Data Endpoint

Datasets, VLM inference, and batch evaluation

The Data API manages datasets and runs VLM inference and evaluation over them.

import asyncio
from eyepop import EyePopSdk

async def main():
    async with EyePopSdk.dataEndpoint(is_async=True) as endpoint:
        datasets = await endpoint.list_datasets()
        print(datasets)

asyncio.run(main())

Some Data API calls need an account: set EYEPOP_ACCOUNT_ID. See Configuration.

infer_asset and evaluate_dataset are marked experimental in the SDK and may change.

VLM inference on one asset

import asyncio

from eyepop.data.data_types import InferRequest, TranscodeMode

async def main():
    async with EyePopSdk.dataEndpoint(is_async=True) as endpoint:
        job = await endpoint.infer_asset(
            asset_uuid="your-asset-uuid",
            infer_request=InferRequest(text_prompt="Describe this image."),
            transcode_mode=TranscodeMode.image_cover_1024,
        )
        while result := await job.predict():
            print(result)

asyncio.run(main())

Batch dataset evaluation

import asyncio

from eyepop.data.data_types import EvaluateRequest, InferRequest

request = EvaluateRequest(
    dataset_uuid="your-dataset-uuid",
    infer=InferRequest(text_prompt="How many people are in this image?"),
)

async def main():
    async with EyePopSdk.dataEndpoint(is_async=True, job_queue_length=4) as endpoint:
        job = await endpoint.evaluate_dataset(evaluate_request=request)
        response = await job.response
        print(response.model_dump_json(indent=2))

asyncio.run(main())

Next steps

• Composable Pops — chain models into an inference pipeline for the worker endpoint

• Running Inference — process media directly instead