hubai_sdk.utils.hub

module documentation

Undocumented

Function	`full_slug_to_id`	Resolve a full HubAI slug to a resource ID using the slug API.
Function	`get_configs`	Sets up the configuration.
Function	`get_output_dir_name`	Build an output directory path for a conversion result.
Function	`get_resource_id`	Resolve a UUID, HubAI ID, or slug into a resource ID.
Function	`get_resource_info`	Fetch raw HubAI resource data for a resolved identifier.
Function	`get_target_specific_options`	Build export options for a target from a parsed config.
Function	`get_variant_name`	Derive a variant name, appending input resolution when known.
Function	`get_version_number`	Return the next patch version for a model's variants.
Function	`init_dirs`	Create the SDK working directories if they do not exist.
Function	`is_hubai_id`	Return whether an identifier looks like a HubAI resource ID.
Function	`is_valid_uuid`	Return whether a string is a valid UUID.
Function	`print_hub_ls`	Render a list of HubAI resources as a rich table.
Function	`print_hub_resource_info`	Render a single HubAI resource in CLI-friendly rich output.
Function	`raise_for_hub_error`	Translate a requests `HTTPError` into an SDK-specific exception.
Function	`resolve_resource_id`	Resolve an identifier and translate lookup failures to SDK errors.
Function	`run_cli`	Run a CLI action and convert SDK exceptions into clean exits.
Function	`slug_to_id`	Resolve a simple slug to a resource ID by querying HubAI.
Function	`wait_for_job`	Poll a HubAI job until it completes or fails.
Type Variable	`T`	Undocumented
Type Alias	`ResourceEndpoint`	Undocumented
Function	`_format_http_error_items`	Format validation-style error items into a single message string.
Function	`_get_http_error_detail`	Extract a readable error message from a requests HTTPError.

def full_slug_to_id(slug: str, endpoint: Literal['models', 'modelVersions', 'modelInstances']) -> str: ¶

Resolve a full HubAI slug to a resource ID using the slug API.

Sets up the configuration.

Parameters
path:`str \| None`	Path to the configuration file or NN Archive.
opts:`list[str] \| dict[str, Any] \| None`	Optional CLI overrides for the config file.

Returns
`tuple[Config, NNArchiveConfig \| None, str \| None]`	Tuple of the parsed modelconverter `Config`, optional `NNArchiveConfig`, and the main stage key.

def get_output_dir_name(target: Target, name: str, output_dir: str | None) -> Path: ¶

Build an output directory path for a conversion result.

def get_resource_id(identifier: str, endpoint: Literal['models', 'modelVersions', 'modelInstances']) -> str: ¶

Resolve a UUID, HubAI ID, or slug into a resource ID.

Parameters
identifier:`str`	User-supplied UUID, HubAI ID, or slug.
endpoint:`Literal['models', 'modelVersions', 'modelInstances']`	Resource endpoint to resolve against.

Returns
`str`	The resolved HubAI resource ID.

Raises
`ValueError`	If the identifier cannot be resolved.

def get_resource_info(identifier: str, endpoint: ResourceEndpoint) -> dict[str, Any]: ¶

Fetch raw HubAI resource data for a resolved identifier.

def get_target_specific_options(target: Target, cfg: SingleStageConfig, tool_version: str | None = None) -> dict[str, Any]: ¶

Build export options for a target from a parsed config.

def get_variant_name(cfg: SingleStageConfig, model_type: ModelType, name: str) -> str: ¶

Derive a variant name, appending input resolution when known.

def get_version_number(model_id: str) -> str: ¶

Return the next patch version for a model's variants.

def init_dirs(): ¶

Create the SDK working directories if they do not exist.

def is_hubai_id(identifier: str) -> bool: ¶

Return whether an identifier looks like a HubAI resource ID.

def is_valid_uuid(uuid_string: str) -> bool: ¶

Return whether a string is a valid UUID.

def print_hub_ls(data: list[dict[str, Any]], keys: list[str], rename: dict[str, str] | None = None): ¶

Render a list of HubAI resources as a rich table.

def print_hub_resource_info(model: dict[str, Any], keys: list[str], json: bool, rename: dict[str, str] | None = None, **kwargs): ¶

Render a single HubAI resource in CLI-friendly rich output.

def raise_for_hub_error(exc: HTTPError, *, identifier: str | None = None, endpoint: ResourceEndpoint | Literal['jobs'] | None = None, conflict_message: str | None = None) -> NoReturn: ¶

Translate a requests HTTPError into an SDK-specific exception.

def resolve_resource_id(identifier: str, endpoint: ResourceEndpoint) -> str: ¶

Resolve an identifier and translate lookup failures to SDK errors.

def run_cli(action: Callable[[], T]) -> T: ¶

Run a CLI action and convert SDK exceptions into clean exits.

def slug_to_id(slug: str, endpoint: Literal['models', 'modelVersions', 'modelInstances']) -> str: ¶

Resolve a simple slug to a resource ID by querying HubAI.

def wait_for_job(job_id: str) -> JobMessageResponse: ¶

Poll a HubAI job until it completes or fails.

Parameters
job_id:`str`	HubAI job ID.

Returns
`JobMessageResponse`	The completed job response.

Raises
`HubApiError`	If the job fails or ends in a terminal non-success status.

T = ¶

Undocumented

Value

TypeVar('T')

ResourceEndpoint = ¶

Undocumented

Value

Literal['models', 'modelVersions', 'modelInstances']

def _format_http_error_items(items: list[object]) -> str | None: ¶

Format validation-style error items into a single message string.

def _get_http_error_detail(exc: HTTPError) -> str: ¶

Extract a readable error message from a requests HTTPError.