d3

What is d3

vllm-sr/d3 is an open-source model repository based on the vLLM inference engine, focused on efficient deployment and inference of large language models.

Key features

• Supports text or JSON input and can attach multiple images and videos simultaneously
• A single call can handle multiple decision types: multiple-choice, yes/no, scoring, and returns the probability for each answer
• Outputs decision probabilities directly without generating text
• Supports reading multiple PNG, JPEG and WebP images as well as MP4, WebM, MOV and MKV videos
• Can serve as the 27B multimodal decision foundation model for vLLM Semantic Router

Who it is for

Developers who need to perform efficient multimodal decision inference, suitable for building selection, judgement or scoring applications.

How to get started

1. Visit Official Space: vllm-sr/Decision-3.0 for online trial

Compiled from public descriptions on the tool's official website / README / launch page. Pricing and user numbers have not been verified; please check the official website. Some sections were drafted with AI assistance. Written in Chinese and English; other languages are machine-translated.