Merge branch 'cont-metrics' of https://github.com/EleutherAI/lm-evaluation-harness into alt_worlds

lm_eval/tasks/bigbench/generate_until/unnatural_in_context_learning.yaml

+# Generated by utils.py
+dataset_name: unnatural_in_context_learning_zero_shot
+include: ../generate_until_template_yaml
+task: bigbench_unnatural_in_context_learning_generate_until

lm_eval/tasks/bigbench/generate_until/vitaminc_fact_verification.yaml

+# Generated by utils.py
+dataset_name: vitaminc_fact_verification_zero_shot
+include: ../generate_until_template_yaml
+task: bigbench_vitaminc_fact_verification_generate_until

lm_eval/tasks/bigbench/generate_until/what_is_the_tao.yaml

+# Generated by utils.py
+dataset_name: what_is_the_tao_zero_shot
+include: ../generate_until_template_yaml
+task: bigbench_what_is_the_tao_generate_until

lm_eval/tasks/bigbench/generate_until/which_wiki_edit.yaml

+# Generated by utils.py
+dataset_name: which_wiki_edit_zero_shot
+include: ../generate_until_template_yaml
+task: bigbench_which_wiki_edit_generate_until

lm_eval/tasks/bigbench/generate_until/winowhy.yaml

+# Generated by utils.py
+dataset_name: winowhy_zero_shot
+include: ../generate_until_template_yaml
+task: bigbench_winowhy_generate_until

lm_eval/tasks/bigbench/generate_until/word_sorting.yaml

+# Generated by utils.py
+dataset_name: word_sorting_zero_shot
+include: ../generate_until_template_yaml
+task: bigbench_word_sorting_generate_until

lm_eval/tasks/bigbench/generate_until/word_unscrambling.yaml

+# Generated by utils.py
+dataset_name: word_unscrambling_zero_shot
+include: ../generate_until_template_yaml
+task: bigbench_word_unscrambling_generate_until

lm_eval/tasks/bigbench/generate_until_template_yaml

+group: bigbench
+dataset_path: bigbench # will switch to `hails/bigbench` when all tasks are pushed
+output_type: generate_until
+dataset_kwargs:
+  # num_shots: 0 # TODO: num of shots for `bigbench` HF dataset should be controlled through this, not through the typical methods
+  # subtask_name: null
+test_split: default
+doc_to_text: inputs
+doc_to_target: "{{targets[0]}}"
+generation_kwargs:
+  max_length: 128
+metric_list:
+  - metric: exact_match
+    aggregation: mean
+    higher_is_better: true
+    ignore_punctuation: true
+  - metric: !function aux_metric.token_edit_distance # pip install textdistance
+    aggregation: mean
+    higher_is_better: false

lm_eval/tasks/bigbench/greedy_until/abstract_narrative_understanding.yaml

-# Generated by utils.py
-dataset_name: abstract_narrative_understanding_zero_shot
-include: ../greedy_until_template_yaml
-task: bigbench_abstract_narrative_understanding_greedy_until

lm_eval/tasks/bigbench/greedy_until/anachronisms.yaml

-# Generated by utils.py
-dataset_name: anachronisms_zero_shot
-include: ../greedy_until_template_yaml
-task: bigbench_anachronisms_greedy_until

lm_eval/tasks/bigbench/greedy_until/analogical_similarity.yaml

-# Generated by utils.py
-dataset_name: analogical_similarity_zero_shot
-include: ../greedy_until_template_yaml
-task: bigbench_analogical_similarity_greedy_until

lm_eval/tasks/bigbench/greedy_until/analytic_entailment.yaml

-# Generated by utils.py
-dataset_name: analytic_entailment_zero_shot
-include: ../greedy_until_template_yaml
-task: bigbench_analytic_entailment_greedy_until

lm_eval/tasks/bigbench/greedy_until/arithmetic.yaml

-# Generated by utils.py
-dataset_name: arithmetic_zero_shot
-include: ../greedy_until_template_yaml
-task: bigbench_arithmetic_greedy_until

lm_eval/tasks/bigbench/greedy_until/ascii_word_recognition.yaml

-# Generated by utils.py
-dataset_name: ascii_word_recognition_zero_shot
-include: ../greedy_until_template_yaml
-task: bigbench_ascii_word_recognition_greedy_until

lm_eval/tasks/bigbench/greedy_until/authorship_verification.yaml

-# Generated by utils.py
-dataset_name: authorship_verification_zero_shot
-include: ../greedy_until_template_yaml
-task: bigbench_authorship_verification_greedy_until

lm_eval/tasks/bigbench/greedy_until/auto_categorization.yaml

-# Generated by utils.py
-dataset_name: auto_categorization_zero_shot
-include: ../greedy_until_template_yaml
-task: bigbench_auto_categorization_greedy_until

lm_eval/tasks/bigbench/greedy_until/auto_debugging.yaml

-# Generated by utils.py
-dataset_name: auto_debugging_zero_shot
-include: ../greedy_until_template_yaml
-task: bigbench_auto_debugging_greedy_until

lm_eval/tasks/bigbench/greedy_until/bbq_lite_json.yaml

-# Generated by utils.py
-dataset_name: bbq_lite_json_zero_shot
-include: ../greedy_until_template_yaml
-task: bigbench_bbq_lite_json_greedy_until

lm_eval/tasks/bigbench/greedy_until/bridging_anaphora_resolution_barqa.yaml

-# Generated by utils.py
-dataset_name: bridging_anaphora_resolution_barqa_zero_shot
-include: ../greedy_until_template_yaml
-task: bigbench_bridging_anaphora_resolution_barqa_greedy_until

lm_eval/tasks/bigbench/greedy_until/causal_judgment.yaml

-# Generated by utils.py
-dataset_name: causal_judgment_zero_shot
-include: ../greedy_until_template_yaml
-task: bigbench_causal_judgment_greedy_until