Commit
•
65fb8a4
1
Parent(s):
a113b49
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,29 @@
|
|
| 1 |
---
|
| 2 |
license: apache-2.0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
license: apache-2.0
|
| 3 |
+
task_categories:
|
| 4 |
+
- sentence-similarity
|
| 5 |
+
pretty_name: Wikipedia UForm Embeddings for Nearest Neighbors Search
|
| 6 |
+
size_categories:
|
| 7 |
+
- 1M<n<10M
|
| 8 |
---
|
| 9 |
+
|
| 10 |
+
## Dataset Summary
|
| 11 |
+
|
| 12 |
+
This dataset contains 256-dimensional vectors for a 1M sample of Wikipedia for Approximate Nearest Neighbors Search benchmarks.
|
| 13 |
+
|
| 14 |
+
### Usage
|
| 15 |
+
|
| 16 |
+
```
|
| 17 |
+
git lfs install
|
| 18 |
+
git clone https://huggingface.co/datasets/unum-cloud/ann-wiki-1m
|
| 19 |
+
```
|
| 20 |
+
|
| 21 |
+
### Dataset Structure
|
| 22 |
+
|
| 23 |
+
The dataset contains three matrices:
|
| 24 |
+
|
| 25 |
+
- base: `base.1M.fbin` with 1M vectors to construct the index.
|
| 26 |
+
- query: `query.public.100K.fbin` with 100K vectors to lookup in the index.
|
| 27 |
+
- truth: `groundtruth.public.100K.ibin` with 10x results for every one of the 100K queries.
|
| 28 |
+
|
| 29 |
+
Use the [ashvardanian/read_matrix.py](https://gist.github.com/ashvardanian/301b0614252941ac8a3137ac72a18892) Gist to parse the files.
|