Skip to content

Config

Configuration module of the Wordcab Transcribe.

Settings

Configuration settings for the Wordcab Transcribe API.

Source code in src/wordcab_transcribe/config.py 
 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
@dataclass
class Settings:
    """Configuration settings for the Wordcab Transcribe API."""

    # General configuration
    project_name: str
    version: str
    description: str
    api_prefix: str
    debug: bool
    # Models configuration
    # Whisper
    whisper_model: str
    compute_type: str
    extra_languages: Union[List[str], None]
    extra_languages_model_paths: Union[Dict[str, str], None]
    # Diarization
    window_lengths: List[float]
    shift_lengths: List[float]
    multiscale_weights: List[float]
    # ASR type configuration
    asr_type: Literal["async", "live", "only_transcription", "only_diarization"]
    # Endpoint configuration
    cortex_endpoint: bool
    # API authentication configuration
    username: str
    password: str
    openssl_key: str
    openssl_algorithm: str
    access_token_expire_minutes: int
    # Cortex configuration
    cortex_api_key: str
    # Svix configuration
    svix_api_key: str
    svix_app_id: str
    # Remote servers configuration
    transcribe_server_urls: Union[List[str], None]
    diarize_server_urls: Union[List[str], None]

    @field_validator("project_name")
    def project_name_must_not_be_none(cls, value: str):  # noqa: B902, N805
        """Check that the project_name is not None."""
        if value is None:
            raise ValueError(
                "`project_name` must not be None, please verify the `.env` file."
            )

        return value

    @field_validator("version")
    def version_must_not_be_none(cls, value: str):  # noqa: B902, N805
        """Check that the version is not None."""
        if value is None:
            raise ValueError(
                "`version` must not be None, please verify the `.env` file."
            )

        return value

    @field_validator("description")
    def description_must_not_be_none(cls, value: str):  # noqa: B902, N805
        """Check that the description is not None."""
        if value is None:
            raise ValueError(
                "`description` must not be None, please verify the `.env` file."
            )

        return value

    @field_validator("api_prefix")
    def api_prefix_must_not_be_none(cls, value: str):  # noqa: B902, N805
        """Check that the api_prefix is not None."""
        if value is None:
            raise ValueError(
                "`api_prefix` must not be None, please verify the `.env` file."
            )

        return value

    @field_validator("compute_type")
    def compute_type_must_be_valid(cls, value: str):  # noqa: B902, N805
        """Check that the model precision is valid."""
        compute_type_values = [
            "int8",
            "int8_float16",
            "int8_bfloat16",
            "int16",
            "float16",
            "bfloat16",
            "float32",
        ]
        if value not in compute_type_values:
            raise ValueError(
                f"{value} is not a valid compute type. Choose one of"
                f" {compute_type_values}."
            )

        return value

    @field_validator("openssl_algorithm")
    def openssl_algorithm_must_be_valid(cls, value: str):  # noqa: B902, N805
        """Check that the OpenSSL algorithm is valid."""
        if value not in {"HS256", "HS384", "HS512"}:
            raise ValueError(
                "openssl_algorithm must be a valid algorithm, please verify the `.env`"
                " file."
            )

        return value

    @field_validator("access_token_expire_minutes")
    def access_token_expire_minutes_must_be_valid(cls, value: int):  # noqa: B902, N805
        """Check that the access token expiration is valid. Only if debug is False."""
        if value <= 0:
            raise ValueError(
                "access_token_expire_minutes must be positive, please verify the `.env`"
                " file."
            )

        return value

    def __post_init__(self):
        """Post initialization checks."""
        if self.debug is False:
            if self.username == "admin" or self.username is None:  # noqa: S105
                logger.warning(
                    f"Username is set to `{self.username}`, which is not secure for"
                    " production."
                )
            if self.password == "admin" or self.password is None:  # noqa: S105
                logger.warning(
                    f"Password is set to `{self.password}`, which is not secure for"
                    " production."
                )
            if (
                self.openssl_key == "0123456789abcdefghijklmnopqrstuvwyz"  # noqa: S105
                or self.openssl_key is None
            ):
                logger.warning(
                    f"OpenSSL key is set to `{self.openssl_key}`, which is the default"
                    " encryption key. It's absolutely not secure for production."
                    " Please change it in the `.env` file. You can generate a new key"
                    " with `openssl rand -hex 32`."
                )

        if (
            len(self.window_lengths)
            != len(self.shift_lengths)
            != len(self.multiscale_weights)
        ):
            raise ValueError(
                "Length of window_lengths, shift_lengths and multiscale_weights must"
                f" be the same.\nFound: {len(self.window_lengths)},"
                f" {len(self.shift_lengths)}, {len(self.multiscale_weights)}"
            )

__post_init__()

Post initialization checks.

Source code in src/wordcab_transcribe/config.py 
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
def __post_init__(self):
    """Post initialization checks."""
    if self.debug is False:
        if self.username == "admin" or self.username is None:  # noqa: S105
            logger.warning(
                f"Username is set to `{self.username}`, which is not secure for"
                " production."
            )
        if self.password == "admin" or self.password is None:  # noqa: S105
            logger.warning(
                f"Password is set to `{self.password}`, which is not secure for"
                " production."
            )
        if (
            self.openssl_key == "0123456789abcdefghijklmnopqrstuvwyz"  # noqa: S105
            or self.openssl_key is None
        ):
            logger.warning(
                f"OpenSSL key is set to `{self.openssl_key}`, which is the default"
                " encryption key. It's absolutely not secure for production."
                " Please change it in the `.env` file. You can generate a new key"
                " with `openssl rand -hex 32`."
            )

    if (
        len(self.window_lengths)
        != len(self.shift_lengths)
        != len(self.multiscale_weights)
    ):
        raise ValueError(
            "Length of window_lengths, shift_lengths and multiscale_weights must"
            f" be the same.\nFound: {len(self.window_lengths)},"
            f" {len(self.shift_lengths)}, {len(self.multiscale_weights)}"
        )

access_token_expire_minutes_must_be_valid(value)

Check that the access token expiration is valid. Only if debug is False.

Source code in src/wordcab_transcribe/config.py 
144
145
146
147
148
149
150
151
152
153
@field_validator("access_token_expire_minutes")
def access_token_expire_minutes_must_be_valid(cls, value: int):  # noqa: B902, N805
    """Check that the access token expiration is valid. Only if debug is False."""
    if value <= 0:
        raise ValueError(
            "access_token_expire_minutes must be positive, please verify the `.env`"
            " file."
        )

    return value

api_prefix_must_not_be_none(value)

Check that the api_prefix is not None.

Source code in src/wordcab_transcribe/config.py 
103
104
105
106
107
108
109
110
111
@field_validator("api_prefix")
def api_prefix_must_not_be_none(cls, value: str):  # noqa: B902, N805
    """Check that the api_prefix is not None."""
    if value is None:
        raise ValueError(
            "`api_prefix` must not be None, please verify the `.env` file."
        )

    return value

compute_type_must_be_valid(value)

Check that the model precision is valid.

Source code in src/wordcab_transcribe/config.py 
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
@field_validator("compute_type")
def compute_type_must_be_valid(cls, value: str):  # noqa: B902, N805
    """Check that the model precision is valid."""
    compute_type_values = [
        "int8",
        "int8_float16",
        "int8_bfloat16",
        "int16",
        "float16",
        "bfloat16",
        "float32",
    ]
    if value not in compute_type_values:
        raise ValueError(
            f"{value} is not a valid compute type. Choose one of"
            f" {compute_type_values}."
        )

    return value

description_must_not_be_none(value)

Check that the description is not None.

Source code in src/wordcab_transcribe/config.py 
 93
 94
 95
 96
 97
 98
 99
100
101
@field_validator("description")
def description_must_not_be_none(cls, value: str):  # noqa: B902, N805
    """Check that the description is not None."""
    if value is None:
        raise ValueError(
            "`description` must not be None, please verify the `.env` file."
        )

    return value

openssl_algorithm_must_be_valid(value)

Check that the OpenSSL algorithm is valid.

Source code in src/wordcab_transcribe/config.py 
133
134
135
136
137
138
139
140
141
142
@field_validator("openssl_algorithm")
def openssl_algorithm_must_be_valid(cls, value: str):  # noqa: B902, N805
    """Check that the OpenSSL algorithm is valid."""
    if value not in {"HS256", "HS384", "HS512"}:
        raise ValueError(
            "openssl_algorithm must be a valid algorithm, please verify the `.env`"
            " file."
        )

    return value

project_name_must_not_be_none(value)

Check that the project_name is not None.

Source code in src/wordcab_transcribe/config.py 
73
74
75
76
77
78
79
80
81
@field_validator("project_name")
def project_name_must_not_be_none(cls, value: str):  # noqa: B902, N805
    """Check that the project_name is not None."""
    if value is None:
        raise ValueError(
            "`project_name` must not be None, please verify the `.env` file."
        )

    return value

version_must_not_be_none(value)

Check that the version is not None.

Source code in src/wordcab_transcribe/config.py 
83
84
85
86
87
88
89
90
91
@field_validator("version")
def version_must_not_be_none(cls, value: str):  # noqa: B902, N805
    """Check that the version is not None."""
    if value is None:
        raise ValueError(
            "`version` must not be None, please verify the `.env` file."
        )

    return value
Last update: 2023-10-12
Created: 2023-10-12