llama-index

1
"""Evaluation modules."""
2

3
from llama_index.legacy.evaluation.answer_relevancy import AnswerRelevancyEvaluator
4
from llama_index.legacy.evaluation.base import (
5
    BaseEvaluator,
6
    EvaluationResult,
7
)
8
from llama_index.legacy.evaluation.batch_runner import BatchEvalRunner
9
from llama_index.legacy.evaluation.context_relevancy import ContextRelevancyEvaluator
10
from llama_index.legacy.evaluation.correctness import CorrectnessEvaluator
11
from llama_index.legacy.evaluation.dataset_generation import (
12
    DatasetGenerator,
13
    QueryResponseDataset,
14
)
15
from llama_index.legacy.evaluation.faithfulness import (
16
    FaithfulnessEvaluator,
17
    ResponseEvaluator,
18
)
19
from llama_index.legacy.evaluation.guideline import GuidelineEvaluator
20
from llama_index.legacy.evaluation.notebook_utils import get_retrieval_results_df
21
from llama_index.legacy.evaluation.pairwise import PairwiseComparisonEvaluator
22
from llama_index.legacy.evaluation.relevancy import (
23
    QueryResponseEvaluator,
24
    RelevancyEvaluator,
25
)
26
from llama_index.legacy.evaluation.retrieval.base import (
27
    BaseRetrievalEvaluator,
28
    RetrievalEvalResult,
29
)
30
from llama_index.legacy.evaluation.retrieval.evaluator import (
31
    MultiModalRetrieverEvaluator,
32
    RetrieverEvaluator,
33
)
34
from llama_index.legacy.evaluation.retrieval.metrics import (
35
    MRR,
36
    HitRate,
37
    RetrievalMetricResult,
38
    resolve_metrics,
39
)
40
from llama_index.legacy.evaluation.semantic_similarity import (
41
    SemanticSimilarityEvaluator,
42
)
43
from llama_index.legacy.evaluation.tonic_validate.answer_consistency import (
44
    AnswerConsistencyEvaluator,
45
)
46
from llama_index.legacy.evaluation.tonic_validate.answer_consistency_binary import (
47
    AnswerConsistencyBinaryEvaluator,
48
)
49
from llama_index.legacy.evaluation.tonic_validate.answer_similarity import (
50
    AnswerSimilarityEvaluator,
51
)
52
from llama_index.legacy.evaluation.tonic_validate.augmentation_accuracy import (
53
    AugmentationAccuracyEvaluator,
54
)
55
from llama_index.legacy.evaluation.tonic_validate.augmentation_precision import (
56
    AugmentationPrecisionEvaluator,
57
)
58
from llama_index.legacy.evaluation.tonic_validate.retrieval_precision import (
59
    RetrievalPrecisionEvaluator,
60
)
61
from llama_index.legacy.evaluation.tonic_validate.tonic_validate_evaluator import (
62
    TonicValidateEvaluator,
63
)
64

65
# import dataset generation too
66
from llama_index.legacy.finetuning.embeddings.common import (
67
    EmbeddingQAFinetuneDataset,
68
    generate_qa_embedding_pairs,
69
)
70

71
# aliases for generate_qa_embedding_pairs
72
generate_question_context_pairs = generate_qa_embedding_pairs
73
LabelledQADataset = EmbeddingQAFinetuneDataset
74

75
__all__ = [
76
    "BaseEvaluator",
77
    "AnswerRelevancyEvaluator",
78
    "ContextRelevancyEvaluator",
79
    "EvaluationResult",
80
    "FaithfulnessEvaluator",
81
    "RelevancyEvaluator",
82
    "RelevanceEvaluator",
83
    "DatasetGenerator",
84
    "QueryResponseDataset",
85
    "GuidelineEvaluator",
86
    "CorrectnessEvaluator",
87
    "SemanticSimilarityEvaluator",
88
    "PairwiseComparisonEvaluator",
89
    "BatchEvalRunner",
90
    # legacy: kept for backward compatibility
91
    "QueryResponseEvaluator",
92
    "ResponseEvaluator",
93
    # retrieval
94
    "generate_qa_embedding_pairs",
95
    "generate_question_context_pairs",
96
    "EmbeddingQAFinetuneDataset",
97
    "BaseRetrievalEvaluator",
98
    "RetrievalEvalResult",
99
    "RetrieverEvaluator",
100
    "MultiModalRetrieverEvaluator",
101
    "RetrievalMetricResult",
102
    "resolve_metrics",
103
    "HitRate",
104
    "MRR",
105
    "get_retrieval_results_df",
106
    "LabelledQADataset",
107
    # tonic_validate evaluators
108
    "AnswerConsistencyEvaluator",
109
    "AnswerConsistencyBinaryEvaluator",
110
    "AnswerSimilarityEvaluator",
111
    "AugmentationAccuracyEvaluator",
112
    "AugmentationPrecisionEvaluator",
113
    "RetrievalPrecisionEvaluator",
114
    "TonicValidateEvaluator",
115
]
116