LaunchLLM / evaluation /__init__.py
Bmccloud22's picture
Deploy LaunchLLM - Production AI Training Platform
ec8f374 verified
"""
Evaluation Module
This module provides comprehensive model evaluation and benchmarking tools.
Includes benchmark building, certification tests, gap analysis, and training recommendations.
"""
from .metrics import Metrics, calculate_perplexity, calculate_bleu
from .evaluator import ModelEvaluator, EvaluationConfig
from .benchmark import Benchmark, BenchmarkSuite
from .benchmark_builder import Benchmark as BenchmarkBuilder, BenchmarkSuite as BenchmarkBuilderSuite
from .certification_tests import (
CertificationTestBuilder,
FinancialCertificationTests,
MedicalCertificationTests,
LegalCertificationTests,
EducationCertificationTests,
get_certification_tests_for_domain
)
from .gap_analyzer import GapAnalyzer
from .training_recommender import TrainingRecommender
__all__ = [
# Original exports
'Metrics',
'calculate_perplexity',
'calculate_bleu',
'ModelEvaluator',
'EvaluationConfig',
'Benchmark',
'BenchmarkSuite',
# New exports
'BenchmarkBuilder',
'BenchmarkBuilderSuite',
'CertificationTestBuilder',
'FinancialCertificationTests',
'MedicalCertificationTests',
'LegalCertificationTests',
'EducationCertificationTests',
'get_certification_tests_for_domain',
'GapAnalyzer',
'TrainingRecommender'
]