fix: Added tests for checking datasets (embeddings-benchmark#307)

* fix: Fixed hf_hub_name for WikiCitiesClustering

* Added points for this PR and a 3 other minor dataset fixes

* feat: Added tests which validated that datasets are available

* fix: Updated hf references and revisions to multiple datasets

* Added points for submission

* fix: Added suggestions from the review

* Apply suggestions from code review

Co-authored-by: Imene Kerboua <[email protected]>

* fix: sped up async test for whether datasets exist

* fix: Updated revisions

* fix: reuploaded scandeval datasets

* fix: Applied formatter

---------

Co-authored-by: Imene Kerboua <[email protected]>

Author: KennethEnevoldsen

docs/mmteb/points.md

@@ -2,7 +2,7 @@
 
 | GitHub            | Total points | New dataset | New task | Dataset annotations | (Bug)fixes | Running Models | Review PR |  Paper Writing | Ideation | Coordination |
 | ----------------- | ------------ | ----------- | -------- | ------------------- | ---------- | -------------- |  -------- | -------------- | -------- | ------------- |
-| KennethEnevoldsen |              |   38        |          |                     |         4  |                |           |                |         |               |
+| KennethEnevoldsen |              |   38        |          |                     |         8  |                |           |                |         |               |
 | x-tabdeveloping   |              |    2        |          |                     |            |                |           |                |        |               |
 
 Note that coordination and ideation is not included in the total points, but is used to determine first and last authors. 

mteb/abstasks/TaskMetadata.py

@@ -9,7 +9,6 @@
     BeforeValidator,
     TypeAdapter,
     field_validator,
-    model_validator,
 )
 from typing_extensions import Annotated, Literal
 
@@ -172,7 +171,11 @@ def _check_dataset_path_is_specified(cls, dataset):
     @field_validator("dataset")
     def _check_dataset_revision_is_specified(cls, dataset):
         if "revision" not in dataset:
-            raise ValueError("You must explicitly specify a revision for the dataset (either a SHA or None).")
+            raise ValueError(
+                "You must explicitly specify a revision for the dataset (either a SHA or None)."
+            )
         if dataset["revision"] is None:
-            logging.warning("It is encourage to specify a dataset revision for reproducability")
+            logging.warning(
+                "It is encourage to specify a dataset revision for reproducability"
+            )
         return dataset