support python 3.12 (#1605)

* use cudaStreamSynchronize(stream) * update runtime * update build docker * update script
InternLM · May 16, 2024 · a31639e · a31639e
1 parent ec1a39e
commit a31639e
Show file tree

Hide file tree

Showing 7 changed files with 10 additions and 10 deletions.
diff --git a/.github/workflows/cuda11.8-whl-release.yml b/.github/workflows/cuda11.8-whl-release.yml
@@ -13,7 +13,7 @@ jobs:
   linux-build:
     strategy:
       matrix:
-        pyver: [py38, py39, py310, py311]
+        pyver: [py38, py39, py310, py311, py312]
     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']
+        pyver: ['3.8', '3.9', '3.10', '3.11', '3.12']
     runs-on: windows-latest
     steps:
       - name: Checkout repository

diff --git a/.github/workflows/pypi.yml b/.github/workflows/pypi.yml
@@ -13,7 +13,7 @@ jobs:
   linux-build:
     strategy:
       matrix:
-        pyver: [py38, py39, py310, py311]
+        pyver: [py38, py39, py310, py311, py312]
     runs-on: ubuntu-latest
     env:
       PYTHON_VERSION: ${{ matrix.pyver }}
@@ -54,7 +54,7 @@ jobs:
   windows-build:
     strategy:
       matrix:
-        pyver: ['3.8', '3.9', '3.10', '3.11']
+        pyver: ['3.8', '3.9', '3.10', '3.11', '3.12']
     runs-on: windows-latest
     steps:
       - name: Checkout repository

diff --git a/builder/manywheel/Dockerfile_2014 b/builder/manywheel/Dockerfile_2014
@@ -33,7 +33,9 @@ RUN /opt/conda/bin/conda create -n py38 python=3.8 -yq && \
     /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/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
 
 FROM base as mpi
 ADD manywheel/scripts/install_openmpi.sh install_openmpi.sh

diff --git a/builder/manywheel/build_all_wheel.sh b/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; do
+    for pyver in py38 py39 py310 py311 py312; do
         bash ${TOPDIR}/manywheel/build_wheel.sh ${pyver} ${PLAT_NAME} ${DOCKER_TAG} ${OUTPUT_FOLDER} \
             |& tee ${PLAT_NAME}.${pyver}.cuda${cuver}.log.txt
     done

diff --git a/builder/manywheel/entrypoint_build.sh b/builder/manywheel/entrypoint_build.sh
@@ -11,6 +11,7 @@ source /opt/conda/bin/activate
 conda activate $PYTHON_VERSION
 
 cd lmdeploy
+rm -rf lmdeploy/lib
 mkdir -p build && cd build && rm -rf *
 bash ../generate.sh make
 make -j$(nproc) && make install

diff --git a/requirements/runtime.txt b/requirements/runtime.txt
@@ -13,6 +13,7 @@ sentencepiece
 shortuuid
 tiktoken
 torch<=2.2.2,>=2.0.0
+torchvision<=0.17.2,>=0.15.0
 transformers
 triton>=2.1.0,<=2.2.0; sys_platform == "linux"
 uvicorn
diff --git a/src/turbomind/kernels/stop_criteria_kernels.cu b/src/turbomind/kernels/stop_criteria_kernels.cu
@@ -151,11 +151,7 @@ void invokeLengthCriterion(bool*           finished,
 
     length_criterion<<<grid, block, 0, stream>>>(
         finished, should_stop, h_pinned_finished_sum_, sequence_limit_length, batch_size, beam_width, step);
-#ifdef _MSC_VER
     cudaStreamSynchronize(stream);
-#else
-    while (((volatile int*)h_pinned_finished_sum_)[0] == -1) {};
-#endif
     sync_check_cuda_error();
 
     *should_stop = h_pinned_finished_sum_[0] == batch_size * beam_width;