43 lines
1.0 KiB
Python
43 lines
1.0 KiB
Python
"""Utilities for execution-feedback code generation experiments."""
|
|
|
|
from .agent import AgentResult, AttemptRecord, ExecutionFeedbackAgent
|
|
from .backends import OpenAICompatibleGenerator
|
|
from .dataset import CodeRLDataset, collate_code_rl
|
|
from .reward import CodeReward, score_code_batch, score_verification
|
|
from .evaluation import (
|
|
EvaluationReport,
|
|
aggregate_reports,
|
|
estimate_pass_at_k,
|
|
evaluate_candidates,
|
|
)
|
|
from .verifier import (
|
|
CodeTask,
|
|
TestCase,
|
|
VerificationResult,
|
|
VerificationStatus,
|
|
extract_python_code,
|
|
verify_code,
|
|
)
|
|
|
|
__all__ = [
|
|
"AgentResult",
|
|
"AttemptRecord",
|
|
"CodeReward",
|
|
"CodeRLDataset",
|
|
"CodeTask",
|
|
"EvaluationReport",
|
|
"ExecutionFeedbackAgent",
|
|
"OpenAICompatibleGenerator",
|
|
"TestCase",
|
|
"VerificationResult",
|
|
"VerificationStatus",
|
|
"aggregate_reports",
|
|
"collate_code_rl",
|
|
"estimate_pass_at_k",
|
|
"evaluate_candidates",
|
|
"extract_python_code",
|
|
"score_verification",
|
|
"score_code_batch",
|
|
"verify_code",
|
|
]
|