Merge pull request #34 from myhloli/hybrid_flash

This commit is contained in:
Xiaomeng Zhao
2026-06-06 12:46:57 +08:00
committed by GitHub
38 changed files with 1040 additions and 278 deletions
+1 -1
View File
@@ -33,7 +33,7 @@ A WebUI developed based on Gradio, with a simple interface and only core parsing
<tr>
<th rowspan="2" style="text-align:center; vertical-align:middle;">Parsing Backend</th>
<th rowspan="2" style="text-align:center; vertical-align:middle;">pipeline</th>
<th colspan="2" style="text-align:center;">*-auto-engine</th>
<th colspan="2" style="text-align:center;">*-engine</th>
<th colspan="2" style="text-align:center;">*-http-client</th>
</tr>
<tr>
+2 -2
View File
@@ -12,8 +12,8 @@ Options:
-o, --output PATH Output directory (required)
--api-url TEXT MinerU FastAPI base URL; if omitted, `mineru` starts a temporary local `mineru-api`
-m, --method [auto|txt|ocr] Parsing method: auto (default), txt, ocr (pipeline and hybrid* backend only)
-b, --backend [pipeline|hybrid-auto-engine|hybrid-http-client|vlm-auto-engine|vlm-http-client]
Parsing backend (default: hybrid-auto-engine)
-b, --backend [pipeline|vlm-engine|hybrid-engine|hybrid-flash-engine|vlm-http-client|hybrid-http-client|hybrid-flash-http-client]
Parsing backend (default: hybrid-flash-engine)
-l, --lang [ch|ch_server|ch_lite|en|korean|japan|chinese_cht|ta|te|ka|th|el|latin|arabic|east_slavic|cyrillic|devanagari]
Specify document language (improves OCR accuracy, pipeline and hybrid* backend only)
-u, --url TEXT OpenAI-compatible backend URL passed through to the server when using http-client
+1 -1
View File
@@ -33,7 +33,7 @@
<tr>
<th rowspan="2" style="text-align:center; vertical-align:middle;">解析后端</th>
<th rowspan="2" style="text-align:center; vertical-align:middle;">pipeline</th>
<th colspan="2" style="text-align:center;">*-auto-engine</th>
<th colspan="2" style="text-align:center;">*-engine</th>
<th colspan="2" style="text-align:center;">*-http-client</th>
</tr>
<tr>
+3 -3
View File
@@ -116,7 +116,7 @@ docker run -u root --name mineru_docker --privileged=true \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
<td>🟢</td>
</tr>
@@ -132,7 +132,7 @@ docker run -u root --name mineru_docker --privileged=true \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
<td>🟢</td>
</tr>
@@ -148,7 +148,7 @@ docker run -u root --name mineru_docker --privileged=true \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
<td>🟢</td>
</tr>
+3 -3
View File
@@ -57,7 +57,7 @@ docker run -it --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
</tr>
<tr>
@@ -70,7 +70,7 @@ docker run -it --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
</tr>
<tr>
@@ -83,7 +83,7 @@ docker run -it --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
</tr>
<tr>
@@ -95,7 +95,7 @@ docker run --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟡</td>
<td>🟡</td>
</tr>
@@ -111,7 +111,7 @@ docker run --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🔴</td>
<td>🟢</td>
</tr>
@@ -127,7 +127,7 @@ docker run --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🔴</td>
<td>🟢</td>
</tr>
+3 -3
View File
@@ -55,7 +55,7 @@ docker run -u root --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
</tr>
<tr>
@@ -68,7 +68,7 @@ docker run -u root --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
</tr>
<tr>
@@ -81,7 +81,7 @@ docker run -u root --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
</tr>
<tr>
+3 -3
View File
@@ -62,7 +62,7 @@ docker run -u root --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
</tr>
<tr>
@@ -75,7 +75,7 @@ docker run -u root --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
</tr>
<tr>
@@ -88,7 +88,7 @@ docker run -u root --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
</tr>
<tr>
@@ -69,7 +69,7 @@ docker run --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
</tr>
<tr>
@@ -82,7 +82,7 @@ docker run --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
</tr>
<tr>
@@ -95,7 +95,7 @@ docker run --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
</tr>
<tr>
@@ -69,7 +69,7 @@ docker run -u root --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
</tr>
<tr>
@@ -82,7 +82,7 @@ docker run -u root --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
</tr>
<tr>
@@ -95,7 +95,7 @@ docker run -u root --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
</tr>
<tr>
+3 -3
View File
@@ -88,7 +88,7 @@ docker run --ipc host \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
<td>🟢</td>
</tr>
@@ -104,7 +104,7 @@ docker run --ipc host \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
<td>🟢</td>
</tr>
@@ -120,7 +120,7 @@ docker run --ipc host \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
<td>🟢</td>
</tr>
@@ -63,7 +63,7 @@ docker run -u root --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
</tr>
<tr>
@@ -76,7 +76,7 @@ docker run -u root --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🔴</td>
</tr>
<tr>
@@ -89,7 +89,7 @@ docker run -u root --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🔴</td>
</tr>
<tr>
+3 -3
View File
@@ -79,7 +79,7 @@ docker run --privileged=true \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
<td>🟢</td>
</tr>
@@ -95,7 +95,7 @@ docker run --privileged=true \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
<td>🟢</td>
</tr>
@@ -111,7 +111,7 @@ docker run --privileged=true \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
<td>🟢</td>
</tr>
@@ -65,7 +65,7 @@ docker run -dit --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
</tr>
<tr>
@@ -78,7 +78,7 @@ docker run -dit --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
</tr>
<tr>
@@ -91,7 +91,7 @@ docker run -dit --name mineru_docker \
<td>🟢</td>
</tr>
<tr>
<td>&lt;vlm/hybrid&gt;-auto-engine</td>
<td>&lt;vlm/hybrid&gt;-engine</td>
<td>🟢</td>
</tr>
<tr>
+11 -11
View File
@@ -72,7 +72,7 @@
## 4. MinerU功能
> [!NOTE]
> - `VastAI`加速卡仅支持使用`vlm-auto-engine`和`vlm-http-client`形式进行`VLM`模型推理加速
> - `VastAI`加速卡仅支持使用`vlm-engine`和`vlm-http-client`形式进行`VLM`模型推理加速
- 进入容器
```bash
@@ -83,15 +83,15 @@
- 模型准备,参考官方介绍:[model_source.md](https://github.com/opendatalab/MinerU/blob/master/docs/zh/usage/model_source.md)
- 方式一:`vlm-auto-engine`
- 方式一:`vlm-engine`
```bash
export MINERU_MODEL_SOURCE=modelscope
# step1, 以`vlm-auto-engine`方式启动MinerU解析任务
# step1, 以`vlm-engine`方式启动MinerU解析任务
mineru -p image.png \
-o ./output \
-b vlm-auto-engine \
-b vlm-engine \
--http-timeout 1200 \
--tensor-parallel-size 2 \
--enforce_eager \
@@ -147,11 +147,11 @@
<td>🔴</td>
</tr>
<tr>
<td>hybrid-auto-engine</td>
<td>hybrid-engine</td>
<td>🔴</td>
</tr>
<tr>
<td>vlm-auto-engine</td>
<td>vlm-engine</td>
<td>🟢</td>
</tr>
<tr>
@@ -168,11 +168,11 @@
<td>🔴</td>
</tr>
<tr>
<td>hybrid-auto-engine</td>
<td>hybrid-engine</td>
<td>🔴</td>
</tr>
<tr>
<td>vlm-auto-engine</td>
<td>vlm-engine</td>
<td>🟢</td>
</tr>
<tr>
@@ -189,11 +189,11 @@
<td>🔴</td>
</tr>
<tr>
<td>hybrid-auto-engine</td>
<td>hybrid-engine</td>
<td>🔴</td>
</tr>
<tr>
<td>vlm-auto-engine</td>
<td>vlm-engine</td>
<td>🟢</td>
</tr>
<tr>
@@ -212,4 +212,4 @@
> - 🟢: 支持,运行较稳定,精度与NVIDIA GPU基本一致
> - 🟡: 支持但较不稳定,在某些场景下可能出现异常,或精度存在一定差异
> - 🔴: 不支持,无法运行,或精度存在较大差异
> - `vlm-auto-engine`VastAI仅支持vLLM后端
> - `vlm-engine`VastAI仅支持vLLM后端
+2 -2
View File
@@ -12,8 +12,8 @@ Options:
-o, --output PATH 输出目录(必填)
--api-url TEXT MinerU FastAPI 服务地址;不传时自动拉起本地临时 mineru-api
-m, --method [auto|txt|ocr] 解析方法:auto(默认)、txt、ocr(仅用于 pipeline 与 hybrid* 后端)
-b, --backend [pipeline|hybrid-auto-engine|hybrid-http-client|vlm-auto-engine|vlm-http-client]
解析后端(默认为 hybrid-auto-engine
-b, --backend [pipeline|vlm-engine|hybrid-engine|hybrid-flash-engine|vlm-http-client|hybrid-http-client|hybrid-flash-http-client]
解析后端(默认为 hybrid-flash-engine
-l, --lang [ch|ch_server|ch_lite|en|korean|japan|chinese_cht|ta|te|ka|th|el|latin|arabic|east_slavic|cyrillic|devanagari]
指定文档语言(可提升 OCR 准确率,仅用于 pipeline 与 hybrid* 后端)
-u, --url TEXT 当使用 http-client 时,传给服务端后端的 OpenAI 兼容地址