remove python3.8 support and add python3.13 support

Author: lvhan028

.github/workflows/cuda11.8-whl-release.yml

@@ -13,7 +13,7 @@ jobs:
   linux-build:
     strategy:
       matrix:
-        pyver: [py38, py39, py310, py311, py312]
+        pyver: [py39, py310, py311, py312, py313]
     runs-on: ubuntu-latest
     env:
       PYTHON_VERSION: ${{ matrix.pyver }}
@@ -56,7 +56,7 @@ jobs:
   windows-build:
     strategy:
       matrix:
-        pyver: ['3.8', '3.9', '3.10', '3.11', '3.12']
+        pyver: ['3.9', '3.10', '3.11', '3.12', '3.13']
     runs-on: windows-latest
     steps:
       - name: Checkout repository

.github/workflows/pypi.yml

@@ -13,7 +13,7 @@ jobs:
   linux-build:
     strategy:
       matrix:
-        pyver: [py38, py39, py310, py311, py312]
+        pyver: [py39, py310, py311, py312, py313]
     runs-on: ubuntu-latest
     env:
       PYTHON_VERSION: ${{ matrix.pyver }}
@@ -55,7 +55,7 @@ jobs:
   windows-build:
     strategy:
       matrix:
-        pyver: ['3.8', '3.9', '3.10', '3.11', '3.12']
+        pyver: ['3.9', '3.10', '3.11', '3.12', '3.13']
     runs-on: windows-latest
     steps:
       - name: Checkout repository
@@ -112,10 +112,10 @@ jobs:
           merge-multiple: true
       - name: Display artifacts
         run: ls artifact/ -lh
-      - name: Set up python3.8
+      - name: Set up python 3.10
         uses: actions/setup-python@v4
         with:
-          python-version: '3.8'
+          python-version: '3.10'
       - name: Upload to pypi
         run: |
           pip install twine

.github/workflows/windows-x64-gpu.yml

@@ -37,7 +37,7 @@ jobs:
       - name: Set up python
         uses: actions/setup-python@v4
         with:
-          python-version: '3.8'
+          python-version: '3.10'
       - name: Install python packages
         run: |
           pip install -r requirements/build.txt

README.md

@@ -187,10 +187,10 @@ They differ in the types of supported models and the inference data type. Please
 
 ## Installation
 
-It is recommended installing lmdeploy using pip in a conda environment (python 3.8 - 3.12):
+It is recommended installing lmdeploy using pip in a conda environment (python 3.9 - 3.13):
 
 ```shell
-conda create -n lmdeploy python=3.8 -y
+conda create -n lmdeploy python=3.10 -y
 conda activate lmdeploy
 pip install lmdeploy
 ```

README_ja.md

@@ -185,10 +185,10 @@ LMDeployは、[TurboMind](./docs/en/inference/turbomind.md)および[PyTorch](./
 
 ## インストール
 
-クリーンなconda環境（Python 3.8 - 3.12）でlmdeployをインストールすることをお勧めします。
+クリーンなconda環境（Python 3.9 - 3.12）でlmdeployをインストールすることをお勧めします。
 
 ```shell
-conda create -n lmdeploy python=3.8 -y
+conda create -n lmdeploy python=3.10 -y
 conda activate lmdeploy
 pip install lmdeploy
 ```

README_zh-CN.md

@@ -189,10 +189,10 @@ LMDeploy 支持 2 种推理引擎： [TurboMind](./docs/zh_cn/inference/turbomin
 
 ## 安装
 
-我们推荐在一个干净的conda环境下（python3.8 - 3.12），安装 lmdeploy：
+我们推荐在一个干净的conda环境下（python3.9 - 3.12），安装 lmdeploy：
 
 ```shell
-conda create -n lmdeploy python=3.8 -y
+conda create -n lmdeploy python=3.10 -y
 conda activate lmdeploy
 pip install lmdeploy
 ```

builder/manywheel/Dockerfile_2014

@@ -26,16 +26,16 @@ FROM base as conda
 ADD manywheel/scripts/install_conda.sh install_conda.sh
 RUN bash ./install_conda.sh && rm install_conda.sh
 RUN
-RUN /opt/conda/bin/conda create -n py38 python=3.8 -yq && \
-    /opt/conda/envs/py38/bin/pip install pybind11 && \
-    /opt/conda/bin/conda create -n py39 python=3.9 -yq && \
+RUN /opt/conda/bin/conda create -n py39 python=3.9 -yq && \
     /opt/conda/envs/py39/bin/pip install pybind11 && \
     /opt/conda/bin/conda create -n py310 python=3.10 -yq && \
     /opt/conda/envs/py310/bin/pip install pybind11 && \
     /opt/conda/bin/conda create -n py311 python=3.11 -yq && \
     /opt/conda/envs/py311/bin/pip install pybind11 && \
     /opt/conda/bin/conda create -n py312 python=3.12 -yq && \
     /opt/conda/envs/py312/bin/pip install pybind11
+    /opt/conda/bin/conda create -n py313 python=3.13 -yq && \
+    /opt/conda/envs/py313/bin/pip install pybind11
 
 FROM base as mpi
 ADD manywheel/scripts/install_openmpi.sh install_openmpi.sh

builder/manywheel/build_all_wheel.sh

@@ -10,7 +10,7 @@ PLAT_NAME=manylinux2014_x86_64
 for cuver in ${CUDA_VER}; do
     DOCKER_TAG=cuda${cuver}
     OUTPUT_FOLDER=cuda${cuver}_dist
-    for pyver in py38 py39 py310 py311 py312; do
+    for pyver in py39 py310 py311 py312 py313; do
         bash ${TOPDIR}/manywheel/build_wheel.sh ${pyver} ${PLAT_NAME} ${DOCKER_TAG} ${OUTPUT_FOLDER} \
             |& tee ${PLAT_NAME}.${pyver}.cuda${cuver}.log.txt
     done

docs/en/get_started/installation.md

@@ -12,10 +12,10 @@ It supports LLMs and VLMs deployment on both Linux and Windows platform, with mi
 
 ## Install with pip (Recommend)
 
-It is recommended installing lmdeploy using pip in a conda environment (python 3.8 - 3.12):
+It is recommended installing lmdeploy using pip in a conda environment (python 3.9 - 3.13):
 
 ```shell
-conda create -n lmdeploy python=3.8 -y
+conda create -n lmdeploy python=3.10 -y
 conda activate lmdeploy
 pip install lmdeploy
 ```

docs/zh_cn/get_started/installation.md

@@ -12,10 +12,10 @@ Ada Lovelace(sm89): 40 系列
 
 ## 使用 pip 安装（推荐）
 
-我们推荐在一个干净的conda环境下（python3.8 - 3.12），安装 lmdeploy：
+我们推荐在一个干净的conda环境下（python3.9 - 3.13），安装 lmdeploy：
 
 ```shell
-conda create -n lmdeploy python=3.8 -y
+conda create -n lmdeploy python=3.10 -y
 conda activate lmdeploy
 pip install lmdeploy
 ```

setup.py

@@ -150,11 +150,11 @@ def gen_packages_items():
         },
         has_ext_modules=check_ext_modules,
         classifiers=[
-            'Programming Language :: Python :: 3.8',
             'Programming Language :: Python :: 3.9',
             'Programming Language :: Python :: 3.10',
             'Programming Language :: Python :: 3.11',
             'Programming Language :: Python :: 3.12',
+            'Programming Language :: Python :: 3.13',
             'Intended Audience :: Developers',
             'Intended Audience :: Education',
             'Intended Audience :: Science/Research',