Add multilingual HellaSwag task (#1228)

* add hellaswag_nl

* add other languages and update readme to hellaswag

* refactor as new task

* update readme

* add endline to yaml files and readme.md

* add group, change folder location and update yaml file

* rename default hellaswag yaml file

* fix whitespace error in some labels

* downgrade log level of whitespace checking

---------
Co-authored-by: JorgeDeCorte <jorge.decorte@ravago.be>
Co-authored-by: Hailey Schoelkopf <65563625+haileyschoelkopf@users.noreply.github.com>

lm_eval/tasks/okapi/hellaswag_multilingual/hellaswag_mr.yaml

+include: _hellaswag_yaml
+task: hellaswag_mr
+dataset_path: alexandrainst/m_hellaswag
+dataset_name: mr
+training_split: null
+validation_split: val

lm_eval/tasks/okapi/hellaswag_multilingual/hellaswag_ne.yaml

+include: _hellaswag_yaml
+task: hellaswag_ne
+dataset_path: alexandrainst/m_hellaswag
+dataset_name: ne
+training_split: null
+validation_split: val

lm_eval/tasks/okapi/hellaswag_multilingual/hellaswag_nl.yaml

+include: _hellaswag_yaml
+task: hellaswag_nl
+dataset_path: alexandrainst/m_hellaswag
+dataset_name: nl
+training_split: null
+validation_split: val

lm_eval/tasks/okapi/hellaswag_multilingual/hellaswag_pt.yaml

+include: _hellaswag_yaml
+task: hellaswag_pt
+dataset_path: alexandrainst/m_hellaswag
+dataset_name: pt
+training_split: null
+validation_split: val

lm_eval/tasks/okapi/hellaswag_multilingual/hellaswag_ro.yaml

+include: _hellaswag_yaml
+task: hellaswag_ro
+dataset_path: alexandrainst/m_hellaswag
+dataset_name: ro
+training_split: null
+validation_split: val

lm_eval/tasks/okapi/hellaswag_multilingual/hellaswag_ru.yaml

+include: _hellaswag_yaml
+task: hellaswag_ru
+dataset_path: alexandrainst/m_hellaswag
+dataset_name: ru
+training_split: null
+validation_split: val

lm_eval/tasks/okapi/hellaswag_multilingual/hellaswag_sk.yaml

+include: _hellaswag_yaml
+task: hellaswag_sk
+dataset_path: alexandrainst/m_hellaswag
+dataset_name: sk
+training_split: null
+validation_split: val

lm_eval/tasks/okapi/hellaswag_multilingual/hellaswag_sr.yaml

+include: _hellaswag_yaml
+task: hellaswag_sr
+dataset_path: alexandrainst/m_hellaswag
+dataset_name: sr
+training_split: null
+validation_split: val

lm_eval/tasks/okapi/hellaswag_multilingual/hellaswag_sv.yaml

+include: _hellaswag_yaml
+task: hellaswag_sv
+dataset_path: alexandrainst/m_hellaswag
+dataset_name: sv
+training_split: null
+validation_split: val

lm_eval/tasks/okapi/hellaswag_multilingual/hellaswag_ta.yaml

+include: _hellaswag_yaml
+task: hellaswag_ta
+dataset_path: alexandrainst/m_hellaswag
+dataset_name: ta
+training_split: null
+validation_split: val

lm_eval/tasks/okapi/hellaswag_multilingual/hellaswag_te.yaml

+include: _hellaswag_yaml
+task: hellaswag_te
+dataset_path: alexandrainst/m_hellaswag
+dataset_name: te
+training_split: null
+validation_split: val

lm_eval/tasks/okapi/hellaswag_multilingual/hellaswag_uk.yaml

+include: _hellaswag_yaml
+task: hellaswag_uk
+dataset_path: alexandrainst/m_hellaswag
+dataset_name: uk
+training_split: null
+validation_split: val

lm_eval/tasks/okapi/hellaswag_multilingual/hellaswag_vi.yaml

+include: _hellaswag_yaml
+task: hellaswag_vi
+dataset_path: alexandrainst/m_hellaswag
+dataset_name: vi
+training_split: null
+validation_split: val

lm_eval/tasks/okapi/hellaswag_multilingual/utils.py

+import datasets
+import re
+
+
+def preprocess(text):
+    text = text.strip()
+    # NOTE: Brackets are artifacts of the WikiHow dataset portion of HellaSwag.
+    text = text.replace(" [title]", ". ")
+    text = re.sub("\\[.*?\\]", "", text)
+    text = text.replace("  ", " ")
+    return text
+
+
+def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:
+    def _process_doc(doc):
+        ctx = doc["ctx_a"] + " " + doc["ctx_b"].capitalize()
+        out_doc = {
+            "query": preprocess(doc["activity_label"] + ": " + ctx),
+            "choices": [preprocess(ending) for ending in doc["endings"]],
+            "gold": int(doc["label"]),
+        }
+        return out_doc
+
+    return dataset.map(_process_doc)