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
90d44580
Commit
90d44580
authored
Jul 22, 2025
by
Baber
Browse files
remove doc_to_choice from `boolq-seq2seq`
parent
3fd12675
Changes
17
Hide whitespace changes
Inline
Side-by-side
Showing
17 changed files
with
34 additions
and
34 deletions
+34
-34
lm_eval/tasks/super_glue/boolq/default.yaml
lm_eval/tasks/super_glue/boolq/default.yaml
+1
-1
lm_eval/tasks/super_glue/boolq/seq2seq.yaml
lm_eval/tasks/super_glue/boolq/seq2seq.yaml
+3
-3
lm_eval/tasks/super_glue/boolq/t5-prompt.yaml
lm_eval/tasks/super_glue/boolq/t5-prompt.yaml
+3
-3
lm_eval/tasks/super_glue/cb/default.yaml
lm_eval/tasks/super_glue/cb/default.yaml
+2
-2
lm_eval/tasks/super_glue/cb/t5-prompt.yaml
lm_eval/tasks/super_glue/cb/t5-prompt.yaml
+3
-3
lm_eval/tasks/super_glue/copa/default.yaml
lm_eval/tasks/super_glue/copa/default.yaml
+1
-1
lm_eval/tasks/super_glue/copa/t5-prompt.yaml
lm_eval/tasks/super_glue/copa/t5-prompt.yaml
+3
-3
lm_eval/tasks/super_glue/multirc/default.yaml
lm_eval/tasks/super_glue/multirc/default.yaml
+1
-1
lm_eval/tasks/super_glue/multirc/t5-prompt.yaml
lm_eval/tasks/super_glue/multirc/t5-prompt.yaml
+2
-2
lm_eval/tasks/super_glue/record/default.yaml
lm_eval/tasks/super_glue/record/default.yaml
+1
-1
lm_eval/tasks/super_glue/record/t5-prompt.yaml
lm_eval/tasks/super_glue/record/t5-prompt.yaml
+1
-1
lm_eval/tasks/super_glue/rte/default.yaml
lm_eval/tasks/super_glue/rte/default.yaml
+2
-2
lm_eval/tasks/super_glue/rte/t5-prompt.yaml
lm_eval/tasks/super_glue/rte/t5-prompt.yaml
+3
-3
lm_eval/tasks/super_glue/wic/default.yaml
lm_eval/tasks/super_glue/wic/default.yaml
+2
-2
lm_eval/tasks/super_glue/wic/t5-prompt.yaml
lm_eval/tasks/super_glue/wic/t5-prompt.yaml
+3
-3
lm_eval/tasks/super_glue/wsc/default.yaml
lm_eval/tasks/super_glue/wsc/default.yaml
+2
-2
lm_eval/tasks/super_glue/wsc/t5-prompt.yaml
lm_eval/tasks/super_glue/wsc/t5-prompt.yaml
+1
-1
No files found.
lm_eval/tasks/super_glue/boolq/default.yaml
View file @
90d44580
tag
:
-
super-glue-lm-eval-v1
task
:
boolq
dataset_path
:
super_glue
dataset_path
:
aps/
super_glue
dataset_name
:
boolq
output_type
:
multiple_choice
training_split
:
train
...
...
lm_eval/tasks/super_glue/boolq/seq2seq.yaml
View file @
90d44580
tag
:
-
super-glue-lm-eval-v1-seq2seq
task
:
"
boolq-seq2seq"
dataset_path
:
super_glue
dataset_path
:
aps/
super_glue
dataset_name
:
boolq
output_type
:
generate_until
training_split
:
train
validation_split
:
validation
doc_to_text
:
"
{{passage}}
\n
Question:
{{question}}?
\n
Answer:"
doc_to_target
:
label
doc_to_choice
:
[
'
no
'
,
'
yes
'
]
doc_to_target
:
"
{{
['
no',
'
yes'][label|int]
}}"
doc_to_choice
:
[
"
no
"
,
"
yes
"
]
target_delimiter
:
"
"
generation_kwargs
:
until
:
...
...
lm_eval/tasks/super_glue/boolq/t5-prompt.yaml
View file @
90d44580
tag
:
-
super-glue-t5-prompt
task
:
super_glue-boolq-t5-prompt
dataset_path
:
super_glue
dataset_path
:
aps/
super_glue
dataset_name
:
boolq
training_split
:
train
validation_split
:
validation
output_type
:
generate_until
doc_to_text
:
"
boolq
passage:
{{passage}}
question:
{{question}}"
doc_to_target
:
label
doc_to_choice
:
[
'
False
'
,
'
True
'
]
doc_to_target
:
"
{{['False',
'True'][label|int]}}"
doc_to_choice
:
[
"
False
"
,
"
True
"
]
generation_kwargs
:
until
:
-
"
</s>"
...
...
lm_eval/tasks/super_glue/cb/default.yaml
View file @
90d44580
tag
:
-
super-glue-lm-eval-v1
task
:
cb
dataset_path
:
super_glue
dataset_path
:
aps/
super_glue
dataset_name
:
cb
output_type
:
multiple_choice
training_split
:
train
validation_split
:
validation
doc_to_text
:
"
{{premise}}
\n
Question:
{{hypothesis}}.
True,
False,
or
Neither?
\n
Answer:"
doc_to_target
:
label
doc_to_choice
:
[
'
True
'
,
'
False
'
,
'
Neither
'
]
doc_to_choice
:
[
"
True
"
,
"
False
"
,
"
Neither
"
]
metric_list
:
-
metric
:
acc
-
metric
:
f1
...
...
lm_eval/tasks/super_glue/cb/t5-prompt.yaml
View file @
90d44580
tag
:
-
super-glue-t5-prompt
task
:
super_glue-cb-t5-prompt
dataset_path
:
super_glue
dataset_path
:
aps/
super_glue
dataset_name
:
cb
training_split
:
train
validation_split
:
validation
output_type
:
generate_until
doc_to_text
:
"
cb
hypothesis:
{{hypothesis}}
premise:
{{premise}}"
doc_to_target
:
label
doc_to_choice
:
[
'
entailment
'
,
'
contradiction
'
,
'
neutral
'
]
doc_to_target
:
"
{{
['entailment',
'contradiction',
'neutral'][label|int]
}}"
doc_to_choice
:
[
"
entailment
"
,
"
contradiction
"
,
"
neutral
"
]
generation_kwargs
:
until
:
-
"
</s>"
...
...
lm_eval/tasks/super_glue/copa/default.yaml
View file @
90d44580
tag
:
-
super-glue-lm-eval-v1
task
:
copa
dataset_path
:
super_glue
dataset_path
:
aps/
super_glue
dataset_name
:
copa
output_type
:
multiple_choice
training_split
:
train
...
...
lm_eval/tasks/super_glue/copa/t5-prompt.yaml
View file @
90d44580
tag
:
-
super-glue-t5-prompt
task
:
super_glue-copa-t5-prompt
dataset_path
:
super_glue
dataset_path
:
aps/
super_glue
dataset_name
:
copa
training_split
:
train
validation_split
:
validation
output_type
:
generate_until
doc_to_text
:
"
copa
choice1:
{{choice1}}
choice2:
{{choice2}}
premise:
{{premise}}
question:
{{question}}"
doc_to_target
:
label
doc_to_choice
:
[
'
choice1
'
,
'
choice2
'
]
doc_to_target
:
"
{{
[choice1,
choice2][label|int]
}}"
doc_to_choice
:
[
"
choice1
"
,
"
choice2
"
]
generation_kwargs
:
until
:
-
"
</s>"
...
...
lm_eval/tasks/super_glue/multirc/default.yaml
View file @
90d44580
tag
:
-
super-glue-lm-eval-v1
task
:
multirc
dataset_path
:
super_glue
dataset_path
:
aps/
super_glue
dataset_name
:
multirc
output_type
:
multiple_choice
training_split
:
train
...
...
lm_eval/tasks/super_glue/multirc/t5-prompt.yaml
View file @
90d44580
tag
:
-
super-glue-t5-prompt
task
:
super_glue-multirc-t5-prompt
dataset_path
:
super_glue
dataset_path
:
aps/
super_glue
dataset_name
:
multirc
training_split
:
train
validation_split
:
validation
output_type
:
generate_until
doc_to_text
:
"
multirc
question:
{{question}}
answer:
{{answer}}
paragraph:
{{paragraph}}"
doc_to_target
:
label
doc_to_target
:
"
{%
set
group_id
=
idx.question|string
%}{{[group_id+'_False',
group_id+'_True'][
label
]}}"
doc_to_choice
:
"
{%
set
group_id
=
idx.question|string
%}{{[group_id+'_False',
group_id+'_True']}}"
generation_kwargs
:
until
:
...
...
lm_eval/tasks/super_glue/record/default.yaml
View file @
90d44580
tag
:
-
super-glue-lm-eval-v1
task
:
record
dataset_path
:
super_glue
dataset_path
:
aps/
super_glue
dataset_name
:
record
output_type
:
multiple_choice
training_split
:
train
...
...
lm_eval/tasks/super_glue/record/t5-prompt.yaml
View file @
90d44580
tag
:
-
super-glue-t5-prompt
task
:
super_glue-record-t5-prompt
dataset_path
:
super_glue
dataset_path
:
aps/
super_glue
dataset_name
:
record
validation_split
:
validation
output_type
:
generate_until
...
...
lm_eval/tasks/super_glue/rte/default.yaml
View file @
90d44580
tag
:
-
super-glue-lm-eval-v1
task
:
sglue_rte
dataset_path
:
super_glue
dataset_path
:
aps/
super_glue
dataset_name
:
rte
output_type
:
multiple_choice
training_split
:
train
validation_split
:
validation
doc_to_text
:
"
{{premise}}
\n
Question:
{{hypothesis}}
True
or
False?
\n
Answer:"
doc_to_target
:
label
doc_to_choice
:
[
'
True
'
,
'
False
'
]
doc_to_choice
:
[
"
True
"
,
"
False
"
]
metric_list
:
-
metric
:
acc
metadata
:
...
...
lm_eval/tasks/super_glue/rte/t5-prompt.yaml
View file @
90d44580
tag
:
-
super-glue-t5-prompt
task
:
super_glue-rte-t5-prompt
dataset_path
:
super_glue
dataset_path
:
aps/
super_glue
dataset_name
:
rte
training_split
:
train
validation_split
:
validation
output_type
:
generate_until
doc_to_text
:
"
rte
hypothesis:
{{hypothesis}}
premise:
{{premise}}"
doc_to_target
:
label
doc_to_choice
:
[
'
entailment
'
,
'
not_entailment
'
]
doc_to_target
:
"
{{
[entailment,
not_entailment][label|int]
}}"
doc_to_choice
:
[
"
entailment
"
,
"
not_entailment
"
]
generation_kwargs
:
until
:
-
"
</s>"
...
...
lm_eval/tasks/super_glue/wic/default.yaml
View file @
90d44580
tag
:
-
super-glue-lm-eval-v1
task
:
"
wic"
dataset_path
:
super_glue
dataset_path
:
aps/
super_glue
dataset_name
:
wic
output_type
:
multiple_choice
training_split
:
train
validation_split
:
validation
doc_to_text
:
"
Sentence
1:
{{sentence1}}
\n
Sentence
2:
{{sentence2}}
\n
Question:
Is
the
word
'{{sentence1[start1:end1]}}'
used
in
the
same
way
in
the
two
sentences
above?
\n
Answer:"
doc_to_target
:
label
doc_to_choice
:
[
'
no
'
,
'
yes
'
]
doc_to_choice
:
[
"
no
"
,
"
yes
"
]
metric_list
:
-
metric
:
acc
metadata
:
...
...
lm_eval/tasks/super_glue/wic/t5-prompt.yaml
View file @
90d44580
tag
:
-
super-glue-t5-prompt
task
:
super_glue-wic-t5-prompt
dataset_path
:
super_glue
dataset_path
:
aps/
super_glue
dataset_name
:
wic
training_split
:
train
validation_split
:
validation
output_type
:
generate_until
doc_to_text
:
"
wic
sentence1:
{{sentence1}}
sentence2:
{{sentence2}}
word:
{{word}}"
doc_to_target
:
label
doc_to_choice
:
[
'
False
'
,
'
True
'
]
doc_to_target
:
"
{{
['False',
'True'][label|int]
}}"
doc_to_choice
:
[
"
False
"
,
"
True
"
]
generation_kwargs
:
until
:
-
"
</s>"
...
...
lm_eval/tasks/super_glue/wsc/default.yaml
View file @
90d44580
tag
:
-
super-glue-lm-eval-v1
task
:
wsc
dataset_path
:
super_glue
dataset_path
:
aps/
super_glue
dataset_name
:
wsc.fixed
output_type
:
multiple_choice
training_split
:
train
validation_split
:
validation
doc_to_text
:
!function
preprocess_wsc.default_doc_to_text
doc_to_target
:
label
doc_to_choice
:
[
'
no
'
,
'
yes
'
]
doc_to_choice
:
[
"
no
"
,
"
yes
"
]
metric_list
:
-
metric
:
acc
metadata
:
...
...
lm_eval/tasks/super_glue/wsc/t5-prompt.yaml
View file @
90d44580
tag
:
-
super-glue-t5-prompt
task
:
super_glue-wsc-t5-prompt
dataset_path
:
super_glue
dataset_path
:
aps/
super_glue
dataset_name
:
wsc.fixed
training_split
:
train
validation_split
:
validation
...
...
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