HyperChat™ Python Client

Installation

pip install hyperbee

Usage

import os
from hyperbee import HyperBee
 
client = HyperBee(
    # This is the default and can be omitted
    api_key=os.environ.get("HYPERBEE_API_KEY"),
)
 
chat_completion = client.chat.completions.create(
    messages=[
        {
            "role": "user",
            "content": "Say this is a test",
        }
    ],
    model="hyperchat",
)

While you can provide an api_key keyword argument, we recommend using python-dotenv (opens in a new tab) to add HYPERBEE_API_KEY="My API Key" to your .env file so that your API Key is not stored in source control.

The default API base_url is the Chat Completions API, i.e., https://api.hyperbee.ai/v1/. Make sure you add base_url="https://api-rag.hyperbee.ai/v1/" to your client initialization if you want to use the RAG API.

Async usage

Simply import AsyncHyperBee instead of HyperBee and use await with each API call:

import os, asyncio
from hyperbee import AsyncHyperBee
 
client = AsyncHyperBee(
    # This is the default and can be omitted
    api_key=os.environ.get("HYPERBEE_API_KEY"),
)
 
 
async def main() -> None:
    chat_completion = await client.chat.completions.create(
        messages=[
            {
                "role": "user",
                "content": "Say this is a test",
            }
        ],
        model="hyperchat",
    )
 
asyncio.run(main())

Functionality between the synchronous and asynchronous clients is otherwise identical.

Streaming Responses

We provide support for streaming responses using Server Side Events (SSE).

from hyperbee import HyperBee
 
client = HyperBee()
 
stream = client.chat.completions.create(
    model="hyperchat",
    messages=[{"role": "user", "content": "Say this is a test"}],
    stream=True,
)
for chunk in stream:
    print(chunk.choices[0].delta.content or "", end="")

The async client uses the exact same interface.

from hyperbee import AsyncHyperBee
 
client = AsyncHyperBee()
 
 
async def main():
    stream = await client.chat.completions.create(
        model="hyperchat",
        messages=[{"role": "user", "content": "Say this is a test"}],
        stream=True,
    )
    async for chunk in stream:
        print(chunk.choices[0].delta.content or "", end="")
 
 
asyncio.run(main())

Retrieval Augmented Generation

This documentation provides an overview of how to use the HyperBee Python package for managing and interacting with document collections and performing a RAG. The following sections will guide you through setting up the client, creating a namespace, uploading documents, asking questions, listing documents, deleting documents, and deleting a namespace.

Setting Up the Client

To begin using the HyperBee package, you need to initialize the client with your API key. Ensure that your API key is stored in an environment variable named HYPERBEE_API_KEY.

import os
from hyperbee import HyperBee
 
api_key = os.environ["HYPERBEE_API_KEY"]
 
### Initialize the HyperBee client
client = HyperBee(api_key=api_key)

Creating a Namespace

A namespace is created by uploading a document. This can be done using the create_namespace method and leaving the namespace field as blank. Here, we upload a document named Why AI is harder than we think.pdf to create a new namespace. The method returns a status variable indicating the completion of the propagating processes. Receiving a 'Ready' status means that namespace is ready to be used and receiving 'Pending' means that namespace is under construction and it should be waited for 'Ready' status.

file1 = "test_docs/Why AI is harder than we think.pdf"
namespace, status = client.chat.completions.create_namespace(
    [file1], sleepseconds=10, timeoutseconds=60, verbose=True
)
print(f"Namespace created: {namespace}")

Uploading Documents to an Existing Namespace

Once a namespace is created, additional documents can be uploaded to it using the add_to_collection method. Here, we upload opticalflow.pdf to the existing namespace by providing its id as parameter.

file2 = "test_docs/opticalflow.pdf"
client.chat.completions.add_to_collection(
    [file2], namespace, sleepseconds=10, timeoutseconds=60, verbose=True
)

Asking Questions

With documents uploaded to a namespace, you can ask questions related to the content of these documents. Use the create method to interact with the documents in your namespace.

messages = [
    "What is the definition of the optical flow?",
    "What are the reasons for AI being this hard to achieve?",
]
for message_content in messages:
    response = client.chat.completions.create(
        messages=[{"role": "user", "content": message_content}],
        model="hyperchat",
        namespace=namespace,
        stream=False,
    )
    print(f"Question: {message_content}")
    print("Answer:", response.choices[0].message.content)

Listing Documents

To retrieve a list of all documents in a namespace, use the get_remote_doclist method.

documents = client.chat.completions.get_remote_doclist(namespace)
print(f"Document list retrieved for namespace: {namespace}")
print(documents)

Deleting Documents

Documents can be removed from a namespace using the remove_from_collection method. Here, we delete raptor_paper.pdf and recurrent_neural_networks.pdf.

files_to_delete = ["raptor_paper.pdf", "recurrent_neural_networks.pdf"]
client.chat.completions.remove_from_collection(
    files_to_delete, namespace, sleepseconds=10, timeoutseconds=60, verbose=True
)

Deleting a Namespace

Finally, to delete an entire namespace along with all its documents, use the delete_namespace method.

client.chat.completions.delete_namespace(namespace)

Module-level client

We highly recommend instantiating client instances instead of relying on the global client.

We also expose a global client instance that is accessible in a similar fashion to versions prior to v1.

import hyperbee
 
# optional; defaults to `os.environ['HYPERBEE_API_KEY']`
hyperbee.api_key = '...'
 
# all client options can be configured just like the `HyperBee` instantiation counterpart
hyperbee.base_url = "https://..."
hyperbee.default_headers = {"x-foo": "true"}
 
completion = hyperbee.chat.completions.create(
    model="hyperchat",
    messages=[
        {
            "role": "user",
            "content": "How do I output all files in a directory using Python?",
        },
    ],
)
print(completion.choices[0].message.content)

The API is the exact same as the standard client instance based API.

This is intended to be used within REPLs or notebooks for faster iteration, not in application code.

We recommend that you always instantiate a client (e.g., with client = HyperBee()) in application code because:

• It can be difficult to reason about where client options are configured
• It's not possible to change certain client options without potentially causing race conditions
• It's harder to mock for testing purposes
• It's not possible to control cleanup of network connections

Using types

Nested request parameters are TypedDicts (opens in a new tab). Responses are Pydantic models (opens in a new tab), which provide helper methods for things like:

• Serializing back into JSON, model.model_dump_json(indent=2, exclude_unset=True)
• Converting to a dictionary, model.model_dump(exclude_unset=True)

Typed requests and responses provide autocomplete and documentation within your editor. If you would like to see type errors in VS Code to help catch bugs earlier, set python.analysis.typeCheckingMode to basic.

Nested params

Nested parameters are dictionaries, typed using TypedDict, for example:

from hyperbee import HyperBee
 
client = HyperBee()
 
completion = client.chat.completions.create(
    messages=[
        {
            "role": "user",
            "content": "Can you generate an example json object describing a fruit?",
        }
    ],
    model="hyperchat",
    response_format={"type": "json_object"},
)

Handling errors

When the library is unable to connect to the API (for example, due to network connection problems or a timeout), a subclass of hyperbee.APIConnectionError is raised.

When the API returns a non-success status code (that is, 4xx or 5xx response), a subclass of hyperbee.APIStatusError is raised, containing status_code and response properties.

All errors inherit from hyperbee.APIError.

Error codes are as followed:

Retries

Certain errors are automatically retried 2 times by default, with a short exponential backoff. Connection errors (for example, due to a network connectivity problem), 408 Request Timeout, 409 Conflict, 429 Rate Limit, and >=500 Internal errors are all retried by default.

You can use the max_retries option to configure or disable retry settings:

from hyperbee import HyperBee
 
# Configure the default for all requests:
client = HyperBee(
    # default is 2
    max_retries=0,
)
 
# Or, configure per-request:
client.with_options(max_retries=5).chat.completions.create(
    messages=[
        {
            "role": "user",
            "content": "How can I get the name of the current day in Node.js?",
        }
    ],
    model="hyperchat",
)

Timeouts

By default requests time out after 10 minutes. You can configure this with a timeout option, which accepts a float or an httpx.Timeout (opens in a new tab) object:

from hyperbee import HyperBee
 
# Configure the default for all requests:
client = HyperBee(
    # 20 seconds (default is 10 minutes)
    timeout=20.0,
)
 
# More granular control:
client = HyperBee(
    timeout=httpx.Timeout(60.0, read=5.0, write=10.0, connect=2.0),
)
 
# Override per-request:
client.with_options(timeout=5 * 1000).chat.completions.create(
    messages=[
        {
            "role": "user",
            "content": "How can I list all files in a directory using Python?",
        }
    ],
    model="hyperchat",
)

On timeout, an APITimeoutError is thrown.

Note that requests that time out are retried twice by default.

Advanced

Logging

We use the standard library logging (opens in a new tab) module.

You can enable logging by setting the environment variable HYPERBEE_LOG to debug.

$ export HYPERBEE_LOG=debug

How to tell whether None means null or missing

In an API response, a field may be explicitly null, or missing entirely; in either case, its value is None in this library. You can differentiate the two cases with .model_fields_set:

if response.my_field is None:
  if 'my_field' not in response.model_fields_set:
    print('Got json like {}, without a "my_field" key present at all.')
  else:
    print('Got json like {"my_field": null}.')

Configuring the HTTP client

You can directly override the httpx client (opens in a new tab) to customize it for your use case, including:

• Support for proxies
• Custom transports
• Additional advanced (opens in a new tab) functionality

import httpx
from hyperbee import HyperBee
 
client = HyperBee(
    # Or use the `HYPERBEE_BASE_URL` env var
    base_url="http://my.test.server.example.com:8083",
    http_client=httpx.Client(
        proxies="http://my.test.proxy.example.com",
        transport=httpx.HTTPTransport(local_address="0.0.0.0"),
    ),
)

Managing HTTP resources

By default the library closes underlying HTTP connections whenever the client is garbage collected (opens in a new tab). You can manually close the client using the .close() method if desired, or with a context manager that closes when exiting.

Requirements

Python 3.7 or higher.