Merge branch 'upstream' into 'mmlu-pro'

add tokenizer logs info (#1731)

See merge request shijie.yu/lm-evaluation-harness!4

lm_eval/tasks/arc_mt/README.md

+# arc mt
+
+arc mt is an implementation of tasks to support machine translated arc
+challenge evals, to improve eval support across a number of additional
+languages.
+
+The main page for the effort is
+[here](https://huggingface.co/datasets/LumiOpen/arc_challenge_mt) and we will
+include more data and analysis there.
+
+Initial datasets include a number of European languages, and we plan to expand
+more in the future.

lm_eval/tasks/arc_mt/arc_challenge_mt_da.yaml

+include: arc_challenge_mt_fi.yaml
+task: arc_challenge_mt_da
+dataset_name: da

lm_eval/tasks/arc_mt/arc_challenge_mt_de.yaml

+include: arc_challenge_mt_fi.yaml
+task: arc_challenge_mt_de
+dataset_name: de

lm_eval/tasks/arc_mt/arc_challenge_mt_el.yaml

+include: arc_challenge_mt_fi.yaml
+task: arc_challenge_mt_el
+dataset_name: el

lm_eval/tasks/arc_mt/arc_challenge_mt_es.yaml

+include: arc_challenge_mt_fi.yaml
+task: arc_challenge_mt_es
+dataset_name: es

lm_eval/tasks/arc_mt/arc_challenge_mt_fi.yaml

+tag:
+  - arc_challenge_mt
+task: arc_challenge_mt_fi
+dataset_path: LumiOpen/arc_challenge_mt
+dataset_name: fi
+output_type: multiple_choice
+training_split: train
+validation_split: validation
+test_split: test
+doc_to_text: "Question: {{question}}\nAnswer:"
+doc_to_target: "{{choices.label.index(answerKey)}}"
+doc_to_choice: "{{choices.text}}"
+should_decontaminate: true
+doc_to_decontamination_query: "Question: {{question}}\nAnswer:"
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+metadata:
+  version: 1.0

lm_eval/tasks/arc_mt/arc_challenge_mt_hu.yaml

+include: arc_challenge_mt_fi.yaml
+task: arc_challenge_mt_hu
+dataset_name: hu

lm_eval/tasks/arc_mt/arc_challenge_mt_is.yaml

+group:
+  - arc_challenge_mt
+task: arc_challenge_mt_is
+dataset_path: mideind/icelandic-arc-challenge
+output_type: multiple_choice
+training_split: train
+validation_split: validation
+test_split: test
+doc_to_text: "Question: {{question}}\nAnswer:"
+doc_to_target: "{{choices.label.index(answerKey)}}"
+doc_to_choice: "{{choices.text}}"
+should_decontaminate: true
+doc_to_decontamination_query: "Question: {{question}}\nAnswer:"
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+metadata:
+  version: 1.0

lm_eval/tasks/arc_mt/arc_challenge_mt_it.yaml

+include: arc_challenge_mt_fi.yaml
+task: arc_challenge_mt_it
+dataset_name: it

lm_eval/tasks/arc_mt/arc_challenge_mt_nb.yaml

+include: arc_challenge_mt_fi.yaml
+task: arc_challenge_mt_nb
+dataset_name: nb

lm_eval/tasks/arc_mt/arc_challenge_mt_pl.yaml

+include: arc_challenge_mt_fi.yaml
+task: arc_challenge_mt_pl
+dataset_name: pl

lm_eval/tasks/arc_mt/arc_challenge_mt_pt.yaml

+include: arc_challenge_mt_fi.yaml
+task: arc_challenge_mt_pt
+dataset_name: pt

lm_eval/tasks/arc_mt/arc_challenge_mt_sv.yaml

+include: arc_challenge_mt_fi.yaml
+task: arc_challenge_mt_sv
+dataset_name: sv

lm_eval/tasks/arithmetic/README.md

@@ -27,9 +27,9 @@ Homepage: https://github.com/openai/gpt-3/tree/master/data
 }
 ```
 
-### Groups and Tasks
+### Groups, Tags, and Tasks
 
-#### Groups
+#### Tags
 
 * `arithmetic`: Evaluates `1dc` to `5ds`
 

lm_eval/tasks/arithmetic/arithmetic_1dc.yaml

-group:
+tag:
   - arithmetic
 task: arithmetic_1dc
 dataset_path: EleutherAI/arithmetic

lm_eval/tasks/asdiv/README.md

@@ -32,7 +32,7 @@ Homepage: https://github.com/chaochun/nlu-asdiv-dataset
 }
 ```
 
-### Groups and Tasks
+### Groups, Tags, and Tasks
 
 #### Groups
 

lm_eval/tasks/babi/README.md

@@ -21,12 +21,16 @@ Homepage: https://github.com/facebookarchive/bAbI-tasks
 }
 ```
 
-### Groups and Tasks
+### Groups, Tags, and Tasks
 
 #### Groups
 
 * Not part of a group yet
 
+#### Tags
+
+* No tags applied.
+
 #### Tasks
 
 * `babi`

lm_eval/tasks/basqueglue/README.md

@@ -43,11 +43,15 @@ Homepage: `https://github.com/hitz-zentroa/latxa`
 }
 ```
 
-### Groups and Tasks
+### Groups, Tags, and Tasks
 
 #### Groups
 
-* `basque-glue`: First version of the implementation
+None.
+
+#### Tags
+
+* `basque-glue`: First version of the implementation. Calls all subtasks, but does not average.
 
 #### Tasks
 

lm_eval/tasks/basqueglue/bec.yaml

-group: basque-glue
+tag: basque-glue
 task: bec2016eu
 dataset_path: orai-nlp/basqueGLUE
 dataset_name: bec

lm_eval/tasks/basqueglue/bhtc.yaml

-group: basque-glue
+tag: basque-glue
 task: bhtc_v2
 dataset_path: orai-nlp/basqueGLUE
 dataset_name: bhtc