ms per query, R@1 0.9359, missing rate 0.0000 | ||
| efSearch 64 bounded queue True 0.004 ms per query, R@1 0.9855, missing rate 0.0000 | ||
| efSearch 64 bounded queue False 0.005 ms per query, R@1 0.9855, missing rate 0.0000 | ||
| efSearch 128 bounded queue True 0.008 ms per query, R@1 0.9974, missing rate 0.0000 | ||
| efSearch 128 bounded queue False 0.008 ms per query, R@1 0.9974, missing rate 0.0000 | ||
| efSearch 256 bounded queue True 0.013 ms per query, R@1 0.9998, missing rate 0.0000 | ||
| efSearch 256 bounded queue False 0.013 ms per query, R@1 0.9998, missing rate 0.0000 | ||
|
|
||
|
|
||
| Effect on Increasing Variance: | ||
| Std=1 | ||
| efSearch 16 bounded queue True 0.002 ms per query, R@1 0.8337, missing rate 0.0000 | ||
| efSearch 16 bounded queue False 0.001 ms per query, R@1 0.8337, missing rate 0.0000 | ||
| efSearch 32 bounded queue True 0.002 ms per query, R@1 0.9365, missing rate 0.0000 | ||
| efSearch 32 bounded queue False 0.002 ms per query, R@1 0.9365, missing rate 0.0000 | ||
| efSearch 64 bounded queue True 0.004 ms per query, R@1 0.9844, missing rate 0.0000 | ||
| efSearch 64 bounded queue False 0.004 ms per query, R@1 0.9844, missing rate 0.0000 | ||
| efSearch 128 bounded queue True 0.007 ms per query, R@1 0.9972, missing rate 0.0000 | ||
| efSearch 128 bounded queue False 0.007 ms per query, R@1 0.9972, missing rate 0.0000 | ||
| efSearch 256 bounded queue True 0.013 ms per query, R@1 0.9997, missing rate 0.0000 | ||
| efSearch 256 bounded queue False 0.013 ms per query, R@1 0.9997, missing rate 0.0000 | ||
| Std=2 | ||
| efSearch 16 bounded queue True 0.001 ms per query, R@1 0.7662, missing rate 0.0000 | ||
| efSearch 16 bounded queue False 0.002 ms per query, R@1 0.7662, missing rate 0.0000 | ||
| efSearch 32 bounded queue True 0.002 ms per query, R@1 0.8961, missing rate 0.0000 | ||
| efSearch 32 bounded queue False 0.002 ms per query, R@1 0.8961, missing rate 0.0000 | ||
| efSearch 64 bounded queue True 0.004 ms per query, R@1 0.9688, missing rate 0.0000 | ||
| efSearch 64 bounded queue False 0.004 ms per query, R@1 0.9688, missing rate 0.0000 | ||
| efSearch 128 bounded queue True 0.008 ms per query, R@1 0.9933, missing rate 0.0000 | ||
| efSearch 128 bounded queue False 0.008 ms per query, R@1 0.9933, missing rate 0.0000 | ||
| efSearch 256 bounded queue True 0.015 ms per query, R@1 0.9984, missing rate 0.0000 | ||
| efSearch 256 bounded queue False 0.015 ms per query, R@1 0.9984, missing rate 0.0000 | ||
| Std=3 | ||
| efSearch 16 bounded queue True 0.002 ms per query, R@1 0.7324, missing rate 0.0000 | ||
| efSearch 16 bounded queue False 0.002 ms per query, R@1 0.7324, missing rate 0.0000 | ||
| efSearch 32 bounded queue True 0.003 ms per query, R@1 0.8702, missing rate 0.0000 | ||
| efSearch 32 bounded queue False 0.003 ms per query, R@1 0.8702, missing rate 0.0000 | ||
| efSearch 64 bounded queue True 0.005 ms per query, R@1 0.9541, missing rate 0.0000 | ||
| efSearch 64 bounded queue False 0.005 ms per query, R@1 0.9541, missing rate 0.0000 | ||
| efSearch 128 bounded queue True 0.009 ms per query, R@1 0.9895, missing rate 0.0000 | ||
| efSearch 128 bounded queue False 0.009 ms per query, R@1 0.9895, missing rate 0.0000 | ||
| efSearch 256 bounded queue True 0.017 ms per query, R@1 0.9974, missing rate 0.0000 | ||
| efSearch 256 bounded queue False 0.016 ms per query, R@1 0.9974, missing rate 0.0000 | ||
| Std=4 | ||
| efSearch 16 bounded queue True 0.002 ms per query, R@1 0.7081, missing rate 0.0000 | ||
| efSearch 16 bounded queue False 0.002 ms per query, R@1 0.7081, missing rate 0.0000 | ||
| efSearch 32 bounded queue True 0.003 ms per query, R@1 0.8555, missing rate 0.0000 | ||
| efSearch 32 bounded queue False 0.003 ms per query, R@1 0.8555, missing rate 0.0000 | ||
| efSearch 64 bounded queue True 0.005 ms per query, R@1 0.9472, missing rate 0.0000 | ||
| efSearch 64 bounded queue False 0.005 ms per query, R@1 0.9472, missing rate 0.0000 | ||
| efSearch 128 bounded queue True 0.010 ms per query, R@1 0.9870, missing rate 0.0000 | ||
| efSearch 128 bounded queue False 0.010 ms per query, R@1 0.9870, missing rate 0.0000 | ||
| efSearch 256 bounded queue True 0.019 ms per query, R@1 0.9966, missing rate 0.0000 | ||
| efSearch 256 bounded queue False 0.018 ms per query, R@1 0.9966, missing rate 0.0000 |
10 changes: 10 additions & 0 deletions
10
benchmark/experimentation_bench/vector_index/raw_data/q1_data.txt
This file contains hidden or 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
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,10 @@ | ||
| efSearch 16 bounded queue True 0.008 ms per query, R@1 0.9054, missing rate 0.0000 | ||
| efSearch 16 bounded queue False 0.008 ms per query, R@1 0.9054, missing rate 0.0000 | ||
| efSearch 32 bounded queue True 0.013 ms per query, R@1 0.9599, missing rate 0.0000 | ||
| efSearch 32 bounded queue False 0.012 ms per query, R@1 0.9599, missing rate 0.0000 | ||
| efSearch 64 bounded queue True 0.021 ms per query, R@1 0.9859, missing rate 0.0000 | ||
| efSearch 64 bounded queue False 0.022 ms per query, R@1 0.9859, missing rate 0.0000 | ||
| efSearch 128 bounded queue True 0.039 ms per query, R@1 0.9962, missing rate 0.0000 | ||
| efSearch 128 bounded queue False 0.038 ms per query, R@1 0.9962, missing rate 0.0000 | ||
| efSearch 256 bounded queue True 0.065 ms per query, R@1 0.9989, missing rate 0.0000 | ||
| efSearch 256 bounded queue False 0.062 ms per query, R@1 0.9989, missing rate 0.0000 |
35 changes: 35 additions & 0 deletions
35
benchmark/experimentation_bench/vector_index/raw_data/q2_data.txt
This file contains hidden or 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
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,35 @@ | ||
| M=16 | ||
| efSearch 16 bounded queue True 0.004 ms per query, R@1 0.8015, missing rate 0.0000, memory usage 1280532.00 KB | ||
| efSearch 16 bounded queue False 0.004 ms per query, R@1 0.8017, missing rate 0.0000, memory usage 1280532.00 KB | ||
| efSearch 32 bounded queue True 0.007 ms per query, R@1 0.8912, missing rate 0.0000, memory usage 1280532.00 KB | ||
| efSearch 32 bounded queue False 0.006 ms per query, R@1 0.8913, missing rate 0.0000, memory usage 1280532.00 KB | ||
| efSearch 64 bounded queue True 0.010 ms per query, R@1 0.9490, missing rate 0.0000, memory usage 1280532.00 KB | ||
| efSearch 64 bounded queue False 0.011 ms per query, R@1 0.9491, missing rate 0.0000, memory usage 1280532.00 KB | ||
| efSearch 128 bounded queue True 0.020 ms per query, R@1 0.9797, missing rate 0.0000, memory usage 1280532.00 KB | ||
| efSearch 128 bounded queue False 0.019 ms per query, R@1 0.9797, missing rate 0.0000, memory usage 1280532.00 KB | ||
| efSearch 256 bounded queue True 0.037 ms per query, R@1 0.9930, missing rate 0.0000, memory usage 1280532.00 KB | ||
| efSearch 256 bounded queue False 0.033 ms per query, R@1 0.9930, missing rate 0.0000, memory usage 1280532.00 KB | ||
|
|
||
| M=24 | ||
| efSearch 16 bounded queue True 0.007 ms per query, R@1 0.8827, missing rate 0.0000, memory usage 1336632.00 KB | ||
| efSearch 16 bounded queue False 0.006 ms per query, R@1 0.8828, missing rate 0.0000, memory usage 1336632.00 KB | ||
| efSearch 32 bounded queue True 0.010 ms per query, R@1 0.9487, missing rate 0.0000, memory usage 1336632.00 KB | ||
| efSearch 32 bounded queue False 0.010 ms per query, R@1 0.9488, missing rate 0.0000, memory usage 1336632.00 KB | ||
| efSearch 64 bounded queue True 0.016 ms per query, R@1 0.9815, missing rate 0.0000, memory usage 1336632.00 KB | ||
| efSearch 64 bounded queue False 0.017 ms per query, R@1 0.9815, missing rate 0.0000, memory usage 1336632.00 KB | ||
| efSearch 128 bounded queue True 0.030 ms per query, R@1 0.9944, missing rate 0.0000, memory usage 1336632.00 KB | ||
| efSearch 128 bounded queue False 0.029 ms per query, R@1 0.9944, missing rate 0.0000, memory usage 1336632.00 KB | ||
| efSearch 256 bounded queue True 0.053 ms per query, R@1 0.9987, missing rate 0.0000, memory usage 1336632.00 KB | ||
| efSearch 256 bounded queue False 0.055 ms per query, R@1 0.9987, missing rate 0.0000, memory usage 1336632.00 KB | ||
|
|
||
| M=32 | ||
| efSearch 16 bounded queue True 0.009 ms per query, R@1 0.9088, missing rate 0.0000, memory usage 1405588.00 KB | ||
| efSearch 16 bounded queue False 0.008 ms per query, R@1 0.9088, missing rate 0.0000, memory usage 1405588.00 KB | ||
| efSearch 32 bounded queue True 0.012 ms per query, R@1 0.9623, missing rate 0.0000, memory usage 1405588.00 KB | ||
| efSearch 32 bounded queue False 0.012 ms per query, R@1 0.9623, missing rate 0.0000, memory usage 1405588.00 KB | ||
| efSearch 64 bounded queue True 0.020 ms per query, R@1 0.9872, missing rate 0.0000, memory usage 1405588.00 KB | ||
| efSearch 64 bounded queue False 0.021 ms per query, R@1 0.9873, missing rate 0.0000, memory usage 1405588.00 KB | ||
| efSearch 128 bounded queue True 0.036 ms per query, R@1 0.9964, missing rate 0.0000, memory usage 1405588.00 KB | ||
| efSearch 128 bounded queue False 0.035 ms per query, R@1 0.9964, missing rate 0.0000, memory usage 1405588.00 KB | ||
| efSearch 256 bounded queue True 0.063 ms per query, R@1 0.9992, missing rate 0.0000, memory usage 1405588.00 KB | ||
| efSearch 256 bounded queue False 0.063 ms per query, R@1 0.9992, missing rate 0.0000, memory usage 1405588.00 KB |
Oops, something went wrong.
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.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think the previous ground truth is more correct
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
When the variance increases, the vectors become more dispersed, making it harder to capture all neighboring points, and the recall rate decreases. Does this reasoning make sense?