Memory leak during DINO training.

[lolikonloli]

Divice info

---

sys.platform linux Python 3.10.0 (default, Mar 3 2022, 09:58:08) [GCC 7.5.0] numpy 1.22.4 detectron2 0.6 @/home/lolikonloli/code/detection/package/detrex/detectron2/detectron2 Compiler GCC 11.4 CUDA compiler CUDA 11.8 detectron2 arch flags 7.5 DETECTRON2_ENV_MODULE PyTorch 2.0.1+cu118 @/home/lolikonloli/anaconda3/envs/pl_det/lib/python3.10/site-packages/torch PyTorch debug build False GPU available Yes GPU 0,1 NVIDIA GeForce RTX 2080 Ti (arch=7.5) Driver version 535.104.05 CUDA_HOME /usr/local/cuda-11.8 Pillow 9.3.0 torchvision 0.15.2+cu118 @/home/lolikonloli/anaconda3/envs/pl_det/lib/python3.10/site-packages/torchvision torchvision arch flags 3.5, 5.0, 6.0, 7.0, 7.5, 8.0, 8.6 fvcore 0.1.5.post20221221 iopath 0.1.9 cv2 4.8.0

PyTorch built with:

• GCC 9.3
• C++ Version: 201703
• Intel(R) oneAPI Math Kernel Library Version 2022.2-Product Build 20220804 for Intel(R) 64 architecture applications
• Intel(R) MKL-DNN v2.7.3 (Git Hash 6dbeffbae1f23cbbeae17adb7b5b13f1f37c080e)
• OpenMP 201511 (a.k.a. OpenMP 4.5)
• LAPACK is enabled (usually provided by MKL)
• NNPACK is enabled
• CPU capability usage: AVX2
• CUDA Runtime 11.8
• NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86;-gencode;arch=compute_90,code=sm_90
• CuDNN 8.7
• Magma 2.6.1
• Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.8, CUDNN_VERSION=8.7.0, CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++, CXX_FLAGS= -D_GLIBCXX_USE_CXX11_ABI=0 -fabi-version=11 -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOROCTRACER -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -O2 -fPIC -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Werror=bool-operation -Wnarrowing -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wunused-local-typedefs -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Werror=cast-function-type -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_DISABLE_GPU_ASSERTS=ON, TORCH_VERSION=2.0.1, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=1, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF

  describe

  Memory continuously increases during DINO training with two 2080ti GPUs until it gets killed by the system.

[rentainhe]

Hello, it's a normal, because of the multi-scale training and denoising query, the model's memory usage is not that stable, it may takes about more than 12GB of 2080Ti, you can try to use fp16 training or lower the total_batch_size to skip this issue, or you can try to add activation checkpoint to reduce the memory usage of the total model