Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
## Dataset Summary
|
| 2 |
|
| 3 |
**SciFact-Fa** is a Persian (Farsi) dataset designed for the **Retrieval** task, with a focus on **scientific fact verification**. It is a translated version of the original English **SciFact** dataset used in the **BEIR benchmark** and is part of the **FaMTEB** (Farsi Massive Text Embedding Benchmark) under the BEIR-Fa collection.
|
|
|
|
| 1 |
+
---
|
| 2 |
+
configs:
|
| 3 |
+
- config_name: default
|
| 4 |
+
data_files:
|
| 5 |
+
- split: train
|
| 6 |
+
path: qrels/train.jsonl
|
| 7 |
+
- split: test
|
| 8 |
+
path: qrels/test.jsonl
|
| 9 |
+
- config_name: corpus
|
| 10 |
+
data_files:
|
| 11 |
+
- split: corpus
|
| 12 |
+
path: corpus.jsonl
|
| 13 |
+
- config_name: queries
|
| 14 |
+
data_files:
|
| 15 |
+
- split: queries
|
| 16 |
+
path: queries.jsonl
|
| 17 |
+
---
|
| 18 |
## Dataset Summary
|
| 19 |
|
| 20 |
**SciFact-Fa** is a Persian (Farsi) dataset designed for the **Retrieval** task, with a focus on **scientific fact verification**. It is a translated version of the original English **SciFact** dataset used in the **BEIR benchmark** and is part of the **FaMTEB** (Farsi Massive Text Embedding Benchmark) under the BEIR-Fa collection.
|