Browse Source

Merge pull request #3534 from opendatalab/release-2.5.2

Release 2.5.2
Xiaomeng Zhao 2 months ago
parent
commit
100ace2e99
3 changed files with 4 additions and 4 deletions
  1. 1 1
      README.md
  2. 2 2
      README_zh-CN.md
  3. 1 1
      mineru/backend/vlm/vlm_middle_json_mkcontent.py

+ 1 - 1
README.md

@@ -44,7 +44,7 @@
 
 
 # Changelog
 # Changelog
 
 
-- 2025/09/19 2.5.1 Released
+- 2025/09/19 2.5.2 Released
 
 
   We are officially releasing MinerU2.5, currently the most powerful multimodal large model for document parsing.
   We are officially releasing MinerU2.5, currently the most powerful multimodal large model for document parsing.
   With only 1.2B parameters, MinerU2.5's accuracy on the OmniDocBench benchmark comprehensively surpasses top-tier multimodal models like Gemini 2.5 Pro, GPT-4o, and Qwen2.5-VL-72B. It also significantly outperforms leading specialized models such as dots.ocr, MonkeyOCR, and PP-StructureV3.
   With only 1.2B parameters, MinerU2.5's accuracy on the OmniDocBench benchmark comprehensively surpasses top-tier multimodal models like Gemini 2.5 Pro, GPT-4o, and Qwen2.5-VL-72B. It also significantly outperforms leading specialized models such as dots.ocr, MonkeyOCR, and PP-StructureV3.

+ 2 - 2
README_zh-CN.md

@@ -44,9 +44,9 @@
 
 
 # 更新记录
 # 更新记录
 
 
-- 2025/09/19 2.5.1 发布
+- 2025/09/19 2.5.2 发布
   我们正式发布 MinerU2.5,当前最强文档解析多模态大模型。仅凭 1.2B 参数,MinerU2.5 在 OmniDocBench 文档解析评测中,精度已全面超越 Gemini2.5-Pro、GPT-4o、Qwen2.5-VL-72B等顶级多模态大模型,并显著领先于主流文档解析专用模型(如 dots.ocr, MonkeyOCR, PP-StructureV3 等)。
   我们正式发布 MinerU2.5,当前最强文档解析多模态大模型。仅凭 1.2B 参数,MinerU2.5 在 OmniDocBench 文档解析评测中,精度已全面超越 Gemini2.5-Pro、GPT-4o、Qwen2.5-VL-72B等顶级多模态大模型,并显著领先于主流文档解析专用模型(如 dots.ocr, MonkeyOCR, PP-StructureV3 等)。
-  模型已发布至[HuggingFace](https://huggingface.co/opendatalab/MinerU2.5-2509-1.2B)和[ModelScope](https://huggingface.co/opendatalab/MinerU2.5-2509-1.2B)平台,欢迎大家下载使用!
+  模型已发布至[HuggingFace](https://huggingface.co/opendatalab/MinerU2.5-2509-1.2B)和[ModelScope](https://modelscope.cn/models/opendatalab/MinerU2.5-2509-1.2B)平台,欢迎大家下载使用!
   - 核心亮点
   - 核心亮点
     - 极致能效,性能SOTA: 以 1.2B 的轻量化规模,实现了超越百亿乃至千亿级模型的SOTA性能,重新定义了文档解析的能效比。
     - 极致能效,性能SOTA: 以 1.2B 的轻量化规模,实现了超越百亿乃至千亿级模型的SOTA性能,重新定义了文档解析的能效比。
     - 先进架构,全面领先: 通过 “两阶段推理” (解耦布局分析与内容识别) 与 原生高分辨率架构 的结合,在布局分析、文本识别、公式识别、表格识别及阅读顺序五大方面均达到 SOTA 水平。
     - 先进架构,全面领先: 通过 “两阶段推理” (解耦布局分析与内容识别) 与 原生高分辨率架构 的结合,在布局分析、文本识别、公式识别、表格识别及阅读顺序五大方面均达到 SOTA 水平。

+ 1 - 1
mineru/backend/vlm/vlm_middle_json_mkcontent.py

@@ -54,7 +54,7 @@ def mk_blocks_to_markdown(para_blocks, make_mode, formula_enable, table_enable,
         elif para_type == BlockType.LIST:
         elif para_type == BlockType.LIST:
             for block in para_block['blocks']:
             for block in para_block['blocks']:
                 item_text = merge_para_with_text(block, formula_enable=formula_enable, img_buket_path=img_buket_path)
                 item_text = merge_para_with_text(block, formula_enable=formula_enable, img_buket_path=img_buket_path)
-                para_text += f"{item_text}\n"
+                para_text += f"{item_text}  \n"
         elif para_type == BlockType.TITLE:
         elif para_type == BlockType.TITLE:
             title_level = get_title_level(para_block)
             title_level = get_title_level(para_block)
             para_text = f'{"#" * title_level} {merge_para_with_text(para_block)}'
             para_text = f'{"#" * title_level} {merge_para_with_text(para_block)}'