File size: 4,986 Bytes
d7ae59e dbabef2 d7ae59e | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 |
from __future__ import annotations
import json
import pickle
from pathlib import Path
import numpy as np
from sentence_transformers import SentenceTransformer
# =========================================================
# EMBEDDING GENERATOR
# =========================================================
class EmbeddingGenerator:
def __init__(
self,
model_name: str = "BAAI/bge-large-en-v1.5"
):
self.model = SentenceTransformer(
model_name
)
# =====================================================
# GENERATE
# =====================================================
def generate(
self,
chunks: list,
batch_size: int = 32
):
texts = [
chunk.text
for chunk in chunks
]
embeddings = self.model.encode(
texts,
batch_size=batch_size,
show_progress_bar=True,
convert_to_numpy=True,
normalize_embeddings=True
)
return embeddings
# =====================================================
# SAVE NUMPY
# =====================================================
def save_embeddings(
self,
embeddings,
output_path: str
):
np.save(
output_path,
embeddings
)
# =====================================================
# SAVE COMPLETE DATASET
# =====================================================
def save_dataset(
self,
chunks,
embeddings,
output_file: str
):
records = []
for chunk, embedding in zip(
chunks,
embeddings
):
records.append(
{
"chunk_id":
chunk.chunk_id,
"chunk_type":
chunk.chunk_type,
"text":
chunk.text,
"metadata":
chunk.metadata,
"references":
chunk.references,
"embedding":
embedding.tolist()
}
)
with open(
output_file,
"w",
encoding="utf8"
) as f:
json.dump(
records,
f,
ensure_ascii=False
)
# =====================================================
# PICKLE
# =====================================================
def save_pickle(
self,
chunks,
embeddings,
output_file
):
data = []
for chunk, embedding in zip(
chunks,
embeddings
):
data.append(
{
"chunk_id":
chunk.chunk_id,
"chunk_type":
chunk.chunk_type,
"text":
chunk.text,
"metadata":
chunk.metadata,
"references":
chunk.references,
"embedding":
embedding
}
)
with open(
output_file,
"wb"
) as f:
pickle.dump(
data,
f
)
# =========================================================
# EXAMPLE
# =========================================================
if __name__ == "__main__":
from constitution_parser import ConstitutionParser
from db.parsers.consitution.chunker import LegalChunker
with open(
"../../pdfs/constitution.txt",
"r",
encoding="utf8"
) as f:
text = f.read()
# -----------------------------------------
# Parse
# -----------------------------------------
parser = ConstitutionParser()
constitution = parser.parse(
text
)
# -----------------------------------------
# Chunk
# -----------------------------------------
chunker = LegalChunker()
chunks = chunker.chunk_constitution(
constitution
)
print(
"Chunks:",
len(chunks)
)
# -----------------------------------------
# Embeddings
# -----------------------------------------
generator = EmbeddingGenerator()
embeddings = generator.generate(
chunks
)
print(
"Embeddings Shape:",
embeddings.shape
)
# -----------------------------------------
# Save
# -----------------------------------------
generator.save_embeddings(
embeddings,
"constitution_embeddings.npy"
)
generator.save_pickle(
chunks,
embeddings,
"constitution_dataset.pkl"
)
print(
"Saved successfully"
)
|