18 lines
697 B
Markdown
18 lines
697 B
Markdown
|
German Passage Retrieval (MIRACL)
|
||
|
|
=====================================
|
||
|
|
|
||
|
|
German dev dataset from the [MIRACL challenge](https://project-miracl.github.io).
|
||
|
|
|
||
|
|
The dataset consists of 305 queries in the German language.
|
||
|
|
|
||
|
|
The `expected.tsv` file should contain tab-separated identifiers of passages relevant to the queries (max 10 per query). The identifiers correspond to passages that are located in the German split of the dataset found [here](https://huggingface.co/datasets/miracl/miracl-corpus).
|
||
|
|
|
||
|
|
Directory structure
|
||
|
|
-------------------
|
||
|
|
|
||
|
|
* `README.md` — this file
|
||
|
|
* `config.txt` — configuration file
|
||
|
|
* `test-A` — directory with test data
|
||
|
|
* `test-A/in.tsv` — input data for the test set
|
||
|
|
|