Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fix to Cosine Similiarity To Probability Clipping #326

Merged
merged 1 commit into from
Jan 28, 2025
Merged

Fix to Cosine Similiarity To Probability Clipping #326

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
15 changes: 10 additions & 5 deletions adalflow/adalflow/components/retriever/faiss_retriever.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,5 +1,6 @@
"""Semantic search/embedding-based retriever using FAISS."""

import faiss
from typing import (
List,
Optional,
Expand Down Expand Up @@ -29,17 +30,18 @@
from adalflow.utils.lazy_import import safe_import, OptionalPackages

safe_import(OptionalPackages.FAISS.value[0], OptionalPackages.FAISS.value[1])
import faiss

log = logging.getLogger(__name__)

FAISSRetrieverDocumentEmbeddingType = Union[List[float], np.ndarray] # single embedding
# single embedding
FAISSRetrieverDocumentEmbeddingType = Union[List[float], np.ndarray]
FAISSRetrieverDocumentsType = Sequence[FAISSRetrieverDocumentEmbeddingType]

FAISSRetrieverEmbeddingQueryType = Union[
List[float], List[List[float]], np.ndarray
] # single embedding or list of embeddings
FAISSRetrieverQueryType = Union[RetrieverStrQueryType, FAISSRetrieverEmbeddingQueryType]
FAISSRetrieverQueryType = Union[RetrieverStrQueryType,
FAISSRetrieverEmbeddingQueryType]
FAISSRetrieverQueriesType = Sequence[FAISSRetrieverQueryType]
FAISSRetrieverQueriesStrType = Sequence[RetrieverStrQueryType]
FAISSRetrieverQueriesEmbeddingType = Sequence[FAISSRetrieverEmbeddingQueryType]
Expand Down Expand Up @@ -161,7 +163,8 @@ def build_index_from_documents(
If you are using Document format, pass them as [doc.vector for doc in documents]
"""
if document_map_func:
assert callable(document_map_func), "document_map_func should be callable"
assert callable(
document_map_func), "document_map_func should be callable"
documents = [document_map_func(doc) for doc in documents]
try:
self.documents = documents
Expand Down Expand Up @@ -194,6 +197,7 @@ def build_index_from_documents(
raise e

def _convert_cosine_similarity_to_probability(self, D: np.ndarray) -> np.ndarray:
D = np.clip(D, -1, 1)
D = (D + 1) / 2
D = np.round(D, 3)
return D
Expand Down Expand Up @@ -295,7 +299,8 @@ def retrieve_string_queries(
output: RetrieverOutputType = [
RetrieverOutput(doc_indices=[], query=query) for query in queries
]
retrieved_output: RetrieverOutputType = self._to_retriever_output(Ind, D)
retrieved_output: RetrieverOutputType = self._to_retriever_output(
Ind, D)

# fill in the doc_indices and score for valid queries
for i, per_query_output in enumerate(retrieved_output):
Expand Down
Loading