Package Details: tensorrt 10.7.0.23-1

Git Clone URL: https://aur.archlinux.org/tensorrt.git (read-only, click to copy)
Package Base: tensorrt
Description: A platform for high-performance deep learning inference on NVIDIA hardware
Upstream URL: https://developer.nvidia.com/tensorrt/
Keywords: ai artificial intelligence nvidia
Licenses: Apache-2.0, LicenseRef-NVIDIA-TensorRT-SLA
Submitter: dbermond
Maintainer: dbermond
Last Packager: dbermond
Votes: 20
Popularity: 0.53
First Submitted: 2018-07-29 16:17 (UTC)
Last Updated: 2024-12-07 14:13 (UTC)

Latest Comments

1 2 3 4 5 6 .. 10 Next › Last »

dbermond commented on 2024-12-19 03:26 (UTC)

@sots removing the unneeded patch is already on my radar, thanks for pointing anyway. The build jobs/parallelism is a user setting and should be configured in your 'makepkg.conf'.

sots commented on 2024-12-19 01:55 (UTC)

Recommended patch for PKGBUILD:

--- a/PKGBUILD
+++ b/PKGBUILD
@@ -27,7 +27,6 @@ source=("https://developer.nvidia.com/downloads/compute/machine-learning/tensorr
         "https://github.com/google/protobuf/releases/download/v${_protobuf_ver}/protobuf-cpp-${_protobuf_ver}.tar.gz"
         '010-tensorrt-use-local-protobuf-sources.patch'
         '020-tensorrt-fix-python.patch'
-        '030-tensorrt-onnx-tensorrt-disable-missing-source-file.patch'
         'TensorRT-SLA.txt')
 noextract=("protobuf-cpp-${_protobuf_ver}.tar.gz")
 sha256sums=('d7f16520457caaf97ad8a7e94d802f89d77aedf9f361a255f2c216e2a3a40a11'
@@ -41,7 +40,6 @@ sha256sums=('d7f16520457caaf97ad8a7e94d802f89d77aedf9f361a255f2c216e2a3a40a11'
             'dddd73664306d7d895a95e1cf18925b31b52785e468727e4635b45edae5166f9'
             'ba94c0685216fe9566f7989df98b372e72a8da04b66d64380024107f2f7f4a8f'
             '62109d798ecfe286aa12d8c848927baa074b422669fb6248590a7ae00074ab26'
-            '5b90eb795c5d7a209a67c7aad73f3b472650e52c0c42f8cd15b43b5a6c6396c4'
             'c493df1b40477f750d9840a1b0b6cf5ec36c1b6bde3bcd6c7029e27e60df3c58')

 prepare() {
@@ -73,7 +71,7 @@ prepare() {
     patch -d TensorRT -Np1 -i "${srcdir}/020-tensorrt-fix-python.patch"

     # https://github.com/onnx/onnx-tensorrt/issues/979
-    patch -d TensorRT/parsers/onnx -Np1 -i "${srcdir}/030-tensorrt-onnx-tensorrt-disable-missing-source-file.patch"
+    # patch -d TensorRT/parsers/onnx -Np1 -i "${srcdir}/030-tensorrt-onnx-tensorrt-disable-missing-source-file.patch"
 }

 build() {
@@ -88,7 +86,7 @@ build() {
         -DGPU_ARCHS='50 52 53 60 61 62 70 72 75 80 86 87 89 90' \
         -DPROTOBUF_VERSION="$_protobuf_ver" \
         -Wno-dev
-    cmake --build build
+    cmake --build build -j "$(nproc)"

     # python bindings (fails to build with python 3.11)
     #local _pyver

juliusk commented on 2024-12-03 01:21 (UTC) (edited on 2024-12-03 01:21 (UTC) by juliusk) 

CMake Error: The following variables are used in this project, but they are set to NOTFOUND.
Please set them or make sure they are set and tested correctly in the CMake files: nvinfer_LIB_PATH

whiteLinux commented on 2024-11-10 05:12 (UTC)

030-tensorrt-onnx-tensorrt-disable-missing-source-file.patch is no longer needed.The annotated file "errorHelpers.cpp" provide _ZTIN8onnx2trt16OnnxTrtExceptionE define.python-onnx package also include this file,so install python-onnx will fix it

dbermond commented on 2024-11-02 14:47 (UTC)

@jholmer try to rebuild the python-onnx package and use it. For rebuilding, use the PKGBUILD from the official repositories (can be found in this link). If if works, please open a bug report for python-onnx.

FuzzyAtish commented on 2024-10-31 11:07 (UTC)

@jholmer In the past with similar errors what worked for me was to reinstall the python-onnx package. I'm not saying it's a definite solution, but it could work

jholmer commented on 2024-10-20 01:17 (UTC)

I am also receiving the "Could not load library libnvonnxparser.so.10: Unable to open library: libnvonnxparser.so.10 due to /usr/lib/libnvonnxparser.so.10: undefined symbol: _ZTIN8onnx2trt16OnnxTrtExceptionE" error on runtime. I have tried doing a completely clean build. I'm wondering if there is some sort of version incompatibility between this package and a dependency?

dbermond commented on 2024-09-19 21:26 (UTC)

@lu0se yes, the '.so' link to 'libnvonnxparser.so.10' is right, otherwise you would get a 'file not found' error. If you are getting a 'Using existing $srcdir/ tree' warning during makepkg, it means that you are not doing a clean build. You should use makepkg --cleanbuild/-C option for doing a clean build when building your packages, or build it in a clean chroot. Try to do it and see if it works.

lu0se commented on 2024-09-19 18:03 (UTC)

trtexec --onnx=rife_v4.10.onnx

&&&& RUNNING TensorRT.trtexec [TensorRT v100400] [b26] # trtexec --onnx=/usr/lib/vapoursynth/models/rife/rife_v4.10.onnx --device=0 [09/20/2024-02:01:38] [I] Start parsing network model. [09/20/2024-02:01:38] [E] Could not load library libnvonnxparser.so.10: Unable to open library: libnvonnxparser.so.10 due to /usr/lib/libnvonnxparser.so.10: undefined symbol: _ZTIN8onnx2trt16OnnxTrtExceptionE [09/20/2024-02:01:38] [E] Assertion failure: parser.onnxParser != nullptr

is libnvonnxparser.so.10 link right？is it related to WARNING: Using existing $srcdir/ tree

milianw commented on 2024-08-06 16:28 (UTC)

@dbermond: the forum post is not mine. I got the same/similar error when I tried to edit the PKGBUILD manually to try to build the newer tensorrt against cuda 12.5.

1 2 3 4 5 6 .. 10 Next › Last »