Skip to content

Latest commit

 

History

History
272 lines (187 loc) · 7.9 KB

README.md

File metadata and controls

272 lines (187 loc) · 7.9 KB
Raw

Hume AI Python SDK

Integrate Hume APIs directly into your Python application



Migration Guide for Version 0.7.0 and Above

We've released version 0.7.0 of the SDK with significant architectural changes. This update introduces AsyncHumeClient and HumeClient, improves type safety and async support, and provides more granular configuration options. To help you transition, we've prepared a comprehensive migration guide:

View the Migration Guide

Please review this guide before updating, as it covers breaking changes and provides examples for updating your code. Legacy functionality is preserved for backward compatibility. If you have any questions, please open an issue or contact our support team.

Documentation

API reference documentation is available here.

Compatibility

The Hume Python SDK is compatible across several Python versions and operating systems.

  • For the Empathic Voice Interface, Python versions 3.9 through 3.11 are supported on macOS and Linux.

  • For Expression Measurement, Python versions 3.9 through 3.12 are supported on macOS, Linux, and Windows.

Below is a table which shows the version and operating system compatibilities by product:

Python Version Operating System
Empathic Voice Interface 3.9, 3.10, 3.11 macOS, Linux
Expression Measurement 3.9, 3.10, 3.11, 3.12 macOS, Linux, Windows

Installation

pip install hume
# or
poetry add hume

Other Resources

from hume.client import HumeClient

client = HumeClient(
    api_key="YOUR_API_KEY", # Defaults to HUME_API_KEY
)
client.empathic_voice.configs.list_configs()

Async Client

The SDK also exports an async client so that you can make non-blocking calls to our API.

import asyncio

from hume.client import AsyncHumeClient

client = AsyncHumeClient(
    api_key="YOUR_API_KEY",
)

async def main() -> None:
    await client.empathic_voice.configs.list_configs()

asyncio.run(main())

Writing File

Writing files with an async stream of bytes can be tricky in Python! aiofiles can simplify this some. For example, you can download your job artifacts like so:

import aiofiles

from hume import AsyncHumeClient

client = AsyncHumeClient()
async with aiofiles.open('artifacts.zip', mode='wb') as file:
    async for chunk in client.expression_measurement.batch.get_job_artifacts(id="my-job-id"):
        await file.write(chunk)

Legacy SDK

If you want to continue using the legacy SDKs, simply import them from the hume.legacy module.

from hume.legacy import HumeVoiceClient, VoiceConfig

client = HumeVoiceClient("<your-api-key>")
config = client.empathic_voice.configs.get_config_version(
    id="id",
    version=1,
)

Namespaces

This SDK contains the APIs for expression measurement, empathic voice and custom models. Even if you do not plan on using more than one API to start, the SDK provides easy access in case you find additional APIs in the future.

Each API is namespaced accordingly:

from hume.client import HumeClient

client = HumeClient(
    api_key="YOUR_API_KEY",
)

client.expression_measurement. # APIs specific to Expression Measurement

client.emapthic_voice.         # APIs specific to Empathic Voice

Exception Handling

All errors thrown by the SDK will be subclasses of ApiError.

import hume.client

try:
  client.expression_measurement.batch.get_job_predictions(...)
except hume.core.ApiError as e: # Handle all errors
  print(e.status_code)
  print(e.body)

Pagination

Paginated requests will return a SyncPager or AsyncPager, which can be used as generators for the underlying object. For example, list_tools will return a generator over ReturnUserDefinedTool and handle the pagination behind the scenes:

import hume.client

client = HumeClient(
    api_key="YOUR_API_KEY",
)

for tool in client.empathic_voice.tools.list_tools():
  print(tool)

you could also iterate page-by-page:

for page in client.empathic_voice.tools.list_tools().iter_pages():
  print(page.items)

or manually:

pager = client.empathic_voice.tools.list_tools()
# First page
print(pager.items)
# Second page
pager = pager.next_page()
print(pager.items)

WebSockets

We expose a websocket client for interacting with the EVI API as well as Expression Measurement.

When interacting with these clients, you can use them very similarly to how you'd use the common websockets library:

from hume import StreamDataModels

client = AsyncHumeClient(api_key=os.getenv("HUME_API_KEY"))

async with client.expression_measurement.stream.connect(
    options={"config": StreamDataModels(...)}
) as hume_socket:
    print(await hume_socket.get_job_details())

The underlying connection, in this case hume_socket, will support intellisense/autocomplete for the different functions that are available on the socket!

Advanced

Retries

The Hume SDK is instrumented with automatic retries with exponential backoff. A request will be retried as long as the request is deemed retriable and the number of retry attempts has not grown larger than the configured retry limit.

A request is deemed retriable when any of the following HTTP status codes is returned:

  • 408 (Timeout)
  • 409 (Conflict)
  • 429 (Too Many Requests)
  • 5XX (Internal Server Errors)

Use the max_retries request option to configure this behavior.

from hume.client import HumeClient
from hume.core import RequestOptions

client = HumeClient(...)

# Override retries for a specific method
client.expression_measurement.batch.get_job_predictions(...,
    request_options=RequestOptions(max_retries=5)
)

Timeouts

By default, requests time out after 60 seconds. You can configure this with a timeout option at the client or request level.

from hume.client import HumeClient
from hume.core import RequestOptions

client = HumeClient(
    # All timeouts are 20 seconds
    timeout=20.0,
)

# Override timeout for a specific method
client.expression_measurement.batch.get_job_predictions(...,
    request_options=RequestOptions(timeout_in_seconds=20)
)

Custom HTTP client

You can override the httpx client to customize it for your use-case. Some common use-cases include support for proxies and transports.

import httpx

from hume.client import HumeClient

client = HumeClient(
    http_client=httpx.Client(
        proxies="http://my.test.proxy.example.com",
        transport=httpx.HTTPTransport(local_address="0.0.0.0"),
    ),
)

Contributing

While we value open-source contributions to this SDK, this library is generated programmatically.

Additions made directly to this library would have to be moved over to our generation code, otherwise they would be overwritten upon the next generated release. Feel free to open a PR as a proof of concept, but know that we will not be able to merge it as-is. We suggest opening an issue first to discuss with us!

On the other hand, contributions to the README are always very welcome!