whylogs-container-client

A client library for accessing FastAPI.

See the WhyLabs doc site for full documentation, or see the API endpoint documentation for information on a specific API. The structure there mimics the module structure in the generated client.

Usage

First, create a client:

from whylogs_container_client import AuthenticatedClient

# Create an authenticated client for a container running on localhost
# The token field should match the password that you set on your whylogs container deployment.
client = AuthenticatedClient(base_url="http://localhost:8000", token="password", prefix="", auth_header_name="X-API-Key")

from whylogs_container_client import Client

# Can use a regular Client if the container has no password set
client = Client(base_url="http://localhost:8000")

APIs

Things to know:

1. Every API has four ways of calling it.

   1. sync: Blocking request that returns parsed data (if successful) or None
   2. sync_detailed: Blocking request that always returns a Request, optionally with parsed set if the request was successful.
   3. asyncio: Like sync but async instead of blocking
   4. asyncio_detailed: Like sync_detailed but async instead of blocking

2. APIs are grouped by their "tags" as Python modules.

3. APIs that do not have a tag are in whylogs_container_client.api.default

Here are some example requests for common APIs.

Log Data

from datetime import datetime

import whylogs_container_client.api.profile.log as Log
from whylogs_container_client.models import LogMultiple, LogRequest

# Get current time in epoch milliseconds using datetime
time_ms = int(datetime.now().timestamp() * 1000)

data = LogRequest(
    dataset_id="model-141",
    timestamp=time_ms,
    multiple=LogMultiple(
        columns=["custom_metric_1", "custom_metric_2"],
        data=[[1, 2], [3, 4]],
    ),
)

response = Log.sync_detailed(client=client, body=data)
if response.status_code != 200:
    raise Exception(f"Failed to log data. Status code: {response.status_code}")

Validate LLM

from whylogs_container_client.models.evaluation_result import EvaluationResult
from whylogs_container_client.models.llm_validate_request import LLMValidateRequest
from whylogs_container_client.models.validation_result import ValidationResult

request = LLMValidateRequest(
    prompt="?",
    response="I'm sorry you feel that way.",
    dataset_id="model-139",
    id="myid",
)

response = Evaluate.sync_detailed(client=client, body=request)

if not isinstance(response.parsed, EvaluationResult):
    raise Exception(f"Failed to validate data. Status code: {response.status_code}. {response.parsed}")

result: ValidationResult = response.parsed.validation_results

Health check

import whylogs_container_client.api.manage.health as Health

Health.sync_detailed(client=client)

Get Status

import whylogs_container_client.api.manage.status as Status
from whylogs_container_client.models.status_response import StatusResponse

response = Status.sync_detailed(client=client)

if response.parsed is None:
    raise Exception("Unexpected response type")

result: StatusResponse = response.parsed

Certificates

You can customize or disable the certificate verification.

# Example of using a custom certificate bundle
client.verify_ssl = "/path/to/certificate_bundle.pem"

# Adding event hooks to the httpx client
def log_request(request):
    print(f"Request event hook: {request.method} {request.url} - Waiting for response")

def log_response(response):
    request = response.request
    print(f"Response event hook: {request.method} {request.url} - Status {response.status_code}")

client.httpx_args = {"event_hooks": {"request": [log_request], "response": [log_response]}}

Advanced customizations

You can set the httpx client directly, but beware that this will override any existing settings (e.g., base_url):

import httpx
from whylogs_container_client import Client

client = Client(
    base_url="https://api.example.com",
)
# Note that base_url needs to be re-set, as would any shared cookies, headers, etc.
client.set_httpx_client(httpx.Client(base_url="https://api.example.com", proxies="http://localhost:8030"))