Merge pull request #1031 from EleutherAI/versioning

[Refactor] Versioning

lm_eval/tasks/code_x_glue/code-text/php.yaml

@@ -17,3 +17,5 @@ metric_list:
   - metric: !function bleu.smoothed_bleu_4
     aggregation: mean
     higher_is_better: True
+metadata:
+  - version: 0.0

lm_eval/tasks/code_x_glue/code-text/python.yaml

@@ -17,3 +17,5 @@ metric_list:
   - metric: !function bleu.smoothed_bleu_4
     aggregation: mean
     higher_is_better: True
+metadata:
+  - version: 0.0

lm_eval/tasks/code_x_glue/code-text/ruby.yaml

@@ -17,3 +17,5 @@ metric_list:
   - metric: !function bleu.smoothed_bleu_4
     aggregation: mean
     higher_is_better: True
+metadata:
+  - version: 2.0

lm_eval/tasks/coqa/default.yaml

@@ -18,3 +18,5 @@ metric_list:
   - metric: f1
     aggregation: mean
     higher_is_better: true
+metadata:
+  - version: 2.0

lm_eval/tasks/crows_pairs/crows_pairs_english.yaml

@@ -19,3 +19,5 @@ metric_list:
   - metric: pct_stereotype
     aggregation: mean
     higher_is_better: false
+metadata:
+  - version: 1.0

lm_eval/tasks/csatqa/_default_csatqa_yaml

@@ -13,3 +13,5 @@ metric_list:
   - metric: acc_norm
     aggregation: mean
     higher_is_better: true
+metadata:
+  - version: 0.0

lm_eval/tasks/drop/default.yaml

@@ -20,3 +20,5 @@ metric_list:
   - metric: f1
     aggregation: mean
     higher_is_better: true
+metadata:
+  - version: 2.0

lm_eval/tasks/glue/cola/default.yaml

@@ -12,3 +12,5 @@ should_decontaminate: true
 doc_to_decontamination_query: sentence
 metric_list:
   - metric: mcc
+metadata:
+  - version: 1.0

lm_eval/tasks/glue/mnli/default.yaml

@@ -10,3 +10,5 @@ doc_to_target: label
 doc_to_choice: ["True", "Neither", "False"]
 metric_list:
   - metric: acc
+metadata:
+  - version: 1.0

lm_eval/tasks/glue/mrpc/default.yaml

@@ -11,3 +11,5 @@ doc_to_choice: ["no", "yes"]
 metric_list:
   - metric: acc
   - metric: f1
+metadata:
+  - version: 1.0

lm_eval/tasks/glue/qnli/default.yaml

@@ -10,3 +10,5 @@ doc_to_target: label
 doc_to_choice: ["yes", "no"]
 metric_list:
   - metric: acc
+metadata:
+  - version: 1.0

lm_eval/tasks/glue/qqp/default.yaml

@@ -12,3 +12,5 @@ doc_to_choice: ["no", "yes"]
 metric_list:
   - metric: acc
   - metric: f1
+metadata:
+  - version: 1.0

lm_eval/tasks/glue/rte/default.yaml

@@ -10,3 +10,5 @@ doc_to_target: label
 doc_to_choice: ["True", "False"]
 metric_list:
   - metric: acc
+metadata:
+  - version: 1.0

lm_eval/tasks/glue/sst/default.yaml

@@ -10,3 +10,5 @@ doc_to_target: label
 doc_to_choice: ["negative", "positive"]
 metric_list:
   - metric: acc
+metadata:
+  - version: 1.0

lm_eval/tasks/glue/wnli/default.yaml

@@ -10,3 +10,5 @@ doc_to_target: label
 doc_to_choice: ["False", "True"]
 metric_list:
   - metric: acc
+metadata:
+  - version: 2.0

lm_eval/tasks/gsm8k/gsm8k-cot-self-consistency.yaml

@@ -30,3 +30,5 @@ filter_list:
         regex_pattern: "The answer is (\\-?[0-9\\.\\,]*[0-9]+)"
       - function: "majority_vote"
       - function: "take_first"
+metadata:
+  - version: 0.0

lm_eval/tasks/gsm8k/gsm8k-cot.yaml

@@ -40,3 +40,5 @@ filter_list:
       - function: "regex"
         regex_pattern: "The answer is (\\-?[0-9\\.\\,]+)."
       - function: "take_first"
+metadata:
+  - version: 0.0

lm_eval/tasks/gsm8k/gsm8k.yaml

@@ -33,3 +33,5 @@ filter_list:
       - function: "regex"
         regex_pattern: "#### (\\-?[0-9\\.\\,]+)"
       - function: "take_first"
+metadata:
+  - version: 1.0

lm_eval/tasks/headqa/headqa_en.yaml

@@ -19,3 +19,5 @@ metric_list:
   - metric: acc_norm
     aggregation: mean
     higher_is_better: true
+metadata:
+  - version: 1.0

lm_eval/tasks/hellaswag/hellaswag.yaml

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