Merge branch 'master' of github.com:EleutherAI/lm-evaluation-harness

# Conflicts: # lm_eval/tasks/pile.py

Merge branch 'master' of github.com:EleutherAI/lm-evaluation-harness
# Conflicts: # lm_eval/tasks/pile.py
a67c17e0 · Leo Gao · cb3babd3 · 9d87d476 · a67c17e0 · a67c17e0
Commit a67c17e0 authored Jan 03, 2022 by Leo Gao
20 changed files
--- a/tests/tests/testdata/blimp_superlative_quantifiers_1-v0-loglikelihood
+++ b/tests/tests/testdata/blimp_superlative_quantifiers_1-v0-loglikelihood
--- a/tests/tests/testdata/blimp_superlative_quantifiers_1-v0-res.json
+++ b/tests/tests/testdata/blimp_superlative_quantifiers_1-v0-res.json
--- a/tests/tests/testdata/blimp_superlative_quantifiers_2-v0-loglikelihood
+++ b/tests/tests/testdata/blimp_superlative_quantifiers_2-v0-loglikelihood
--- a/tests/tests/testdata/blimp_superlative_quantifiers_2-v0-res.json
+++ b/tests/tests/testdata/blimp_superlative_quantifiers_2-v0-res.json
--- a/tests/tests/testdata/blimp_tough_vs_raising_1-v0-loglikelihood
+++ b/tests/tests/testdata/blimp_tough_vs_raising_1-v0-loglikelihood
--- a/tests/tests/testdata/blimp_tough_vs_raising_1-v0-res.json
+++ b/tests/tests/testdata/blimp_tough_vs_raising_1-v0-res.json
--- a/tests/tests/testdata/blimp_tough_vs_raising_2-v0-loglikelihood
+++ b/tests/tests/testdata/blimp_tough_vs_raising_2-v0-loglikelihood
--- a/tests/tests/testdata/blimp_tough_vs_raising_2-v0-res.json
+++ b/tests/tests/testdata/blimp_tough_vs_raising_2-v0-res.json
--- a/tests/tests/testdata/blimp_transitive-v0-loglikelihood
+++ b/tests/tests/testdata/blimp_transitive-v0-loglikelihood
--- a/tests/tests/testdata/blimp_transitive-v0-res.json
+++ b/tests/tests/testdata/blimp_transitive-v0-res.json
--- a/tests/tests/testdata/blimp_wh_island-v0-loglikelihood
+++ b/tests/tests/testdata/blimp_wh_island-v0-loglikelihood
--- a/tests/tests/testdata/blimp_wh_island-v0-res.json
+++ b/tests/tests/testdata/blimp_wh_island-v0-res.json
--- a/tests/tests/testdata/blimp_wh_questions_object_gap-v0-loglikelihood
+++ b/tests/tests/testdata/blimp_wh_questions_object_gap-v0-loglikelihood
--- a/tests/tests/testdata/blimp_wh_questions_object_gap-v0-res.json
+++ b/tests/tests/testdata/blimp_wh_questions_object_gap-v0-res.json
--- a/tests/tests/testdata/blimp_wh_questions_subject_gap-v0-loglikelihood
+++ b/tests/tests/testdata/blimp_wh_questions_subject_gap-v0-loglikelihood
--- a/tests/tests/testdata/blimp_wh_questions_subject_gap-v0-res.json
+++ b/tests/tests/testdata/blimp_wh_questions_subject_gap-v0-res.json
--- a/tests/tests/testdata/blimp_wh_questions_subject_gap_long_distance-v0-loglikelihood
+++ b/tests/tests/testdata/blimp_wh_questions_subject_gap_long_distance-v0-loglikelihood
--- a/tests/tests/testdata/blimp_wh_questions_subject_gap_long_distance-v0-res.json
+++ b/tests/tests/testdata/blimp_wh_questions_subject_gap_long_distance-v0-res.json
--- a/tests/tests/testdata/blimp_wh_vs_that_no_gap-v0-loglikelihood
+++ b/tests/tests/testdata/blimp_wh_vs_that_no_gap-v0-loglikelihood
--- a/tests/tests/testdata/blimp_wh_vs_that_no_gap-v0-res.json
+++ b/tests/tests/testdata/blimp_wh_vs_that_no_gap-v0-res.json