runpod-workers / worker-blip

✍️ | RunPod Worker for bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation

BLIP | Worker

This worker was derived from the Bliper-Serverless project by kodxana

Bliper-Serverless

This API provides an image captioning service that takes an image or a zip file containing multiple images and generates a caption for each image. The captions are saved into text files named after the input images and packed into a zip file for download.

API Parameters

The API accepts the following parameters:

• data_url (required, string): The URL of the image or zip file containing multiple images to be captioned. The supported image formats are JPEG and PNG.

• max_length (optional, integer, default: 75): The maximum length of the generated captions. It should be a positive integer.

• min_length (optional, integer, default: 5): The minimum length of the generated captions. It should be a positive integer.

API Output

The API returns a JSON object containing the following fields:

• captions_zip_url: The URL of the zip file containing the generated captions. Each caption is saved in a text file named after the input image (e.g., input_image.jpg will have a corresponding caption file named input_image.txt).