Skip to content
Open
Show file tree
Hide file tree
Changes from 5 commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
233 changes: 233 additions & 0 deletions build_open_spec.py

Large diffs are not rendered by default.

32 changes: 16 additions & 16 deletions build_secret_run_spec.py
Original file line number Diff line number Diff line change
Expand Up @@ -12,13 +12,13 @@
{'scenario': 'ethics', 'description': "ethics_utilitarianism:model=neurips/local,data_augmentation=canonical", 'priority': 1},

## Math datasets
{'scenario': 'math', 'description': "math:model=neurips/local,subject=number_theory,level=1,use_official_examples=True", 'priority': 2},
{'scenario': 'math', 'description': "math:model=neurips/local,subject=intermediate_algebra,level=1,use_official_examples=True", 'priority': 2},
{'scenario': 'math', 'description': "math:model=neurips/local,subject=algebra,level=1,use_official_examples=True", 'priority': 2},
{'scenario': 'math', 'description': "math:model=neurips/local,subject=prealgebra,level=1,use_official_examples=True", 'priority': 2},
{'scenario': 'math', 'description': "math:model=neurips/local,subject=geometry,level=1,use_official_examples=True", 'priority': 2},
{'scenario': 'math', 'description': "math:model=neurips/local,subject=counting_and_probability,level=1,use_official_examples=True", 'priority': 2},
{'scenario': 'math', 'description': "math:model=neurips/local,subject=precalculus,level=1,use_official_examples=True", 'priority': 2},
# {'scenario': 'math', 'description': "math:model=neurips/local,subject=number_theory,level=1,use_official_examples=True", 'priority': 2},
# {'scenario': 'math', 'description': "math:model=neurips/local,subject=intermediate_algebra,level=1,use_official_examples=True", 'priority': 2},
# {'scenario': 'math', 'description': "math:model=neurips/local,subject=algebra,level=1,use_official_examples=True", 'priority': 2},
# {'scenario': 'math', 'description': "math:model=neurips/local,subject=prealgebra,level=1,use_official_examples=True", 'priority': 2},
# {'scenario': 'math', 'description': "math:model=neurips/local,subject=geometry,level=1,use_official_examples=True", 'priority': 2},
# {'scenario': 'math', 'description': "math:model=neurips/local,subject=counting_and_probability,level=1,use_official_examples=True", 'priority': 2},
# {'scenario': 'math', 'description': "math:model=neurips/local,subject=precalculus,level=1,use_official_examples=True", 'priority': 2},
#
# {'scenario': 'math', 'description': "math:model=neurips/local,subject=number_theory,level=2,use_official_examples=True", 'priority': 4},
# {'scenario': 'math', 'description': "math:model=neurips/local,subject=intermediate_algebra,level=2,use_official_examples=True", 'priority': 4},
Expand All @@ -44,13 +44,13 @@
# {'scenario': 'math', 'description': "math:model=neurips/local,subject=counting_and_probability,level=4,use_official_examples=True", 'priority': 4},
# {'scenario': 'math', 'description': "math:model=neurips/local,subject=precalculus,level=4,use_official_examples=True", 'priority': 4},

{'scenario': 'math', 'description': "math:model=neurips/local,subject=number_theory,level=5,use_official_examples=True", 'priority': 2},
{'scenario': 'math', 'description': "math:model=neurips/local,subject=intermediate_algebra,level=5,use_official_examples=True", 'priority': 2},
{'scenario': 'math', 'description': "math:model=neurips/local,subject=algebra,level=5,use_official_examples=True", 'priority': 2},
{'scenario': 'math', 'description': "math:model=neurips/local,subject=prealgebra,level=5,use_official_examples=True", 'priority': 2},
{'scenario': 'math', 'description': "math:model=neurips/local,subject=geometry,level=5,use_official_examples=True", 'priority': 2},
{'scenario': 'math', 'description': "math:model=neurips/local,subject=counting_and_probability,level=5,use_official_examples=True", 'priority': 2},
{'scenario': 'math', 'description': "math:model=neurips/local,subject=precalculus,level=5,use_official_examples=True", 'priority': 2},
# {'scenario': 'math', 'description': "math:model=neurips/local,subject=number_theory,level=5,use_official_examples=True", 'priority': 2},
# {'scenario': 'math', 'description': "math:model=neurips/local,subject=intermediate_algebra,level=5,use_official_examples=True", 'priority': 2},
# {'scenario': 'math', 'description': "math:model=neurips/local,subject=algebra,level=5,use_official_examples=True", 'priority': 2},
# {'scenario': 'math', 'description': "math:model=neurips/local,subject=prealgebra,level=5,use_official_examples=True", 'priority': 2},
# {'scenario': 'math', 'description': "math:model=neurips/local,subject=geometry,level=5,use_official_examples=True", 'priority': 2},
# {'scenario': 'math', 'description': "math:model=neurips/local,subject=counting_and_probability,level=5,use_official_examples=True", 'priority': 2},
# {'scenario': 'math', 'description': "math:model=neurips/local,subject=precalculus,level=5,use_official_examples=True", 'priority': 2},

# With chain-of-thought prompting:
{'scenario': 'math', 'description': "math:model=neurips/local,subject=number_theory,level=1,use_chain_of_thought=True", 'priority': 2},
Expand Down Expand Up @@ -133,7 +133,7 @@ def generate_equal_sum_list(V, N):
v = scenario_n_examples_dict[cur_scenario].pop()
entries[i]['max_eval_instances'] = v

with open(f'./run_specs_full_closed_eval_coarse_{args.example_budget}_budget.conf','w') as f:
with open(f'./run_specs_closed_{args.example_budget}_budget.conf','w') as f:
f.write('entries: [\n')
last_scenario = ''
for entry in entries:
Expand All @@ -148,4 +148,4 @@ def generate_equal_sum_list(V, N):
f.write('}\n')
f.write(']')

print(f'Saved ./run_secret_specs_full_coarse_{args.example_budget}_budget.conf')
print(f'Saved ./run_secret_specs_closed_{args.example_budget}_budget.conf')
Original file line number Diff line number Diff line change
Expand Up @@ -14,20 +14,6 @@ entries: [
{description: "ethics_utilitarianism:model=neurips/local,data_augmentation=canonical,max_eval_instances=4",priority: 1}

# math
{description: "math:model=neurips/local,subject=number_theory,level=1,use_official_examples=True,max_eval_instances=0",priority: 1}
{description: "math:model=neurips/local,subject=intermediate_algebra,level=1,use_official_examples=True,max_eval_instances=0",priority: 1}
{description: "math:model=neurips/local,subject=algebra,level=1,use_official_examples=True,max_eval_instances=0",priority: 1}
{description: "math:model=neurips/local,subject=prealgebra,level=1,use_official_examples=True,max_eval_instances=0",priority: 1}
{description: "math:model=neurips/local,subject=geometry,level=1,use_official_examples=True,max_eval_instances=0",priority: 1}
{description: "math:model=neurips/local,subject=counting_and_probability,level=1,use_official_examples=True,max_eval_instances=0",priority: 1}
{description: "math:model=neurips/local,subject=precalculus,level=1,use_official_examples=True,max_eval_instances=0",priority: 1}
{description: "math:model=neurips/local,subject=number_theory,level=5,use_official_examples=True,max_eval_instances=0",priority: 1}
{description: "math:model=neurips/local,subject=intermediate_algebra,level=5,use_official_examples=True,max_eval_instances=1",priority: 1}
{description: "math:model=neurips/local,subject=algebra,level=5,use_official_examples=True,max_eval_instances=1",priority: 1}
{description: "math:model=neurips/local,subject=prealgebra,level=5,use_official_examples=True,max_eval_instances=1",priority: 1}
{description: "math:model=neurips/local,subject=geometry,level=5,use_official_examples=True,max_eval_instances=1",priority: 1}
{description: "math:model=neurips/local,subject=counting_and_probability,level=5,use_official_examples=True,max_eval_instances=1",priority: 1}
{description: "math:model=neurips/local,subject=precalculus,level=5,use_official_examples=True,max_eval_instances=1",priority: 1}
{description: "math:model=neurips/local,subject=number_theory,level=1,use_chain_of_thought=True,max_eval_instances=1",priority: 1}
{description: "math:model=neurips/local,subject=intermediate_algebra,level=1,use_chain_of_thought=True,max_eval_instances=1",priority: 1}
{description: "math:model=neurips/local,subject=algebra,level=1,use_chain_of_thought=True,max_eval_instances=1",priority: 1}
Expand All @@ -36,12 +22,12 @@ entries: [
{description: "math:model=neurips/local,subject=counting_and_probability,level=1,use_chain_of_thought=True,max_eval_instances=1",priority: 1}
{description: "math:model=neurips/local,subject=precalculus,level=1,use_chain_of_thought=True,max_eval_instances=1",priority: 1}
{description: "math:model=neurips/local,subject=number_theory,level=5,use_chain_of_thought=True,max_eval_instances=1",priority: 1}
{description: "math:model=neurips/local,subject=intermediate_algebra,level=5,use_chain_of_thought=True,max_eval_instances=1",priority: 1}
{description: "math:model=neurips/local,subject=algebra,level=5,use_chain_of_thought=True,max_eval_instances=1",priority: 1}
{description: "math:model=neurips/local,subject=prealgebra,level=5,use_chain_of_thought=True,max_eval_instances=1",priority: 1}
{description: "math:model=neurips/local,subject=geometry,level=5,use_chain_of_thought=True,max_eval_instances=1",priority: 1}
{description: "math:model=neurips/local,subject=counting_and_probability,level=5,use_chain_of_thought=True,max_eval_instances=1",priority: 1}
{description: "math:model=neurips/local,subject=precalculus,level=5,use_chain_of_thought=True,max_eval_instances=1",priority: 1}
{description: "math:model=neurips/local,subject=intermediate_algebra,level=5,use_chain_of_thought=True,max_eval_instances=2",priority: 1}
{description: "math:model=neurips/local,subject=algebra,level=5,use_chain_of_thought=True,max_eval_instances=2",priority: 1}
{description: "math:model=neurips/local,subject=prealgebra,level=5,use_chain_of_thought=True,max_eval_instances=2",priority: 1}
{description: "math:model=neurips/local,subject=geometry,level=5,use_chain_of_thought=True,max_eval_instances=2",priority: 1}
{description: "math:model=neurips/local,subject=counting_and_probability,level=5,use_chain_of_thought=True,max_eval_instances=2",priority: 1}
{description: "math:model=neurips/local,subject=precalculus,level=5,use_chain_of_thought=True,max_eval_instances=2",priority: 1}

# cnn
{description: "summarization_cnndm:model=neurips/local,max_eval_instances=20",priority: 1}
Expand Down
34 changes: 34 additions & 0 deletions configs/run_specs_closed_5000_budget.conf
Original file line number Diff line number Diff line change
@@ -0,0 +1,34 @@
entries: [

# summarization
{description: "sam_sum:model=neurips/local,max_eval_instances=1000",priority: 1}

# causation
{description: "corr2cause:model=neurips/local,max_train_instances=1,max_eval_instances=1000",priority: 1}

# ethics
{description: "ethics_justice:model=neurips/local,data_augmentation=canonical,max_eval_instances=200",priority: 1}
{description: "ethics_commonsense:model=neurips/local,data_augmentation=canonical,max_eval_instances=200",priority: 1}
{description: "ethics_virtue:model=neurips/local,data_augmentation=canonical,max_eval_instances=200",priority: 1}
{description: "ethics_deontology:model=neurips/local,data_augmentation=canonical,max_eval_instances=200",priority: 1}
{description: "ethics_utilitarianism:model=neurips/local,data_augmentation=canonical,max_eval_instances=200",priority: 1}

# math
{description: "math:model=neurips/local,subject=number_theory,level=1,use_chain_of_thought=True,max_eval_instances=71",priority: 1}
{description: "math:model=neurips/local,subject=intermediate_algebra,level=1,use_chain_of_thought=True,max_eval_instances=71",priority: 1}
{description: "math:model=neurips/local,subject=algebra,level=1,use_chain_of_thought=True,max_eval_instances=71",priority: 1}
{description: "math:model=neurips/local,subject=prealgebra,level=1,use_chain_of_thought=True,max_eval_instances=71",priority: 1}
{description: "math:model=neurips/local,subject=geometry,level=1,use_chain_of_thought=True,max_eval_instances=71",priority: 1}
{description: "math:model=neurips/local,subject=counting_and_probability,level=1,use_chain_of_thought=True,max_eval_instances=71",priority: 1}
{description: "math:model=neurips/local,subject=precalculus,level=1,use_chain_of_thought=True,max_eval_instances=71",priority: 1}
{description: "math:model=neurips/local,subject=number_theory,level=5,use_chain_of_thought=True,max_eval_instances=71",priority: 1}
{description: "math:model=neurips/local,subject=intermediate_algebra,level=5,use_chain_of_thought=True,max_eval_instances=72",priority: 1}
{description: "math:model=neurips/local,subject=algebra,level=5,use_chain_of_thought=True,max_eval_instances=72",priority: 1}
{description: "math:model=neurips/local,subject=prealgebra,level=5,use_chain_of_thought=True,max_eval_instances=72",priority: 1}
{description: "math:model=neurips/local,subject=geometry,level=5,use_chain_of_thought=True,max_eval_instances=72",priority: 1}
{description: "math:model=neurips/local,subject=counting_and_probability,level=5,use_chain_of_thought=True,max_eval_instances=72",priority: 1}
{description: "math:model=neurips/local,subject=precalculus,level=5,use_chain_of_thought=True,max_eval_instances=72",priority: 1}

# cnn
{description: "summarization_cnndm:model=neurips/local,max_eval_instances=1000",priority: 1}
]
Loading