Open
Description
❓ Question
I follow this tutorial to install Torch-TensorRT, but in the last step:
cuda_version=$(nvcc --version | grep Cuda | grep release | cut -d ',' -f 2 | sed -e 's/ release //g')
export TORCH_INSTALL_PATH=$(python -c "import torch, os; print(os.path.dirname(torch.__file__))")
export SITE_PACKAGE_PATH=${TORCH_INSTALL_PATH::-6}
export CUDA_HOME=/usr/local/cuda-${cuda_version}/
# replace the MODULE.bazel with the jetpack one
cat toolchains/jp_workspaces/MODULE.bazel.tmpl | envsubst > MODULE.bazel
# build and install torch_tensorrt wheel file
python setup.py --use-cxx11-abi install --user
some errors happened:
Run this command to start an interactive shell in an identical sandboxed environment:
(exec env - \
LD_LIBRARY_PATH=/usr/lib/gcc/aarch64-linux-gnu/11:/usr/local/cuda-12.6/lib64: \
PATH=/home/lab223/.cache/bazelisk/downloads/sha256/5a4cc979353671e438b9469b833924c2361e25a580cc278a75877aedc27c1c53/bin:/usr/lib/gcc/aarch64-linux-gnu/11:/home/lab223/anaconda3/envs/rnw/bin:/home/lab223/anaconda3/condabin:/usr/local/cuda-12.6/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games:/usr/local/games:/snap/bin:/snap/bin \
PWD=/proc/self/cwd \
TMPDIR=/tmp \
/home/lab223/.cache/bazel/_bazel_lab223/install/128438993754f9753a1e4f56fdd76124/linux-sandbox -t 15 -w /dev/shm -w /home/lab223/.cache/bazel/_bazel_lab223/3fb6c16c20f38dfc11e57e77e6eea473/sandbox/linux-sandbox/46/execroot/_main -w /tmp -M /home/lab223/.cache/bazel/_bazel_lab223/3fb6c16c20f38dfc11e57e77e6eea473/sandbox/linux-sandbox/46/_hermetic_tmp -m /tmp -S /home/lab223/.cache/bazel/_bazel_lab223/3fb6c16c20f38dfc11e57e77e6eea473/sandbox/linux-sandbox/46/stats.out -D /home/lab223/.cache/bazel/_bazel_lab223/3fb6c16c20f38dfc11e57e77e6eea473/sandbox/linux-sandbox/46/debug.out -- /bin/sh -i)
ERROR: /home/lab223/TensorRT/core/conversion/var/BUILD:20:11: Compiling core/conversion/var/Var.cpp failed: I/O exception during sandboxed execution: /home/lab223/.cache/bazel/_bazel_lab223/3fb6c16c20f38dfc11e57e77e6eea473/sandbox/linux-sandbox/58/execroot/_main/bazel-out/aarch64-opt/bin/external/_main~_repo_rules~libtorch/_virtual_includes/ATen/ATen/core/DeprecatedTypePropertiesRegistry.h (???????)
ERROR: /home/lab223/TensorRT/core/conversion/converters/BUILD:59:11: Compiling core/conversion/converters/NodeConverterRegistry.cpp failed: I/O exception during sandboxed execution: /home/lab223/.cache/bazel/_bazel_lab223/3fb6c16c20f38dfc11e57e77e6eea473/sandbox/linux-sandbox/57/execroot/_main/bazel-out/aarch64-opt/bin/external/_main~_repo_rules~libtorch/_virtual_includes/ATen/ATen/ops/cudnn_batch_norm_ops.h (???????)
ERROR: /home/lab223/TensorRT/core/conversion/converters/BUILD:39:11: Compiling core/conversion/converters/converter_util.cpp failed: I/O exception during sandboxed execution: /home/lab223/.cache/bazel/_bazel_lab223/3fb6c16c20f38dfc11e57e77e6eea473/sandbox/linux-sandbox/56/execroot/_main/external/_main~_repo_rules~libtorch/include/ATen/ops/native_dropout_backward_cpu_dispatch.h (???????)
Target //:libtorchtrt failed to build
INFO: Elapsed time: 1000.299s, Critical Path: 574.06s
INFO: 7984 processes: 7938 internal, 46 linux-sandbox.
ERROR: Build did NOT complete successfully
Environment
Build information about Torch-TensorRT can be found by turning on debug messages
- PyTorch Version (e.g., 1.0): 2.5.0
- CPU Architecture: arm64(Jetson AGX Orin)
- OS (e.g., Linux): Linux
- How you installed PyTorch: pip
- Build command you used (if compiling from source): python setup.py --use-cxx11-abi install --user
- Are you using local sources or building from archives: building from archives
- Python version: 3.10.15
- CUDA version: 12.6
- GPU models and configuration: -
- Any other relevant information: Install torch_tensorrt in the model's anaconda virtual environment
Additional context
It seems a I/O exception.But Jetson still has 11GB of space.please help me!thanks!!!!