Skip to content

Configuration

Relta exposes configuration from the relta.Configuration class. The following parameters are key configuration options:

Option Type Default Description
relta_dir_path Path Path(".relta") Main directory for Relta files
openai_key str - OpenAI API key (set via OPENAI_API_KEY)
debug bool False Enable debug mode (not fully implemented)
github_token str "" GitHub token (set via GITHUB_TOKEN)
github_repo str "" GitHub repository (set via GITHUB_REPO)
github_base_branch str "" GitHub base branch (set via GITHUB_BASE_BRANCH)

The list of all configuration options is available in the source code.

The typical setup is to store Relta-specific configuration in pyproject.toml in a [tool.relta] section and use a .env file for secrets like OPENAI_API_KEY.

Configuration Source Priorities

The relta.Configuration object automatically populates from the following sources, in descending order of priority:

  1. Passing in variables when initializing relta.Configuration to be passed into relta.Client
  2. relta.toml
  3. pyproject.toml in a [tool.relta] section
  4. Environment variables, or a .env file
  5. Default values

API Reference

Bases: BaseSettings

Configuration class for Relta

Any attributes ending with _dir_path will be created when a Client object is initialized.

Source code in src/relta/config.py 
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
class Configuration(BaseSettings):
    """Configuration class for Relta

    Any attributes ending with `_dir_path` will be created when a `Client` object is initialized.

    """

    # Unfortunately, default values cannot be `None`, so you will have to add some extra logic when using config variables that are optional.

    model_config = SettingsConfigDict(
        toml_file="relta.toml",
        pyproject_toml_table_header=("tool", "relta"),
        env_file=".env",
        extra="ignore",
    )

    @classmethod
    def settings_customise_sources(
        cls,
        settings_cls: Type[BaseSettings],
        init_settings: PydanticBaseSettingsSource,
        env_settings: PydanticBaseSettingsSource,
        dotenv_settings: PydanticBaseSettingsSource,
        file_secret_settings: PydanticBaseSettingsSource,
    ) -> tuple[PydanticBaseSettingsSource, ...]:
        return (
            init_settings,
            env_settings,
            TomlConfigSettingsSource(settings_cls),
            PyprojectTomlConfigSettingsSource(settings_cls),
            dotenv_settings,
            file_secret_settings,
        )

    relta_dir_path: Path = Path(".relta")
    relta_semantic_layer_dir_path: Path = relta_dir_path / "semantic_layer"
    transient_data_dir_path: Path = relta_dir_path / "data"
    relta_internal_path: Path = relta_dir_path / "relta_internal.duckdb"
    relta_data_path: Path = relta_dir_path / "relta_data.duckdb"
    chat_memory_path: Path = relta_dir_path / "chat_memory.sqlite"
    semantic_memory_path: Path = relta_dir_path / "semantic_memory.sqlite"
    storage_path: Path = relta_dir_path / "relta.sqlite"
    storage_table: str = "messages"
    storage_session_id_field: str = "session_id"
    openai_key: str = Field(alias="OPENAI_API_KEY")
    json_dumps_kwargs: dict = {"indent": 2}
    yaml_dumps_kwargs: dict = {"sort_keys": False}

    encryption_key: str = Field(alias="ENCRYPTION_KEY", default="")
    auto_generate_encryption_key: bool = Field(
        alias="AUTO_GENERATE_ENCRYPTION_KEY", default=True
    )

    debug: bool = False  # TODO: Further implement debug mode.
    anonymized_telemetry: bool = True
    low_cardinality_cutoff: int = 100

    github_token: str = Field(alias="GITHUB_TOKEN", default="")
    github_repo: str = Field(alias="GITHUB_REPO", default="")
    github_base_branch: str = Field(alias="GITHUB_BASE_BRANCH", default="")

    logfire_token: str = Field(alias="LOGFIRE_TOKEN", default="")