公开知识图谱记录

NeoMME

记录摘要

Single bidirectional Transformer processes both text tokens and raw image patche

来自项目知识图谱的有来源实体；具体工具属性仍需补齐。

基准测试: both model sizes lie on ViDoRe v3 Pareto frontier for nDCG@10 and model size
能力: generates vector representations for input text and/or images
能力: handles visual document retrieval without OCR
能力: uses grouped-query attention, query-key normalization, gated attention, 2D rotary position embeddings, and squared-ReLU MLPs
部署方式: cloud/local
集成: Hugging Face Transformers
集成: Transformers
语言: multilingual

知识图谱当前记录了 14 个已知资料维度，另有 4 个维度尚未记录，关联 2 篇公开资料。这些数字只描述资料覆盖，不代表模型性能或适配性已经验证。