Merge branch 'big-refactor' of https://github.com/EleutherAI/lm-evaluation-harness into xstorycloze

lm_eval/tasks/blimp/tough_vs_raising_1.yaml

+# Generated by utils.py
+dataset_name: tough_vs_raising_1
+include: template_yaml
+task: blimp_tough_vs_raising_1

lm_eval/tasks/blimp/tough_vs_raising_2.yaml

+# Generated by utils.py
+dataset_name: tough_vs_raising_2
+include: template_yaml
+task: blimp_tough_vs_raising_2

lm_eval/tasks/blimp/transitive.yaml

+# Generated by utils.py
+dataset_name: transitive
+include: template_yaml
+task: blimp_transitive

lm_eval/tasks/blimp/wh_island.yaml

+# Generated by utils.py
+dataset_name: wh_island
+include: template_yaml
+task: blimp_wh_island

lm_eval/tasks/blimp/wh_questions_object_gap.yaml

+# Generated by utils.py
+dataset_name: wh_questions_object_gap
+include: template_yaml
+task: blimp_wh_questions_object_gap

lm_eval/tasks/blimp/wh_questions_subject_gap.yaml

+# Generated by utils.py
+dataset_name: wh_questions_subject_gap
+include: template_yaml
+task: blimp_wh_questions_subject_gap

lm_eval/tasks/blimp/wh_questions_subject_gap_long_distance.yaml

+# Generated by utils.py
+dataset_name: wh_questions_subject_gap_long_distance
+include: template_yaml
+task: blimp_wh_questions_subject_gap_long_distance

lm_eval/tasks/blimp/wh_vs_that_no_gap.yaml

+# Generated by utils.py
+dataset_name: wh_vs_that_no_gap
+include: template_yaml
+task: blimp_wh_vs_that_no_gap

lm_eval/tasks/blimp/wh_vs_that_no_gap_long_distance.yaml

+# Generated by utils.py
+dataset_name: wh_vs_that_no_gap_long_distance
+include: template_yaml
+task: blimp_wh_vs_that_no_gap_long_distance

lm_eval/tasks/blimp/wh_vs_that_with_gap.yaml

+# Generated by utils.py
+dataset_name: wh_vs_that_with_gap
+include: template_yaml
+task: blimp_wh_vs_that_with_gap

lm_eval/tasks/blimp/wh_vs_that_with_gap_long_distance.yaml

+# Generated by utils.py
+dataset_name: wh_vs_that_with_gap_long_distance
+include: template_yaml
+task: blimp_wh_vs_that_with_gap_long_distance