InitConfig.py 2.5 KB

1234567891011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556575859606162636465666768697071727374
  1. import logging
  2. import os
  3. from chromadb.utils import embedding_functions
  4. from embedchain.config.BaseConfig import BaseConfig
  5. class InitConfig(BaseConfig):
  6. """
  7. Config to initialize an embedchain `App` instance.
  8. """
  9. def __init__(self, log_level=None, ef=None, db=None, host=None, port=None, id=None):
  10. """
  11. :param log_level: Optional. (String) Debug level
  12. ['DEBUG', 'INFO', 'WARNING', 'ERROR', 'CRITICAL'].
  13. :param ef: Optional. Embedding function to use.
  14. :param db: Optional. (Vector) database to use for embeddings.
  15. :param id: Optional. ID of the app. Document metadata will have this id.
  16. :param host: Optional. Hostname for the database server.
  17. :param port: Optional. Port for the database server.
  18. """
  19. self._setup_logging(log_level)
  20. self.ef = ef
  21. self.db = db
  22. self.host = host
  23. self.port = port
  24. self.id = id
  25. return
  26. def _set_embedding_function(self, ef):
  27. self.ef = ef
  28. return
  29. def _set_embedding_function_to_default(self):
  30. """
  31. Sets embedding function to default (`text-embedding-ada-002`).
  32. :raises ValueError: If the template is not valid as template should contain
  33. $context and $query
  34. """
  35. if os.getenv("OPENAI_API_KEY") is None and os.getenv("OPENAI_ORGANIZATION") is None:
  36. raise ValueError("OPENAI_API_KEY or OPENAI_ORGANIZATION environment variables not provided") # noqa:E501
  37. self.ef = embedding_functions.OpenAIEmbeddingFunction(
  38. api_key=os.getenv("OPENAI_API_KEY"),
  39. organization_id=os.getenv("OPENAI_ORGANIZATION"),
  40. model_name="text-embedding-ada-002",
  41. )
  42. return
  43. def _set_db(self, db):
  44. if db:
  45. self.db = db
  46. return
  47. def _set_db_to_default(self):
  48. """
  49. Sets database to default (`ChromaDb`).
  50. """
  51. from embedchain.vectordb.chroma_db import ChromaDB
  52. self.db = ChromaDB(ef=self.ef, host=self.host, port=self.port)
  53. def _setup_logging(self, debug_level):
  54. level = logging.WARNING # Default level
  55. if debug_level is not None:
  56. level = getattr(logging, debug_level.upper(), None)
  57. if not isinstance(level, int):
  58. raise ValueError(f"Invalid log level: {debug_level}")
  59. logging.basicConfig(format="%(asctime)s [%(name)s] [%(levelname)s] %(message)s", level=level)
  60. self.logger = logging.getLogger(__name__)
  61. return