Spaces:
Running
Running
Create embeddings.py
Browse files- embeddings.py +112 -0
embeddings.py
ADDED
@@ -0,0 +1,112 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import os
|
2 |
+
import json
|
3 |
+
import logging
|
4 |
+
from typing import List
|
5 |
+
|
6 |
+
from txtai.embeddings import Embeddings
|
7 |
+
|
8 |
+
# Set up logging
|
9 |
+
logging.basicConfig(level=logging.INFO)
|
10 |
+
logger = logging.getLogger(__name__)
|
11 |
+
|
12 |
+
|
13 |
+
class EmbeddingsManager:
|
14 |
+
def __init__(self, base_path: str = "./indexes", model_path: str = "avsolatorio/GIST-all-MiniLM-L6-v2"):
|
15 |
+
"""
|
16 |
+
Initializes the EmbeddingsManager.
|
17 |
+
|
18 |
+
Args:
|
19 |
+
base_path (str): Base directory to store indices.
|
20 |
+
model_path (str): Path or identifier for the embeddings model.
|
21 |
+
"""
|
22 |
+
self.base_path = base_path
|
23 |
+
os.makedirs(self.base_path, exist_ok=True)
|
24 |
+
self.model_path = model_path
|
25 |
+
self.embeddings = Embeddings({"path": self.model_path})
|
26 |
+
logger.info(f"Embeddings model loaded from '{self.model_path}'. Base path set to '{self.base_path}'.")
|
27 |
+
|
28 |
+
def create_index(self, index_id: str, documents: List[str]) -> None:
|
29 |
+
"""
|
30 |
+
Creates a new embeddings index with the provided documents.
|
31 |
+
|
32 |
+
Args:
|
33 |
+
index_id (str): Unique identifier for the index.
|
34 |
+
documents (List[str]): List of documents to be indexed.
|
35 |
+
|
36 |
+
Raises:
|
37 |
+
ValueError: If the index already exists.
|
38 |
+
Exception: For any other errors during indexing or saving.
|
39 |
+
"""
|
40 |
+
index_path = os.path.join(self.base_path, index_id)
|
41 |
+
if os.path.exists(index_path):
|
42 |
+
logger.error(f"Index with index_id '{index_id}' already exists at '{index_path}'.")
|
43 |
+
raise ValueError(f"Index with index_id '{index_id}' already exists.")
|
44 |
+
|
45 |
+
try:
|
46 |
+
# Prepare documents for txtai indexing
|
47 |
+
document_tuples = [(i, text, None) for i, text in enumerate(documents)]
|
48 |
+
self.embeddings.index(document_tuples)
|
49 |
+
logger.info(f"Documents indexed for index_id '{index_id}'.")
|
50 |
+
|
51 |
+
# Create index directory
|
52 |
+
os.makedirs(index_path, exist_ok=True)
|
53 |
+
|
54 |
+
# Save embeddings
|
55 |
+
self.embeddings.save(os.path.join(index_path, "embeddings"))
|
56 |
+
logger.info(f"Embeddings saved to '{os.path.join(index_path, 'embeddings')}'.")
|
57 |
+
|
58 |
+
# Save document list
|
59 |
+
with open(os.path.join(index_path, "document_list.json"), "w", encoding='utf-8') as f:
|
60 |
+
json.dump(documents, f, ensure_ascii=False, indent=4)
|
61 |
+
logger.info(f"Document list saved to '{os.path.join(index_path, 'document_list.json')}'.")
|
62 |
+
|
63 |
+
logger.info(f"Index '{index_id}' created and saved successfully.")
|
64 |
+
except Exception as e:
|
65 |
+
logger.error(f"Failed to create index '{index_id}': {e}")
|
66 |
+
raise Exception(f"Failed to create index '{index_id}': {e}")
|
67 |
+
|
68 |
+
def query_index(self, index_id: str, query: str, num_results: int = 5) -> List[str]:
|
69 |
+
"""
|
70 |
+
Queries an existing embeddings index.
|
71 |
+
|
72 |
+
Args:
|
73 |
+
index_id (str): Unique identifier for the index to query.
|
74 |
+
query (str): The search query.
|
75 |
+
num_results (int): Number of top results to return.
|
76 |
+
|
77 |
+
Returns:
|
78 |
+
List[str]: List of top matching documents.
|
79 |
+
|
80 |
+
Raises:
|
81 |
+
FileNotFoundError: If the index does not exist.
|
82 |
+
Exception: For any other errors during querying.
|
83 |
+
"""
|
84 |
+
index_path = os.path.join(self.base_path, index_id)
|
85 |
+
if not os.path.exists(index_path):
|
86 |
+
logger.error(f"Index '{index_id}' not found at '{index_path}'.")
|
87 |
+
raise FileNotFoundError(f"Index '{index_id}' not found.")
|
88 |
+
|
89 |
+
try:
|
90 |
+
# Load embeddings from the index
|
91 |
+
self.embeddings.load(os.path.join(index_path, "embeddings"))
|
92 |
+
logger.info(f"Embeddings loaded from '{os.path.join(index_path, 'embeddings')}' for index '{index_id}'.")
|
93 |
+
|
94 |
+
# Load document list
|
95 |
+
document_list_path = os.path.join(index_path, "document_list.json")
|
96 |
+
if not os.path.exists(document_list_path):
|
97 |
+
logger.error(f"Document list not found at '{document_list_path}'.")
|
98 |
+
raise FileNotFoundError(f"Document list not found for index '{index_id}'.")
|
99 |
+
|
100 |
+
with open(document_list_path, "r", encoding='utf-8') as f:
|
101 |
+
document_list = json.load(f)
|
102 |
+
logger.info(f"Document list loaded from '{document_list_path}'.")
|
103 |
+
|
104 |
+
# Perform the search
|
105 |
+
results = self.embeddings.search(query, num_results)
|
106 |
+
queried_texts = [document_list[idx[0]] for idx in results]
|
107 |
+
logger.info(f"Query executed successfully on index '{index_id}'. Retrieved {len(queried_texts)} results.")
|
108 |
+
|
109 |
+
return queried_texts
|
110 |
+
except Exception as e:
|
111 |
+
logger.error(f"Failed to query index '{index_id}': {e}")
|
112 |
+
raise Exception(f"Failed to query index '{index_id}': {e}")
|