Skip to content
GitLab
Menu
Projects
Groups
Snippets
Loading...
Help
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in / Register
Toggle navigation
Menu
Open sidebar
gaoqiong
lm-evaluation-harness
Commits
6dfe848e
Commit
6dfe848e
authored
Nov 02, 2023
by
lintangsutawika
Browse files
added group and task alias
parent
d1e7a30a
Changes
59
Hide whitespace changes
Inline
Side-by-side
Showing
19 changed files
with
73 additions
and
19 deletions
+73
-19
lm_eval/tasks/mmlu/default/mmlu_marketing.yaml
lm_eval/tasks/mmlu/default/mmlu_marketing.yaml
+4
-1
lm_eval/tasks/mmlu/default/mmlu_medical_genetics.yaml
lm_eval/tasks/mmlu/default/mmlu_medical_genetics.yaml
+4
-1
lm_eval/tasks/mmlu/default/mmlu_miscellaneous.yaml
lm_eval/tasks/mmlu/default/mmlu_miscellaneous.yaml
+4
-1
lm_eval/tasks/mmlu/default/mmlu_moral_disputes.yaml
lm_eval/tasks/mmlu/default/mmlu_moral_disputes.yaml
+4
-1
lm_eval/tasks/mmlu/default/mmlu_moral_scenarios.yaml
lm_eval/tasks/mmlu/default/mmlu_moral_scenarios.yaml
+4
-1
lm_eval/tasks/mmlu/default/mmlu_nutrition.yaml
lm_eval/tasks/mmlu/default/mmlu_nutrition.yaml
+4
-1
lm_eval/tasks/mmlu/default/mmlu_philosophy.yaml
lm_eval/tasks/mmlu/default/mmlu_philosophy.yaml
+4
-1
lm_eval/tasks/mmlu/default/mmlu_prehistory.yaml
lm_eval/tasks/mmlu/default/mmlu_prehistory.yaml
+4
-1
lm_eval/tasks/mmlu/default/mmlu_professional_accounting.yaml
lm_eval/tasks/mmlu/default/mmlu_professional_accounting.yaml
+4
-1
lm_eval/tasks/mmlu/default/mmlu_professional_law.yaml
lm_eval/tasks/mmlu/default/mmlu_professional_law.yaml
+4
-1
lm_eval/tasks/mmlu/default/mmlu_professional_medicine.yaml
lm_eval/tasks/mmlu/default/mmlu_professional_medicine.yaml
+4
-1
lm_eval/tasks/mmlu/default/mmlu_professional_psychology.yaml
lm_eval/tasks/mmlu/default/mmlu_professional_psychology.yaml
+4
-1
lm_eval/tasks/mmlu/default/mmlu_public_relations.yaml
lm_eval/tasks/mmlu/default/mmlu_public_relations.yaml
+4
-1
lm_eval/tasks/mmlu/default/mmlu_security_studies.yaml
lm_eval/tasks/mmlu/default/mmlu_security_studies.yaml
+4
-1
lm_eval/tasks/mmlu/default/mmlu_sociology.yaml
lm_eval/tasks/mmlu/default/mmlu_sociology.yaml
+4
-1
lm_eval/tasks/mmlu/default/mmlu_us_foreign_policy.yaml
lm_eval/tasks/mmlu/default/mmlu_us_foreign_policy.yaml
+4
-1
lm_eval/tasks/mmlu/default/mmlu_virology.yaml
lm_eval/tasks/mmlu/default/mmlu_virology.yaml
+4
-1
lm_eval/tasks/mmlu/default/mmlu_world_religions.yaml
lm_eval/tasks/mmlu/default/mmlu_world_religions.yaml
+4
-1
lm_eval/tasks/mmlu/flan_n_shot/loglikelihood/_mmlu_flan_loglikelihood_template_yaml
...shot/loglikelihood/_mmlu_flan_loglikelihood_template_yaml
+1
-1
No files found.
lm_eval/tasks/mmlu/default/mmlu_marketing.yaml
View file @
6dfe848e
"
dataset_name"
:
"
marketing"
"
dataset_name"
:
"
marketing"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
marketing.
\n\n
"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
marketing.
\n\
\n
"
"
group"
:
"
mmlu_other"
"
group"
:
"
mmlu_other"
"
group_alias"
:
"
other"
"
include"
:
"
_default_template_yaml"
"
include"
:
"
_default_template_yaml"
"
task"
:
"
mmlu_marketing"
"
task"
:
"
mmlu_marketing"
"
task_alias"
:
"
marketing"
lm_eval/tasks/mmlu/default/mmlu_medical_genetics.yaml
View file @
6dfe848e
"
dataset_name"
:
"
medical_genetics"
"
dataset_name"
:
"
medical_genetics"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
medical
genetics.
\n\n
"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
medical
\
\
genetics.
\n\n
"
"
group"
:
"
mmlu_other"
"
group"
:
"
mmlu_other"
"
group_alias"
:
"
other"
"
include"
:
"
_default_template_yaml"
"
include"
:
"
_default_template_yaml"
"
task"
:
"
mmlu_medical_genetics"
"
task"
:
"
mmlu_medical_genetics"
"
task_alias"
:
"
medical_genetics"
lm_eval/tasks/mmlu/default/mmlu_miscellaneous.yaml
View file @
6dfe848e
"
dataset_name"
:
"
miscellaneous"
"
dataset_name"
:
"
miscellaneous"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
miscellaneous.
\n\n
"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
miscellaneous.
\n\
\n
"
"
group"
:
"
mmlu_other"
"
group"
:
"
mmlu_other"
"
group_alias"
:
"
other"
"
include"
:
"
_default_template_yaml"
"
include"
:
"
_default_template_yaml"
"
task"
:
"
mmlu_miscellaneous"
"
task"
:
"
mmlu_miscellaneous"
"
task_alias"
:
"
miscellaneous"
lm_eval/tasks/mmlu/default/mmlu_moral_disputes.yaml
View file @
6dfe848e
"
dataset_name"
:
"
moral_disputes"
"
dataset_name"
:
"
moral_disputes"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
moral
disputes.
\n\n
"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
moral
\
\
disputes.
\n\n
"
"
group"
:
"
mmlu_humanities"
"
group"
:
"
mmlu_humanities"
"
group_alias"
:
"
humanities"
"
include"
:
"
_default_template_yaml"
"
include"
:
"
_default_template_yaml"
"
task"
:
"
mmlu_moral_disputes"
"
task"
:
"
mmlu_moral_disputes"
"
task_alias"
:
"
moral_disputes"
lm_eval/tasks/mmlu/default/mmlu_moral_scenarios.yaml
View file @
6dfe848e
"
dataset_name"
:
"
moral_scenarios"
"
dataset_name"
:
"
moral_scenarios"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
moral
scenarios.
\n\n
"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
moral
\
\
scenarios.
\n\n
"
"
group"
:
"
mmlu_humanities"
"
group"
:
"
mmlu_humanities"
"
group_alias"
:
"
humanities"
"
include"
:
"
_default_template_yaml"
"
include"
:
"
_default_template_yaml"
"
task"
:
"
mmlu_moral_scenarios"
"
task"
:
"
mmlu_moral_scenarios"
"
task_alias"
:
"
moral_scenarios"
lm_eval/tasks/mmlu/default/mmlu_nutrition.yaml
View file @
6dfe848e
"
dataset_name"
:
"
nutrition"
"
dataset_name"
:
"
nutrition"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
nutrition.
\n\n
"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
nutrition.
\n\
\n
"
"
group"
:
"
mmlu_other"
"
group"
:
"
mmlu_other"
"
group_alias"
:
"
other"
"
include"
:
"
_default_template_yaml"
"
include"
:
"
_default_template_yaml"
"
task"
:
"
mmlu_nutrition"
"
task"
:
"
mmlu_nutrition"
"
task_alias"
:
"
nutrition"
lm_eval/tasks/mmlu/default/mmlu_philosophy.yaml
View file @
6dfe848e
"
dataset_name"
:
"
philosophy"
"
dataset_name"
:
"
philosophy"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
philosophy.
\n\n
"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
philosophy.
\n\
\n
"
"
group"
:
"
mmlu_humanities"
"
group"
:
"
mmlu_humanities"
"
group_alias"
:
"
humanities"
"
include"
:
"
_default_template_yaml"
"
include"
:
"
_default_template_yaml"
"
task"
:
"
mmlu_philosophy"
"
task"
:
"
mmlu_philosophy"
"
task_alias"
:
"
philosophy"
lm_eval/tasks/mmlu/default/mmlu_prehistory.yaml
View file @
6dfe848e
"
dataset_name"
:
"
prehistory"
"
dataset_name"
:
"
prehistory"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
prehistory.
\n\n
"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
prehistory.
\n\
\n
"
"
group"
:
"
mmlu_humanities"
"
group"
:
"
mmlu_humanities"
"
group_alias"
:
"
humanities"
"
include"
:
"
_default_template_yaml"
"
include"
:
"
_default_template_yaml"
"
task"
:
"
mmlu_prehistory"
"
task"
:
"
mmlu_prehistory"
"
task_alias"
:
"
prehistory"
lm_eval/tasks/mmlu/default/mmlu_professional_accounting.yaml
View file @
6dfe848e
"
dataset_name"
:
"
professional_accounting"
"
dataset_name"
:
"
professional_accounting"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
professional
accounting.
\n\n
"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
professional
\
\
accounting.
\n\n
"
"
group"
:
"
mmlu_other"
"
group"
:
"
mmlu_other"
"
group_alias"
:
"
other"
"
include"
:
"
_default_template_yaml"
"
include"
:
"
_default_template_yaml"
"
task"
:
"
mmlu_professional_accounting"
"
task"
:
"
mmlu_professional_accounting"
"
task_alias"
:
"
professional_accounting"
lm_eval/tasks/mmlu/default/mmlu_professional_law.yaml
View file @
6dfe848e
"
dataset_name"
:
"
professional_law"
"
dataset_name"
:
"
professional_law"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
professional
law.
\n\n
"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
professional
\
\
law.
\n\n
"
"
group"
:
"
mmlu_humanities"
"
group"
:
"
mmlu_humanities"
"
group_alias"
:
"
humanities"
"
include"
:
"
_default_template_yaml"
"
include"
:
"
_default_template_yaml"
"
task"
:
"
mmlu_professional_law"
"
task"
:
"
mmlu_professional_law"
"
task_alias"
:
"
professional_law"
lm_eval/tasks/mmlu/default/mmlu_professional_medicine.yaml
View file @
6dfe848e
"
dataset_name"
:
"
professional_medicine"
"
dataset_name"
:
"
professional_medicine"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
professional
medicine.
\n\n
"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
professional
\
\
medicine.
\n\n
"
"
group"
:
"
mmlu_other"
"
group"
:
"
mmlu_other"
"
group_alias"
:
"
other"
"
include"
:
"
_default_template_yaml"
"
include"
:
"
_default_template_yaml"
"
task"
:
"
mmlu_professional_medicine"
"
task"
:
"
mmlu_professional_medicine"
"
task_alias"
:
"
professional_medicine"
lm_eval/tasks/mmlu/default/mmlu_professional_psychology.yaml
View file @
6dfe848e
"
dataset_name"
:
"
professional_psychology"
"
dataset_name"
:
"
professional_psychology"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
professional
psychology.
\n\n
"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
professional
\
\
psychology.
\n\n
"
"
group"
:
"
mmlu_social_sciences"
"
group"
:
"
mmlu_social_sciences"
"
group_alias"
:
"
social_sciences"
"
include"
:
"
_default_template_yaml"
"
include"
:
"
_default_template_yaml"
"
task"
:
"
mmlu_professional_psychology"
"
task"
:
"
mmlu_professional_psychology"
"
task_alias"
:
"
professional_psychology"
lm_eval/tasks/mmlu/default/mmlu_public_relations.yaml
View file @
6dfe848e
"
dataset_name"
:
"
public_relations"
"
dataset_name"
:
"
public_relations"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
public
relations.
\n\n
"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
public
\
\
relations.
\n\n
"
"
group"
:
"
mmlu_social_sciences"
"
group"
:
"
mmlu_social_sciences"
"
group_alias"
:
"
social_sciences"
"
include"
:
"
_default_template_yaml"
"
include"
:
"
_default_template_yaml"
"
task"
:
"
mmlu_public_relations"
"
task"
:
"
mmlu_public_relations"
"
task_alias"
:
"
public_relations"
lm_eval/tasks/mmlu/default/mmlu_security_studies.yaml
View file @
6dfe848e
"
dataset_name"
:
"
security_studies"
"
dataset_name"
:
"
security_studies"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
security
studies.
\n\n
"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
security
\
\
studies.
\n\n
"
"
group"
:
"
mmlu_social_sciences"
"
group"
:
"
mmlu_social_sciences"
"
group_alias"
:
"
social_sciences"
"
include"
:
"
_default_template_yaml"
"
include"
:
"
_default_template_yaml"
"
task"
:
"
mmlu_security_studies"
"
task"
:
"
mmlu_security_studies"
"
task_alias"
:
"
security_studies"
lm_eval/tasks/mmlu/default/mmlu_sociology.yaml
View file @
6dfe848e
"
dataset_name"
:
"
sociology"
"
dataset_name"
:
"
sociology"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
sociology.
\n\n
"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
sociology.
\n\
\n
"
"
group"
:
"
mmlu_social_sciences"
"
group"
:
"
mmlu_social_sciences"
"
group_alias"
:
"
social_sciences"
"
include"
:
"
_default_template_yaml"
"
include"
:
"
_default_template_yaml"
"
task"
:
"
mmlu_sociology"
"
task"
:
"
mmlu_sociology"
"
task_alias"
:
"
sociology"
lm_eval/tasks/mmlu/default/mmlu_us_foreign_policy.yaml
View file @
6dfe848e
"
dataset_name"
:
"
us_foreign_policy"
"
dataset_name"
:
"
us_foreign_policy"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
us
foreign
policy.
\n\n
"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
us
\
\
foreign
policy.
\n\n
"
"
group"
:
"
mmlu_social_sciences"
"
group"
:
"
mmlu_social_sciences"
"
group_alias"
:
"
social_sciences"
"
include"
:
"
_default_template_yaml"
"
include"
:
"
_default_template_yaml"
"
task"
:
"
mmlu_us_foreign_policy"
"
task"
:
"
mmlu_us_foreign_policy"
"
task_alias"
:
"
us_foreign_policy"
lm_eval/tasks/mmlu/default/mmlu_virology.yaml
View file @
6dfe848e
"
dataset_name"
:
"
virology"
"
dataset_name"
:
"
virology"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
virology.
\n\n
"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
virology.
\n\
\n
"
"
group"
:
"
mmlu_other"
"
group"
:
"
mmlu_other"
"
group_alias"
:
"
other"
"
include"
:
"
_default_template_yaml"
"
include"
:
"
_default_template_yaml"
"
task"
:
"
mmlu_virology"
"
task"
:
"
mmlu_virology"
"
task_alias"
:
"
virology"
lm_eval/tasks/mmlu/default/mmlu_world_religions.yaml
View file @
6dfe848e
"
dataset_name"
:
"
world_religions"
"
dataset_name"
:
"
world_religions"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
world
religions.
\n\n
"
"
description"
:
"
The
following
are
multiple
choice
questions
(with
answers)
about
world
\
\
religions.
\n\n
"
"
group"
:
"
mmlu_humanities"
"
group"
:
"
mmlu_humanities"
"
group_alias"
:
"
humanities"
"
include"
:
"
_default_template_yaml"
"
include"
:
"
_default_template_yaml"
"
task"
:
"
mmlu_world_religions"
"
task"
:
"
mmlu_world_religions"
"
task_alias"
:
"
world_religions"
lm_eval/tasks/mmlu/flan_n_shot/loglikelihood/_mmlu_flan_loglikelihood_template_yaml
View file @
6dfe848e
...
@@ -12,4 +12,4 @@ metric_list:
...
@@ -12,4 +12,4 @@ metric_list:
higher_is_better: true
higher_is_better: true
- metric: acc_norm
- metric: acc_norm
aggregation: mean
aggregation: mean
higher_is_better: true
higher_is_better: true
\ No newline at end of file
Prev
1
2
3
Next
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
.
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment