diff --git a/.dockerignore b/.dockerignore
index 1d319f0..ddd8b94 100644
--- a/.dockerignore
+++ b/.dockerignore
@@ -54,6 +54,4 @@ examples/
 assets/
 docs/
 tests/
-README.md
 README.docker.md
-LICENSE
diff --git a/Dockerfile b/Dockerfile
index d5d2ef8..9268cdf 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -1,5 +1,5 @@
-# Use NVIDIA CUDA base image with Python 3.10 (CUDA 12.8 for RTX 5080)
-FROM nvidia/cuda:12.8.0-cudnn-runtime-ubuntu22.04
+# Use NVIDIA CUDA base image with Python 3.12 (CUDA 12.8 for RTX 5080)
+FROM nvidia/cuda:12.8.0-base-ubuntu24.04
 
 # Set environment variables
 ENV DEBIAN_FRONTEND=noninteractive \
@@ -8,27 +8,33 @@ ENV DEBIAN_FRONTEND=noninteractive \
     PIP_NO_CACHE_DIR=1 \
     CUDA_VISIBLE_DEVICES=0
 
-# Install Python 3.10 and system dependencies
+# Configure apt to use Tsinghua mirror (清华源)
+RUN sed -i 's@//archive.ubuntu.com@//mirrors.tuna.tsinghua.edu.cn@g' /etc/apt/sources.list.d/ubuntu.sources && \
+    sed -i 's@//security.ubuntu.com@//mirrors.tuna.tsinghua.edu.cn@g' /etc/apt/sources.list.d/ubuntu.sources
+
+# Install Python and system dependencies (Ubuntu 24.04 uses Python 3.12)
 RUN apt-get update && apt-get install -y \
-    python3.10 \
-    python3.10-dev \
+    python3 \
     python3-pip \
+    python3-venv \
     git \
     libglib2.0-0 \
     libsm6 \
     libxext6 \
     libxrender-dev \
     libgomp1 \
-    libgl1-mesa-glx \
     wget \
     && rm -rf /var/lib/apt/lists/*
 
-# Set Python 3.10 as default
-RUN update-alternatives --install /usr/bin/python3 python3 /usr/bin/python3.10 1 && \
-    update-alternatives --install /usr/bin/python python /usr/bin/python3.10 1
+# Create symlink for python command
+RUN ln -sf /usr/bin/python3 /usr/bin/python
 
-# Upgrade pip
-RUN python3 -m pip install --upgrade pip setuptools wheel
+# Configure pip to use Tsinghua mirror (清华源) and allow system-wide installs
+RUN python3 -m pip config set global.index-url https://pypi.tuna.tsinghua.edu.cn/simple && \
+    python3 -m pip config set global.break-system-packages true
+
+# Upgrade pip (ignore system-installed packages)
+RUN pip install --upgrade --ignore-installed pip setuptools wheel
 
 # Set working directory
 WORKDIR /app
@@ -37,9 +43,12 @@ WORKDIR /app
 COPY . /app/
 
 # Install PyTorch with CUDA support first (cu124 is compatible with CUDA 12.8)
-RUN pip install torch==2.6.0 torchvision==0.21.0 --index-url https://download.pytorch.org/whl/cu124
+# Note: PyTorch uses official mirror as Tsinghua doesn't host CUDA builds
+RUN pip install torch torchvision
 
 # Install the package and dependencies
+# Set version manually since .git is excluded by .dockerignore
+ENV SETUPTOOLS_SCM_PRETEND_VERSION=1.0.0
 RUN pip install -e .
 
 # Install additional dependencies for server
@@ -56,5 +65,5 @@ HEALTHCHECK --interval=30s --timeout=10s --start-period=5s --retries=3 \
     CMD python3 -c "import requests; requests.get('http://localhost:8001/', timeout=5)" || exit 1
 
 # Default command to start the server (port 8001)
-CMD ["texteller", "launch", "server", "-p", "8001"]
+CMD ["texteller", "launch", "-p", "8001"]
 
diff --git a/deploy.sh b/deploy.sh
index 5e05e4c..2f72476 100755
--- a/deploy.sh
+++ b/deploy.sh
@@ -59,7 +59,7 @@ check_docker() {
 # Check if NVIDIA Container Toolkit is installed
 check_nvidia_docker() {
     print_info "Checking NVIDIA Container Toolkit..."
-    if ! docker run --rm --gpus all nvidia/cuda:12.1.0-base-ubuntu22.04 nvidia-smi &> /dev/null; then
+    if ! docker run --rm --gpus all nvidia/cuda:12.8.0-base-ubuntu24.04 nvidia-smi &> /dev/null; then
         print_error "NVIDIA Container Toolkit not working properly."
         print_info "Please install it with:"
         echo "  sudo apt-get install -y nvidia-container-toolkit"
@@ -113,8 +113,12 @@ start_container() {
         --name $CONTAINER_NAME \
         --gpus '"device=0"' \
         -p $PORT:8001 \
-        -v "$MODEL_PATH:/root/.cache/huggingface/hub/models--OleehyO--TexTeller:ro" \
+        --shm-size=2g \
+        -v "$HOME/.cache/huggingface:/root/.cache/huggingface:ro" \
         -e CUDA_VISIBLE_DEVICES=0 \
+        -e HF_HOME=/root/.cache/huggingface \
+        -e HF_HUB_OFFLINE=1 \
+        -e TRANSFORMERS_OFFLINE=1 \
         -e RAY_NUM_REPLICAS=1 \
         -e RAY_NCPU_PER_REPLICA=4 \
         -e RAY_NGPU_PER_REPLICA=1 \