20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154 | class SentenceTransformerRerank(BaseNodePostprocessor):
"""
HuggingFace class for cross encoding two sentences/texts.
Args:
model (str): A model name from Hugging Face Hub that can be loaded with AutoModel, or a path to a local model.
device (str, optional): Device (like “cuda”, “cpu”, “mps”, “npu”) that should be used for computation.
If None, checks if a GPU can be used.
cache_folder (str, Path, optional): Path to the folder where cached files are stored. Defaults to None.
top_n (int): Number of nodes to return sorted by score. Defaults to 2.
keep_retrieval_score (bool, optional): Whether to keep the retrieval score in metadata. Defaults to False.
cross_encoder_kwargs (dict, optional): Additional keyword arguments for CrossEncoder initialization. Defaults to None.
"""
model: str = Field(description="Sentence transformer model name.")
top_n: int = Field(description="Number of nodes to return sorted by score.")
keep_retrieval_score: bool = Field(
default=False,
description="Whether to keep the retrieval score in metadata.",
)
cross_encoder_kwargs: dict = Field(
default_factory=dict,
description="Additional keyword arguments for CrossEncoder initialization. "
"device and model should not be included here.",
)
_model: Any = PrivateAttr()
_device: str = PrivateAttr()
def __init__(
self,
model: str = "cross-encoder/stsb-distilroberta-base",
device: Optional[str] = None,
cache_folder: Optional[Union[str, Path]] = None,
top_n: int = 2,
keep_retrieval_score: Optional[bool] = False,
cross_encoder_kwargs: Optional[dict] = None,
):
try:
from sentence_transformers import CrossEncoder
except ImportError:
raise ImportError(
"Cannot import sentence-transformers or torch package,",
"please `pip install torch sentence-transformers`",
)
super().__init__(
top_n=top_n,
model=model,
device=device,
keep_retrieval_score=keep_retrieval_score,
cross_encoder_kwargs=cross_encoder_kwargs or {},
)
init_kwargs = self.cross_encoder_kwargs.copy()
if "device" in init_kwargs or "model" in init_kwargs:
raise ValueError(
"'device' and 'model' should not be specified in 'cross_encoder_kwargs'. "
"Use the top-level 'device' and 'model' parameters instead."
)
# Set default max_length if not provided by the user in kwargs.
if "max_length" not in init_kwargs:
init_kwargs["max_length"] = DEFAULT_SENTENCE_TRANSFORMER_MAX_LENGTH
# Explicit arguments from the constructor take precedence over kwargs
resolved_device = infer_torch_device() if device is None else device
init_kwargs["device"] = resolved_device
self._device = resolved_device
if cache_folder:
init_kwargs["cache_folder"] = cache_folder
self._model = CrossEncoder(
model_name_or_path=model,
**init_kwargs,
)
@classmethod
def class_name(cls) -> str:
return "SentenceTransformerRerank"
def _postprocess_nodes(
self,
nodes: List[NodeWithScore],
query_bundle: Optional[QueryBundle] = None,
) -> List[NodeWithScore]:
dispatcher.event(
ReRankStartEvent(
query=query_bundle,
nodes=nodes,
top_n=self.top_n,
model_name=self.model,
)
)
if query_bundle is None:
raise ValueError("Missing query bundle in extra info.")
if len(nodes) == 0:
return []
query_and_nodes = [
(
query_bundle.query_str,
node.node.get_content(metadata_mode=MetadataMode.EMBED),
)
for node in nodes
]
with self.callback_manager.event(
CBEventType.RERANKING,
payload={
EventPayload.NODES: nodes,
EventPayload.MODEL_NAME: self.model,
EventPayload.QUERY_STR: query_bundle.query_str,
EventPayload.TOP_K: self.top_n,
},
) as event:
scores = self._model.predict(query_and_nodes)
assert len(scores) == len(nodes)
for node, score in zip(nodes, scores):
if self.keep_retrieval_score:
# keep the retrieval score in metadata
node.node.metadata["retrieval_score"] = node.score
node.score = float(score)
new_nodes = sorted(nodes, key=lambda x: -x.score if x.score else 0)[
: self.top_n
]
event.on_end(payload={EventPayload.NODES: new_nodes})
dispatcher.event(ReRankEndEvent(nodes=new_nodes))
return new_nodes
|