AI Briefingd1-3B: Image and text classification and judgment with 3B parameters
LiquidAI/d1-3B
About the project
A lightweight multimodal model with 3B parameters that accepts both image and text inputs to perform classification and judgment tasks. It is designed to operate efficiently on edge devices or in resource-constrained environments by fine-tuning LFM2.5-VL-3B.
It features built-in System One-style fast decision-making and calibration capabilities, making it suitable for scenarios requiring simple classification or confidence evaluation. It supports 16 languages, including Korean, English, and Japanese, and can be easily integrated via the Transformers library.
It is useful for applications where low-latency responses and accurate classification are more important than large-scale inference. It also supports serving via vLLM, enabling rapid deployment in an OpenAI-compatible API format.
LiquidAI/d1-3B
The original page has no description.
image-text-to-text
This introduction was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report errors, attribution issues, or removal requests via Contact.