Update docker (#456)

c69b88f5 · Yang Yong (雍洋) · GitHub · 87625ec2 · c69b88f5 · c69b88f5
Unverified Commit c69b88f5 authored Nov 09, 2025 by Yang Yong (雍洋) Committed by GitHub Nov 09, 2025
5 changed files
--- a/Dockerfile
+++ b/Dockerfile
-FROM pytorch/pytorch:2.8.0-cuda12.8-cudnn9-devel AS base
+FROM nvidia/cuda:12.8.1-cudnn-devel-ubuntu24.04 AS base

 WORKDIR /app

@@ -7,24 +7,35 @@ ENV LANG=C.UTF-8
 ENV LC_ALL=C.UTF-8
 ENV LD_LIBRARY_PATH=/usr/local/lib/x86_64-linux-gnu:$LD_LIBRARY_PATH

-RUN apt-get update && apt-get install -y vim tmux zip unzip wget git git-lfs build-essential libibverbs-dev ca-certificates \
+RUN apt-get update && apt-get install -y vim tmux zip unzip bzip2 wget git git-lfs build-essential libibverbs-dev ca-certificates \
    curl iproute2 libsm6 libxext6 kmod ccache libnuma-dev libssl-dev flex bison libgtk-3-dev libpango1.0-dev \
-    libsoup2.4-dev libnice-dev libopus-dev libvpx-dev libx264-dev libsrtp2-dev libglib2.0-dev libdrm-dev\
+    libsoup2.4-dev libnice-dev libopus-dev libvpx-dev libx264-dev libsrtp2-dev libglib2.0-dev libdrm-dev libjpeg-dev libpng-dev \
    && apt-get clean && rm -rf /var/lib/apt/lists/* && git lfs install

+# install miniconda with Python 3.12
+RUN wget https://repo.anaconda.com/miniconda/Miniconda3-py312_25.9.1-1-Linux-x86_64.sh -O /tmp/miniconda.sh && \
+    bash /tmp/miniconda.sh -b -p /app/miniconda && \
+    rm /tmp/miniconda.sh
+ENV PATH=/app/miniconda/bin:$PATH
+RUN conda tos accept --override-channels --channel https://repo.anaconda.com/pkgs/main
+RUN conda tos accept --override-channels --channel https://repo.anaconda.com/pkgs/r
+
+RUN conda install conda-forge::ffmpeg=8.0.0 -y && conda clean -all -y
+
+# install torch
+RUN pip install --no-cache-dir torch==2.8.0 torchvision==0.23.0 torchaudio==2.8.0 --index-url https://download.pytorch.org/whl/cu128
+
 RUN pip install --no-cache-dir packaging ninja cmake scikit-build-core uv meson ruff pre-commit fastapi uvicorn requests -U

 RUN git clone https://github.com/vllm-project/vllm.git && cd vllm \
-    && python use_existing_torch.py && pip install -r requirements/build.txt \
+    && python use_existing_torch.py && pip install --no-cache-dir -r requirements/build.txt \
    && pip install --no-cache-dir --no-build-isolation -v -e .

 RUN git clone https://github.com/sgl-project/sglang.git && cd sglang/sgl-kernel \
    && make build && make clean

 RUN pip install --no-cache-dir diffusers transformers tokenizers accelerate safetensors opencv-python numpy imageio \
-    imageio-ffmpeg einops loguru qtorch ftfy av decord
-
-RUN conda install conda-forge::ffmpeg=8.0.0 -y && ln -s /opt/conda/bin/ffmpeg /usr/bin/ffmpeg && conda clean -all -y
+    imageio-ffmpeg einops loguru qtorch ftfy av decord matplotlib debugpy

 RUN git clone https://github.com/Dao-AILab/flash-attention.git --recursive

@@ -32,13 +43,31 @@ RUN cd flash-attention && python setup.py install && rm -rf build

 RUN cd flash-attention/hopper && python setup.py install && rm -rf build

-RUN git clone https://github.com/ModelTC/SageAttention.git
+RUN git clone https://github.com/ModelTC/SageAttention-1104.git --depth 1

-RUN cd SageAttention && CUDA_ARCHITECTURES="8.0,8.6,8.9,9.0,12.0" EXT_PARALLEL=4 NVCC_APPEND_FLAGS="--threads 8" MAX_JOBS=32 pip install --no-cache-dir -v -e .
+RUN cd SageAttention-1104 && TORCH_CUDA_ARCH_LIST="8.0,8.6,8.9,9.0,12.0" EXT_PARALLEL=4 NVCC_APPEND_FLAGS="--threads 8" MAX_JOBS=32 python setup.py install && rm -rf build

-RUN git clone https://github.com/KONAKONA666/q8_kernels.git
+RUN cd SageAttention-1104/sageattention3_blackwell && python setup.py install && rm -rf build

-RUN cd q8_kernels && git submodule init && git submodule update && python setup.py install && rm -rf build
+RUN git clone https://github.com/SandAI-org/MagiAttention.git --recursive
+
+RUN cd MagiAttention && TORCH_CUDA_ARCH_LIST="9.0" pip install --no-cache-dir --no-build-isolation -v -e .
+
+RUN git clone https://github.com/ModelTC/FlashVSR.git --depth 1
+
+RUN cd FlashVSR && pip install --no-cache-dir -v -e .
+
+COPY lightx2v_kernel /app/lightx2v_kernel
+
+RUN git clone https://github.com/NVIDIA/cutlass.git --depth 1 && cd /app/lightx2v_kernel && MAX_JOBS=32 && CMAKE_BUILD_PARALLEL_LEVEL=4 \
+    uv build --wheel \
+        -Cbuild-dir=build . \
+        -Ccmake.define.CUTLASS_PATH=/app/cutlass \
+        --verbose \
+        --color=always \
+        --no-build-isolation \
+    && pip install dist/*whl --force-reinstall --no-deps \
+    && rm -rf /app/lightx2v_kernel && rm -rf /app/cutlass

 # cloud deploy
 RUN pip install --no-cache-dir aio-pika asyncpg>=0.27.0 aioboto3>=12.0.0 PyJWT alibabacloud_dypnsapi20170525==1.2.2 redis==6.4.0 tos
@@ -47,7 +76,7 @@ RUN curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh -s -- -y --de
 ENV PATH=/root/.cargo/bin:$PATH

 RUN cd /opt \
-    && wget https://mirrors.tuna.tsinghua.edu.cn/gnu//libiconv/libiconv-1.15.tar.gz \
+    && wget https://mirrors.tuna.tsinghua.edu.cn/gnu/libiconv/libiconv-1.15.tar.gz \
    && tar zxvf libiconv-1.15.tar.gz \
    && cd libiconv-1.15 \
    && ./configure \
@@ -56,7 +85,7 @@ RUN cd /opt \
    && rm -rf /opt/libiconv-1.15

 RUN cd /opt \
-    && git clone https://github.com/GStreamer/gstreamer.git -b 1.24.12 --depth 1 \
+    && git clone https://github.com/GStreamer/gstreamer.git -b 1.27.2 --depth 1 \
    && cd gstreamer \
    && meson setup builddir \
    && meson compile -C builddir \
@@ -65,7 +94,7 @@ RUN cd /opt \
    && rm -rf /opt/gstreamer

 RUN cd /opt \
-    && git clone https://github.com/GStreamer/gst-plugins-rs.git -b gstreamer-1.24.12 --depth 1 \
+    && git clone https://github.com/GStreamer/gst-plugins-rs.git -b gstreamer-1.27.2 --depth 1 \
    && cd gst-plugins-rs \
    && cargo build --package gst-plugin-webrtchttp --release \
    && install -m 644 target/release/libgstwebrtchttp.so $(pkg-config --variable=pluginsdir gstreamer-1.0)/ \
@@ -73,4 +102,13 @@ RUN cd /opt \

 RUN ldconfig

+
+# q8f for base docker
+RUN git clone https://github.com/KONAKONA666/q8_kernels.git --depth 1
+RUN cd q8_kernels && git submodule init && git submodule update && python setup.py install && rm -rf build
+
+# q8f for 5090 docker
+# RUN git clone https://github.com/ModelTC/LTX-Video-Q8-Kernels.git --depth 1
+# RUN cd LTX-Video-Q8-Kernels && git submodule init && git submodule update && python setup.py install && rm -rf build
+
 WORKDIR /workspace
--- a/Dockerfile_5090
+++ b/Dockerfile_5090
+FROM nvidia/cuda:12.8.1-cudnn-devel-ubuntu24.04 AS base
+
+WORKDIR /app
+
+ENV DEBIAN_FRONTEND=noninteractive
+ENV LANG=C.UTF-8
+ENV LC_ALL=C.UTF-8
+ENV LD_LIBRARY_PATH=/usr/local/lib/x86_64-linux-gnu:$LD_LIBRARY_PATH
+
+RUN apt-get update && apt-get install -y vim tmux zip unzip bzip2 wget git git-lfs build-essential libibverbs-dev ca-certificates \
+    curl iproute2 libsm6 libxext6 kmod ccache libnuma-dev libssl-dev flex bison libgtk-3-dev libpango1.0-dev \
+    libsoup2.4-dev libnice-dev libopus-dev libvpx-dev libx264-dev libsrtp2-dev libglib2.0-dev libdrm-dev libjpeg-dev libpng-dev \
+    && apt-get clean && rm -rf /var/lib/apt/lists/* && git lfs install
+
+# install miniconda with Python 3.12
+RUN wget https://repo.anaconda.com/miniconda/Miniconda3-py312_25.9.1-1-Linux-x86_64.sh -O /tmp/miniconda.sh && \
+    bash /tmp/miniconda.sh -b -p /app/miniconda && \
+    rm /tmp/miniconda.sh
+ENV PATH=/app/miniconda/bin:$PATH
+RUN conda tos accept --override-channels --channel https://repo.anaconda.com/pkgs/main
+RUN conda tos accept --override-channels --channel https://repo.anaconda.com/pkgs/r
+
+RUN conda install conda-forge::ffmpeg=8.0.0 -y && conda clean -all -y
+
+# install torch
+RUN pip install --no-cache-dir torch==2.8.0 torchvision==0.23.0 torchaudio==2.8.0 --index-url https://download.pytorch.org/whl/cu128
+
+RUN pip install --no-cache-dir packaging ninja cmake scikit-build-core uv meson ruff pre-commit fastapi uvicorn requests -U
+
+RUN git clone https://github.com/vllm-project/vllm.git && cd vllm \
+    && python use_existing_torch.py && pip install --no-cache-dir -r requirements/build.txt \
+    && pip install --no-cache-dir --no-build-isolation -v -e .
+
+RUN git clone https://github.com/sgl-project/sglang.git && cd sglang/sgl-kernel \
+    && make build && make clean
+
+RUN pip install --no-cache-dir diffusers transformers tokenizers accelerate safetensors opencv-python numpy imageio \
+    imageio-ffmpeg einops loguru qtorch ftfy av decord matplotlib debugpy
+
+RUN git clone https://github.com/Dao-AILab/flash-attention.git --recursive
+
+RUN cd flash-attention && python setup.py install && rm -rf build
+
+RUN cd flash-attention/hopper && python setup.py install && rm -rf build
+
+RUN git clone https://github.com/ModelTC/SageAttention-1104.git --depth 1
+
+RUN cd SageAttention-1104 && TORCH_CUDA_ARCH_LIST="8.0,8.6,8.9,9.0,12.0" EXT_PARALLEL=4 NVCC_APPEND_FLAGS="--threads 8" MAX_JOBS=32 python setup.py install && rm -rf build
+
+RUN cd SageAttention-1104/sageattention3_blackwell && python setup.py install && rm -rf build
+
+RUN git clone https://github.com/SandAI-org/MagiAttention.git --recursive
+
+RUN cd MagiAttention && TORCH_CUDA_ARCH_LIST="9.0" pip install --no-cache-dir --no-build-isolation -v -e .
+
+RUN git clone https://github.com/ModelTC/FlashVSR.git --depth 1
+
+RUN cd FlashVSR && pip install --no-cache-dir -v -e .
+
+COPY lightx2v_kernel /app/lightx2v_kernel
+
+RUN git clone https://github.com/NVIDIA/cutlass.git --depth 1 && cd /app/lightx2v_kernel && MAX_JOBS=32 && CMAKE_BUILD_PARALLEL_LEVEL=4 \
+    uv build --wheel \
+        -Cbuild-dir=build . \
+        -Ccmake.define.CUTLASS_PATH=/app/cutlass \
+        --verbose \
+        --color=always \
+        --no-build-isolation \
+    && pip install dist/*whl --force-reinstall --no-deps \
+    && rm -rf /app/lightx2v_kernel && rm -rf /app/cutlass
+
+# cloud deploy
+RUN pip install --no-cache-dir aio-pika asyncpg>=0.27.0 aioboto3>=12.0.0 PyJWT alibabacloud_dypnsapi20170525==1.2.2 redis==6.4.0 tos
+
+RUN curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh -s -- -y --default-toolchain stable
+ENV PATH=/root/.cargo/bin:$PATH
+
+RUN cd /opt \
+    && wget https://mirrors.tuna.tsinghua.edu.cn/gnu/libiconv/libiconv-1.15.tar.gz \
+    && tar zxvf libiconv-1.15.tar.gz \
+    && cd libiconv-1.15 \
+    && ./configure \
+    && make \
+    && make install \
+    && rm -rf /opt/libiconv-1.15
+
+RUN cd /opt \
+    && git clone https://github.com/GStreamer/gstreamer.git -b 1.27.2 --depth 1 \
+    && cd gstreamer \
+    && meson setup builddir \
+    && meson compile -C builddir \
+    && meson install -C builddir \
+    && ldconfig \
+    && rm -rf /opt/gstreamer
+
+RUN cd /opt \
+    && git clone https://github.com/GStreamer/gst-plugins-rs.git -b gstreamer-1.27.2 --depth 1 \
+    && cd gst-plugins-rs \
+    && cargo build --package gst-plugin-webrtchttp --release \
+    && install -m 644 target/release/libgstwebrtchttp.so $(pkg-config --variable=pluginsdir gstreamer-1.0)/ \
+    && rm -rf /opt/gst-plugins-rs
+
+RUN ldconfig
+
+
+# q8f for base docker
+# RUN git clone https://github.com/KONAKONA666/q8_kernels.git --depth 1
+# RUN cd q8_kernels && git submodule init && git submodule update && python setup.py install && rm -rf build
+
+# q8f for 5090 docker
+RUN git clone https://github.com/ModelTC/LTX-Video-Q8-Kernels.git --depth 1
+RUN cd LTX-Video-Q8-Kernels && git submodule init && git submodule update && python setup.py install && rm -rf build
+
+WORKDIR /workspace
--- a/Dockerfile_deploy
+++ b/Dockerfile_deploy
@@ -5,7 +5,7 @@ RUN cd /opt/lightx2v/deploy/server/frontend \
    && npm install \
    && npm run build

-FROM lightx2v/lightx2v:25101501-cu128 AS base
+FROM lightx2v/lightx2v:25110701-cu128 AS base

 RUN mkdir /workspace/LightX2V
 WORKDIR /workspace/LightX2V

--- a/docs/EN/source/getting_started/quickstart.md
+++ b/docs/EN/source/getting_started/quickstart.md
@@ -27,10 +27,10 @@ We strongly recommend using the Docker environment, which is the simplest and fa

 #### 1. Pull Image

-Visit LightX2V's [Docker Hub](https://hub.docker.com/r/lightx2v/lightx2v/tags), select a tag with the latest date, such as `25101501-cu128`:
+Visit LightX2V's [Docker Hub](https://hub.docker.com/r/lightx2v/lightx2v/tags), select a tag with the latest date, such as `25110701-cu128`:

 ```bash
-docker pull lightx2v/lightx2v:25101501-cu128
+docker pull lightx2v/lightx2v:25110701-cu128
 ```

 We recommend using the `cuda128` environment for faster inference speed. If you need to use the `cuda124` environment, you can use image versions with the `-cu124` suffix:
@@ -51,7 +51,7 @@ For mainland China, if the network is unstable when pulling images, you can pull

 ```bash
 # cuda128
-docker pull registry.cn-hangzhou.aliyuncs.com/yongyang/lightx2v:25101501-cu128
+docker pull registry.cn-hangzhou.aliyuncs.com/yongyang/lightx2v:25110701-cu128

 # cuda124
 docker pull registry.cn-hangzhou.aliyuncs.com/yongyang/lightx2v:25101501-cu124

--- a/docs/ZH_CN/source/getting_started/quickstart.md
+++ b/docs/ZH_CN/source/getting_started/quickstart.md
@@ -27,10 +27,10 @@

 #### 1. 拉取镜像

-访问 LightX2V 的 [Docker Hub](https://hub.docker.com/r/lightx2v/lightx2v/tags)，选择一个最新日期的 tag，比如 `25101501-cu128`：
+访问 LightX2V 的 [Docker Hub](https://hub.docker.com/r/lightx2v/lightx2v/tags)，选择一个最新日期的 tag，比如 `25110701-cu128`：

 ```bash
-docker pull lightx2v/lightx2v:25101501-cu128
+docker pull lightx2v/lightx2v:25110701-cu128
 ```

 我们推荐使用`cuda128`环境，以获得更快的推理速度，若需要使用`cuda124`环境，可以使用带`-cu124`后缀的镜像版本：
@@ -51,7 +51,7 @@ docker run --gpus all -itd --ipc=host --name [容器名] -v [挂载设置] --ent

 ```bash
 # cuda128
-docker pull registry.cn-hangzhou.aliyuncs.com/yongyang/lightx2v:25101501-cu128
+docker pull registry.cn-hangzhou.aliyuncs.com/yongyang/lightx2v:25110701-cu128

 # cuda124
 docker pull registry.cn-hangzhou.aliyuncs.com/yongyang/lightx2v:25101501-cu124