HumeBatchClient

class HumeBatchClient(ClientBase)

Example

from hume import HumeBatchClient
from hume.models.config import FaceConfig
from hume.models.config import ProsodyConfig

client = HumeBatchClient("<your-api-key>")
urls = ["https://storage.googleapis.com/hume-test-data/video/armisen-clip.mp4"]
configs = [FaceConfig(identify_faces=True), ProsodyConfig()]
job = client.submit_job(urls, configs)

print(job)
print("Running...")

job.await_complete()
job.download_predictions("predictions.json")
print("Predictions downloaded to predictions.json")

job.download_artifacts("artifacts.zip")
print("Artifacts downloaded to artifacts.zip")

init

Construct a HumeBatchClient.

def __init__(api_key: str, *args: Any, timeout: int = 10, **kwargs: Any)

Arguments:

api_key str - Hume API key.
timeout int - Time in seconds before canceling long-running Hume API requests.

get_api_type

Get the ApiType of the client.

@classmethod
def get_api_type(cls) -> ApiType

Returns:

ApiType - API type of the client.

get_job

Rehydrate a job based on a Job ID.

def get_job(job_id: str) -> BatchJob

Arguments:

job_id str - ID of the job to rehydrate.

Returns:

BatchJob - Job associated with the given ID.

BatchJob

class BatchJob()

init

Construct a BatchJob.

def __init__(client: "HumeBatchClient", job_id: str)

Arguments:

client HumeBatchClient - HumeBatchClient instance.
job_id str - Job ID.

get_status

Get the status of the job.

def get_status() -> BatchJobStatus

Returns:

BatchJobStatus - The status of the BatchJob.

get_predictions

Get BatchJob predictions.

def get_predictions() -> Any

Returns:

Any - Predictions for the BatchJob.

download_predictions

Download BatchJob predictions file.

def download_predictions(filepath: Union[str, Path]) -> None

Arguments:

filepath Union[str, Path] - Filepath where predictions will be downloaded.

download_artifacts

Download BatchJob artifacts zip file.

def download_artifacts(filepath: Union[str, Path]) -> None

Arguments:

filepath Optional[Union[str, Path]] - Filepath where artifacts will be downloaded.

get_details

Get details for the BatchJob.

def get_details() -> BatchJobDetails

Note that the details for a job may be fetched before the job has completed.
You may want to use job.await_complete() which will wait for the job to
reach a terminal state before returning.

Returns:

BatchJobDetails - Details for the BatchJob.

await_complete

Block until the job has reached a terminal status.

def await_complete(timeout: int = 300) -> BatchJobDetails

Arguments:

timeout int - Maximum time in seconds to await. If the timeout is reached
before the job reaches a terminal state the job will continue to be processed,
but a HumeClientException will be raised to the caller of await_complete.

Raises:

ValueError - If the timeout is not valid.

Returns:

BatchJobDetails - Details for the BatchJob.

repr

Get the string representation of the BatchJob.

def __repr__() -> str

Returns:

The the string representation of the BatchJob.

BatchJobStatus

class BatchJobStatus(Enum)

is_terminal

Check if a status is "terminal".

@classmethod
def is_terminal(cls, status: "BatchJobStatus") -> bool

Arguments:

status BatchJobStatus - Status to check.

Returns:

bool - Whether the status is "terminal".

from_str

Convert a status to a string.

@classmethod
def from_str(cls, status: str) -> "BatchJobStatus"

Arguments:

status str - Status to convert.

Returns:

BatchJobStatus - The enum variant for the given string.

BatchJobState

@dataclass
class BatchJobState()

Arguments:

status BatchJobStatus - Status of the batch job.
created_timestamp_ms Optional[int] - Time when job was created.
started_timestamp_ms Optional[int] - Time when job started.
ended_timestamp_ms Optional[int] - Time when job ended.

BatchJobDetails

class BatchJobDetails()

init

Construct a BatchJobDetails.

def __init__(*,
             configs: Dict[ModelType, ModelConfigBase],
             urls: List[str],
             files: List[str],
             state: BatchJobState,
             callback_url: Optional[str] = None,
             notify: bool = False)

Arguments:

configs Dict[ModelType, ModelConfigBase] - Configurations for the BatchJob.
urls List[str] - URLs processed in the BatchJob.
files List[str] - Files processed in the BatchJob.
state BatchJobState - State of BatchJob.
callback_url Optional[str] - A URL to which a POST request is sent upon job completion.
notify bool - Whether an email notification should be sent upon job completion.

from_response

Construct a BatchJobDetails from a batch API job response.

@classmethod
def from_response(cls, response: Any) -> "BatchJobDetails"

Arguments:

response Any - Batch API job response.

Returns:

BatchJobDetails - A BatchJobDetails based on a batch API job response.

get_status

Get the status of the job

def get_status() -> BatchJobStatus

Returns:

BatchJobStatus - The status of the BatchJob.

get_run_time_ms

Get the total time in milliseconds it took for the job to run if the job is in a terminal state

def get_run_time_ms() -> Optional[int]

Returns:

Optional[int] - Time in milliseconds it took for the job to run. If the job is not in a terminal
state then None is returned.

get_created_time

Get the time the job was created.

def get_created_time() -> Optional[datetime]

Returns:

Optional[datetime] - Datetime when the job was created. If the job has not started
then None is returned.

get_started_time

Get the time the job started running.

def get_started_time() -> Optional[datetime]

Returns:

Optional[datetime] - Datetime when the job started running. If the job has not started
then None is returned.

get_ended_time

Get the time the job stopped running if the job is in a terminal state.

def get_ended_time() -> Optional[datetime]

Returns:

Optional[datetime] - Datetime when the job started running. If the job is not in a terminal
state then None is returned.

TranscriptionConfig

Configuration for speech transcription.

@dataclass
class TranscriptionConfig(ConfigBase["TranscriptionConfig"])

Arguments:

language Optional[str] - By default, we use an automated language detection method for our
Speech Prosody, Language, and NER models. However, if you know what language is being spoken
in your media samples, you can specify it via its BCP-47 tag and potentially obtain more accurate results.
You can specify any of the following: zh, da, nl, en, en-AU,
en-IN, en-NZ, en-GB, fr, fr-CA, de, hi, hi-Latn, id, it, ja, ko, no,
pl, pt, pt-BR, pt-PT, ru, es, es-419, sv, ta, tr, or uk.