diff --git a/.github/workflows/build-docker.yml b/.github/workflows/build-docker.yml
index 28f208d..edcbaec 100644
--- a/.github/workflows/build-docker.yml
+++ b/.github/workflows/build-docker.yml
@@ -26,24 +26,29 @@ env:
   DOCKER_MIRROR_REGISTRY_USERNAME: ${{ vars.DOCKER_MIRROR_REGISTRY_USERNAME }}
   DOCKER_MIRROR_REGISTRY_PASSWORD: ${{ secrets.DOCKER_MIRROR_REGISTRY_PASSWORD }}
 
+
 jobs:
-  qpod_OpenCV:
-    name: 'opencv'
+  qpod_PaddleOCR_cuda112:
+    name: 'paddleocr-cuda112'
     runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@v4
-      - run: source ./tool.sh && build_image opencv latest docker_OpenCV/Dockerfile && push_image
-  
-  # paddleocr-models: temp disable the build caused by the paddle run on CPU server
-  qpod_PaddleOCR_cuda112:
-    name: 'paddleocr-cuda112'
+      - run: |
+          source ./tool.sh && free_diskspace
+          build_image paddleocr-cuda112 latest docker_PaddleOCR/paddle-ocr.Dockerfile --build-arg "BASE_IMG=cuda_11.2"
+          echo 'paddleocr-models: temp disable the build caused by the paddle run on CPU server'
+          echo 'build_image paddleocr-models  latest docker_PaddleOCR/paddle-ocr-models.Dockerfile --build-arg "BASE_IMG=paddleocr-cuda112" --build-arg "BASE_NAMESPACE_SRC=docker.io/library"'
+          push_image
+
+  qpod_PaddleOCR_cuda120:
+    name: 'paddleocr-cuda120,doc-ai-cuda120'
     runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@v4
       - run: |
           source ./tool.sh && free_diskspace
-          build_image paddleocr-cuda112 latest docker_PaddleOCR/Dockerfile --build-arg "BASE_IMG=cuda_11.2"
-          echo 'build_image paddleocr-models  latest docker_PaddleOCR/models.Dockerfile --build-arg "BASE_IMG=paddleocr-cuda112" --build-arg "BASE_NAMESPACE_SRC=docker.io/library"'
+          build_image paddleocr-cuda120 latest docker_PaddleOCR/paddle-ocr.Dockerfile --build-arg "BASE_IMG=cuda_12.0"
+          build_image doc-ai-cuda120    latest docker_PaddleOCR/paddle-ocr.Dockerfile --build-arg "BASE_IMG=py-nlp-cuda120"
           push_image
 
 
@@ -57,29 +62,29 @@ jobs:
           build_image vllm-cuda latest docker_vllm/vllm-cuda.Dockerfile
           push_image
 
-
-  qpod_PaddleOCR_cuda120:
-    name: 'paddleocr-cuda120,doc-ai-cuda120'
+  qpod_aidoc-miner:
+    name: 'aidoc-miner'
     runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@v4
       - run: |
           source ./tool.sh && free_diskspace
-          build_image paddleocr-cuda120 latest docker_PaddleOCR/Dockerfile --build-arg "BASE_IMG=cuda_12.0"
-          build_image doc-ai-cuda120    latest docker_PaddleOCR/Dockerfile --build-arg "BASE_IMG=py-nlp-cuda120"
+          build_image aidoc-miner latest docker_aidoc/miner.Dockerfile --build-arg "BASE_IMG=py-nlp-cuda128"
           push_image
 
 
   qpod_OpenFace:
-    name: 'openface-src,openface'
-    # needs: [ "qpod_OpenCV" ]
+    name: 'opencv,openface-src,openface'
     runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@v4
       - run: |
-          source ./tool.sh
-          export IMG_PREFIX_SRC="docker.io/library" && build_image openface-src latest docker_OpenFace/OpenFace-src.Dockerfile && push_image src
-          export IMG_PREFIX_SRC="${IMG_PREFIX_DST}" && build_image openface     latest docker_OpenFace/OpenFace.Dockerfile     && push_image openface
+          source ./tool.sh &&
+          export IMG_PREFIX_SRC="docker.io/library"
+          build_image openface-src latest docker_OpenFace/OpenFace-src.Dockerfile && push_image openface-src
+          export IMG_PREFIX_SRC="${IMG_PREFIX_DST}"
+          build_image opencv       latest docker_OpenFace/OpenCV.Dockerfile       && push_image opencv
+          build_image openface     latest docker_OpenFace/OpenFace.Dockerfile     && push_image openface
 
   # To build HF model image for a single model, simple run: `build_image_hf_model bert-base-cased`
   qpod_HuggingFaceModels:
@@ -90,16 +95,17 @@ jobs:
       - env:
           HF_MODEL_NAME: ${{ vars.HF_MODEL_NAME }}
         run: |
-          source tool.sh && export IMG_PREFIX_SRC="docker.io/library"
-          source docker_HuggingFace-model/script-setup-huggingface.sh
-          export -f download_hf_model build_image_hf_model build_image_no_tag push_image
+          source tool.sh && export -f build_image_no_tag push_image free_diskspace && free_diskspace
+          source docker_HuggingFace-model/script-setup-huggingface.sh && export -f download_hf_model build_image_hf_model
           HF_MODEL_RANDOM=$(sort --random-sort docker_HuggingFace-model/list_hf_models.txt | head -n1)
-          download_hf_model ${HF_MODEL_NAME:-$HF_MODEL_RANDOM} && build_image_hf_model ${HF_MODEL_NAME}
+          download_hf_model ${HF_MODEL_NAME:-$HF_MODEL_RANDOM}
+          export IMG_PREFIX_SRC="docker.io/library"
+          build_image_hf_model ${HF_MODEL_NAME}
 
 
   ## Sync all images in this build (listed by "names") to mirror registry.
   sync_images:
-    needs: ["qpod_OpenCV", "qpod_HuggingFaceModels", "qpod_PaddleOCR_cuda112", "qpod_PaddleOCR_cuda120", "qpod_vllm-cuda"]
+    needs: ["qpod_OpenFace", "qpod_HuggingFaceModels", "qpod_aidoc-miner", "qpod_PaddleOCR_cuda112", "qpod_PaddleOCR_cuda120", "qpod_vllm-cuda"]
     runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@v4
diff --git a/docker_HuggingFace-model/list_hf_models.txt b/docker_HuggingFace-model/list_hf_models.txt
index 657c5ff..62657af 100644
--- a/docker_HuggingFace-model/list_hf_models.txt
+++ b/docker_HuggingFace-model/list_hf_models.txt
@@ -4,3 +4,4 @@ BAAI/bge-m3
 BAAI/bge-reranker-v2-m3
 Alibaba-NLP/gte-modernbert-base
 Alibaba-NLP/gte-reranker-modernbert-base
+Alibaba-NLP/gte-Qwen2-1.5B-instruct
diff --git a/docker_OpenCV/Dockerfile b/docker_OpenFace/OpenCV.Dockerfile
similarity index 100%
rename from docker_OpenCV/Dockerfile
rename to docker_OpenFace/OpenCV.Dockerfile
diff --git a/docker_OpenCV/work/install_list_OpenCV.apt b/docker_OpenFace/work/install_list_OpenCV.apt
similarity index 100%
rename from docker_OpenCV/work/install_list_OpenCV.apt
rename to docker_OpenFace/work/install_list_OpenCV.apt
diff --git a/docker_PaddleOCR/models.Dockerfile b/docker_PaddleOCR/paddle-ocr-models.Dockerfile
similarity index 100%
rename from docker_PaddleOCR/models.Dockerfile
rename to docker_PaddleOCR/paddle-ocr-models.Dockerfile
diff --git a/docker_PaddleOCR/Dockerfile b/docker_PaddleOCR/paddle-ocr.Dockerfile
similarity index 100%
rename from docker_PaddleOCR/Dockerfile
rename to docker_PaddleOCR/paddle-ocr.Dockerfile
diff --git a/docker_aidoc/miner.Dockerfile b/docker_aidoc/miner.Dockerfile
new file mode 100644
index 0000000..c2c8815
--- /dev/null
+++ b/docker_aidoc/miner.Dockerfile
@@ -0,0 +1,25 @@
+# Distributed under the terms of the Modified BSD License.
+
+# reference: https://github.com/opendatalab/MinerU/blob/master/docker/china/Dockerfile
+
+ARG BASE_NAMESPACE
+ARG BASE_IMG="py-nlp-cuda128"
+FROM ${BASE_NAMESPACE:+$BASE_NAMESPACE/}${BASE_IMG}
+
+LABEL maintainer="haobibo@gmail.com"
+
+RUN set -eux \
+ # ----------
+ && apt-get update \
+ && mkdir -pv /usr/share/man/man1 \
+ && apt-get -qq install -yq --no-install-recommends openjdk-21-jre-headless \
+ && apt-get -qq install -yq --no-install-recommends \
+     libgl1 libglib2.0-0 libxrender1 libsm6 libxext6 \
+     fontconfig ttf-mscorefonts-installer \
+     fonts-noto-cjk fonts-wqy-zenhei fonts-wqy-microhei \
+     libreoffice poppler-utils \
+ && pip install -U magic-pdf[full] modelscope \
+ # ----------
+ && rm -rf /var/lib/apt/lists/* \
+ && source /opt/utils/script-setup.sh \
+ && install__clean && list_installed_packages
diff --git a/docker_vllm/vllm-cuda.Dockerfile b/docker_vllm/vllm-cuda.Dockerfile
index 94ed620..01dca10 100644
--- a/docker_vllm/vllm-cuda.Dockerfile
+++ b/docker_vllm/vllm-cuda.Dockerfile
@@ -1,7 +1,7 @@
 # Distributed under the terms of the Modified BSD License.
 
 ARG BASE_NAMESPACE
-ARG BASE_IMG="cuda_12.4"
+ARG BASE_IMG="torch-cuda126"
 FROM ${BASE_NAMESPACE:+$BASE_NAMESPACE/}${BASE_IMG}
 
 LABEL maintainer="haobibo@gmail.com"