Merge remote-tracking branch 'origin/big-refactor' into big-refactor_dp

lm_eval/tasks/lambada_multilingual/lambada_mt_en.yaml

@@ -16,3 +16,5 @@ metric_list:
   - metric: acc
     aggregation: mean
     higher_is_better: true
+metadata:
+  - version: 1.0

lm_eval/tasks/logiqa/logiqa.yaml

@@ -17,3 +17,5 @@ metric_list:
   - metric: acc_norm
     aggregation: mean
     higher_is_better: true
+metadata:
+  - version: 1.0

lm_eval/tasks/logiqa2/logieval.yaml

@@ -23,3 +23,5 @@ filter_list:
         # https://github.com/openai/evals/blob/305b237cdb3884c7ddb6a5d12cb184a83551fcba/evals/api.py#L84
         regex_pattern: "^\\s*([A-D])"
       - function: "take_first"
+metadata:
+  - version: 0.0

lm_eval/tasks/logiqa2/logiqa2.yaml

@@ -17,3 +17,5 @@ metric_list:
   - metric: acc_norm
     aggregation: mean
     higher_is_better: true
+metadata:
+  - version: 0.0

lm_eval/tasks/mathqa/mathqa.yaml

@@ -18,3 +18,5 @@ metric_list:
   - metric: acc_norm
     aggregation: mean
     higher_is_better: true
+metadata:
+  - version: 1.0

lm_eval/tasks/mc_taco/default.yaml

@@ -11,3 +11,5 @@ doc_to_decontamination_query: "{{question}} {{sentence}}"
 metric_list:
   - metric: acc
   - metric: f1
+metadata:
+  - version: 1.0

lm_eval/tasks/mgsm/direct/direct_yaml

@@ -25,3 +25,5 @@ metric_list:
     higher_is_better: true
     ignore_case: true
     ignore_punctuation: true
+metadata:
+  - version: 0.0

lm_eval/tasks/mgsm/en_cot/cot_yaml

@@ -27,3 +27,5 @@ filter_list:
       - function: "regex"
         regex_pattern: "The answer is (\\-?[0-9\\.\\,]+)"
       - function: "take_first"
+metadata:
+  - version: 0.0

lm_eval/tasks/mgsm/native_cot/cot_yaml

@@ -27,3 +27,5 @@ filter_list:
       - function: "regex"
         regex_pattern: "The answer is (\\-?[0-9\\.\\,]+)"
       - function: "take_first"
+metadata:
+  - version: 1.0

lm_eval/tasks/minerva_math/minerva_math_algebra.yaml

@@ -19,3 +19,5 @@ metric_list:
   - metric: exact_match
     aggregation: mean
     higher_is_better: true
+metadata:
+  - version: 0.0

lm_eval/tasks/mmlu/default/_default_template_yaml

@@ -11,3 +11,5 @@ metric_list:
   - metric: acc
     aggregation: mean
     higher_is_better: true
+metadata:
+  - version: 0.0

lm_eval/tasks/mmlu/flan_cot_fewshot/_mmlu_flan_cot_fewshot_template_yaml

@@ -21,3 +21,5 @@ metric_list:
     higher_is_better: true
     ignore_case: true
     ignore_punctuation: true
+metadata:
+  - version: 0.0

lm_eval/tasks/mmlu/flan_cot_zeroshot/_mmlu_flan_cot_zeroshot_template_yaml

@@ -21,3 +21,5 @@ metric_list:
     higher_is_better: true
     ignore_case: true
     ignore_punctuation: true
+metadata:
+  - version: 0.0

lm_eval/tasks/mmlu/flan_n_shot/generative/_mmlu_flan_generative_template_yaml

@@ -12,3 +12,5 @@ metric_list:
   - metric: exact_match
     aggregation: mean
     higher_is_better: true
+metadata:
+  - version: 0.0

lm_eval/tasks/mmlu/flan_n_shot/loglikelihood/_mmlu_flan_loglikelihood_template_yaml

@@ -12,3 +12,5 @@ metric_list:
   - metric: acc_norm
     aggregation: mean
     higher_is_better: true
+metadata:
+  - version: 0.0

lm_eval/tasks/model_written_evals/advanced_ai_risk/_template_yaml

@@ -10,3 +10,5 @@ should_decontaminate: true
 doc_to_decontamination_query: "{{sentence_good}} {{sentence_bad}}"
 metric_list:
   - metric: acc
+metadata:
+  - version: 0.0

lm_eval/tasks/model_written_evals/persona/_template_yaml

@@ -8,3 +8,5 @@ doc_to_target: 0
 doc_to_choice: "{{[answer_matching_behavior, answer_not_matching_behavior]}}"
 metric_list:
   - metric: acc
+metadata:
+  - version: 0.0

lm_eval/tasks/model_written_evals/sycophancy/sycophancy_on_nlp_survey.yaml

@@ -10,3 +10,5 @@ doc_to_target: 0
 doc_to_choice: "{{[answer_matching_behavior, answer_not_matching_behavior]}}"
 metric_list:
   - metric: acc
+metadata:
+  - version: 0.0

lm_eval/tasks/model_written_evals/sycophancy/sycophancy_on_philpapers2020.yaml

@@ -10,3 +10,5 @@ doc_to_target: 0
 doc_to_choice: "{{[answer_matching_behavior, answer_not_matching_behavior]}}"
 metric_list:
   - metric: acc
+metadata:
+  - version: 0.0

lm_eval/tasks/model_written_evals/sycophancy/sycophancy_on_political_typology_quiz.yaml

@@ -10,3 +10,5 @@ doc_to_target: 0
 doc_to_choice: "{{[answer_matching_behavior, answer_not_matching_behavior]}}"
 metric_list:
   - metric: acc
+metadata:
+  - version: 0.0