Commit 7d09b24c authored by haileyschoelkopf's avatar haileyschoelkopf
Browse files

fix alllll the merge conflicts

parents 96dfe976 6348b947
group:
tag:
- math_word_problems
task: gsm8k
dataset_path: gsm8k
......
group: haerae
dataset_path: HAERAE-HUB/HAE_RAE_BENCH
test_split: test
fewshot_split: test
......
group: haerae
task:
- haerae_gk
- haerae_hi
- haerae_lw
- haerae_rw
- haerae_sn
aggregate_metric_list:
- metric: acc
aggregation: mean
weight_by_size: true
- metric: acc_norm
aggregation: mean
weight_by_size: true
metadata:
version: 1.0
group:
- headqa
tag: headqa
task: headqa_en
dataset_path: EleutherAI/headqa
dataset_name: en
......
group:
tag:
- multiple_choice
task: hellaswag
dataset_path: hellaswag
......
group:
tag:
- hendrycks_ethics
task: ethics_cm
dataset_path: EleutherAI/hendrycks_ethics
......
include: deontology.yaml
group:
tag:
- hendrycks_ethics
task: ethics_justice
dataset_name: justice
......
include: commonsense.yaml
group:
tag:
- hendrycks_ethics
task: ethics_utilitarianism
dataset_name: utilitarianism
......
include: commonsense.yaml
group:
tag:
- hendrycks_ethics
task: ethics_virtue
dataset_name: virtue
......
......@@ -7,3 +7,9 @@ task:
- hendrycks_math_num_theory
- hendrycks_math_prealgebra
- hendrycks_math_precalc
aggregate_metric_list:
- metric: exact_match
aggregation: mean
weight_by_size: true
metadata:
version: 1.0
group:
tag:
- math_word_problems
task: hendrycks_math_algebra
dataset_path: EleutherAI/hendrycks_math
......
group:
tag:
- kmmlu
- kmmlu_hard_cot
dataset_path: HAERAE-HUB/KMMLU-HARD
......
group:
tag:
- kmmlu
- kmmlu_direct
dataset_path: HAERAE-HUB/KMMLU
......
group:
tag:
- kmmlu
- kmmlu_hard_direct
dataset_path: HAERAE-HUB/KMMLU-HARD
......
group:
tag:
- kmmlu
- kmmlu_hard
dataset_path: HAERAE-HUB/KMMLU-HARD
......
group:
tag:
- kobest
task: kobest_boolq
dataset_path: skt/kobest_v1
......
group:
tag:
- kobest
task: kobest_copa
dataset_path: skt/kobest_v1
......
group:
tag:
- kobest
task: kobest_hellaswag
dataset_path: skt/kobest_v1
......
group:
tag:
- kobest
task: kobest_sentineg
dataset_path: skt/kobest_v1
......
group:
tag:
- kobest
task: kobest_wic
dataset_path: skt/kobest_v1
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment