Vectorstore module
Here we handle the application of vectorstore services to retrieval-augmented generation tasks by embedding documents.
DocumentEmbedder
Source code in biochatter/vectorstore.py
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 |
|
__init__(used=False, online=False, chunk_size=1000, chunk_overlap=0, split_by_characters=True, separators=None, n_results=3, model='text-embedding-ada-002', vector_db_vendor=None, connection_args=None, embedding_collection_name=None, metadata_collection_name=None, api_key=None, is_azure=False, azure_deployment=None, azure_endpoint=None, base_url=None, embeddings=None, documentids_workspace=None)
Class that handles the retrieval-augmented generation (RAG) functionality
of BioChatter. It splits text into chunks, embeds them, and stores them in
a vector database. It can then be used to do similarity search on the
database.
Args:
used (bool, optional): whether RAG has been used (ChatGSE setting).
Defaults to False.
online (bool, optional): whether we are running ChatGSE online.
Defaults to False.
chunk_size (int, optional): size of chunks to split text into.
Defaults to 1000.
chunk_overlap (int, optional): overlap between chunks. Defaults to 0.
split_by_characters (bool, optional): whether to split by characters
or tokens. Defaults to True.
separators (Optional[list], optional): list of separators to use when
splitting by characters. Defaults to [" ", ",", "
"].
n_results (int, optional): number of results to return from
similarity search. Defaults to 3.
model (Optional[str], optional): name of model to use for embeddings.
Defaults to 'text-embedding-ada-002'.
vector_db_vendor (Optional[str], optional): name of vector database
to use. Defaults to Milvus.
connection_args (Optional[dict], optional): arguments to pass to
vector database connection. Defaults to None.
api_key (Optional[str], optional): OpenAI API key. Defaults to None.
base_url (Optional[str], optional): base url of OpenAI API.
embeddings (Optional[OpenAIEmbeddings | XinferenceEmbeddings],
optional): Embeddings object to use. Defaults to OpenAI.
documentids_workspace (Optional[List[str]], optional): a list of document IDs
that defines the scope within which rag operations (remove, similarity search,
and get all) occur. Defaults to None, which means the operations will be
performed across all documents in the database.
is_azure (Optional[bool], optional): if we are using Azure
azure_deployment (Optional[str], optional): Azure embeddings model deployment,
should work with azure_endpoint when is_azure is True
azure_endpoint (Optional[str], optional): Azure endpoint, should work with
azure_deployment when is_azure is True
Source code in biochatter/vectorstore.py
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 |
|
save_document(doc)
This function saves document to the vector database Args: doc List[Document]: document content, read with DocumentReader load_document(), or document_from_pdf(), document_from_txt() Returns: str: document id, which can be used to remove an uploaded document with remove_document()
Source code in biochatter/vectorstore.py
DocumentReader
Source code in biochatter/vectorstore.py
364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 |
|
document_from_pdf(pdf)
Receive a byte representation of a pdf file and return a list of Documents with metadata.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
pdf |
bytes
|
byte representation of pdf file |
required |
Returns:
Type | Description |
---|---|
list[Document]
|
List[Document]: list of documents |
Source code in biochatter/vectorstore.py
document_from_txt(txt)
Receive a byte representation of a txt file and return a list of Documents with metadata.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
txt |
bytes
|
byte representation of txt file |
required |
Returns:
Type | Description |
---|---|
list[Document]
|
List[Document]: list of documents |
Source code in biochatter/vectorstore.py
load_document(path)
Loads a document from a path; accepts txt and pdf files. Txt files are loaded as-is, pdf files are converted to text using fitz.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
path |
str
|
path to document |
required |
Returns:
Type | Description |
---|---|
list[Document]
|
List[Document]: list of documents |
Source code in biochatter/vectorstore.py
XinferenceDocumentEmbedder
Bases: DocumentEmbedder
Source code in biochatter/vectorstore.py
236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 |
|
__init__(used=False, chunk_size=1000, chunk_overlap=0, split_by_characters=True, separators=None, n_results=3, model='auto', vector_db_vendor=None, connection_args=None, embedding_collection_name=None, metadata_collection_name=None, api_key='none', base_url=None, documentids_workspace=None)
Extension of the DocumentEmbedder class that uses Xinference for embeddings.
Args:
used (bool, optional): whether RAG has been used (ChatGSE setting).
chunk_size (int, optional): size of chunks to split text into.
chunk_overlap (int, optional): overlap between chunks.
split_by_characters (bool, optional): whether to split by characters
or tokens.
separators (Optional[list], optional): list of separators to use when
splitting by characters.
n_results (int, optional): number of results to return from
similarity search.
model (Optional[str], optional): name of model to use for embeddings.
Can be "auto" to use the first available model.
vector_db_vendor (Optional[str], optional): name of vector database
to use.
connection_args (Optional[dict], optional): arguments to pass to
vector database connection.
embedding_collection_name (Optional[str], optional): name of
collection to store embeddings in.
metadata_collection_name (Optional[str], optional): name of
collection to store metadata in.
api_key (Optional[str], optional): Xinference API key.
base_url (Optional[str], optional): base url of Xinference API.
documentids_workspace (Optional[List[str]], optional): a list of document IDs
that defines the scope within which rag operations (remove, similarity search,
and get all) occur. Defaults to None, which means the operations will be
performed across all documents in the database.
Source code in biochatter/vectorstore.py
237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 |
|
list_models_by_type(type)
Return all models of a certain type that are currently available on the Xinference server.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
type |
str
|
type of model to list (e.g. "embedding", "chat") |
required |
Returns:
Type | Description |
---|---|
List[str]: list of model names |
Source code in biochatter/vectorstore.py
load_models()
Return all models that are currently available on the Xinference server.
Returns:
Name | Type | Description |
---|---|---|
dict |
dict of models |