Remarks
- Release has been extended with library versions built against CUDA 11.8
- Release has been been extended with
.whl
Python packages
Installation (amd64/arm64)
CPU
TF_VERSION=2.11.0
ARCH=$(dpkg --print-architecture)
wget https://github.com/ika-rwth-aachen/libtensorflow_cc/releases/download/v${TF_VERSION}/libtensorflow-cc_${TF_VERSION}_${ARCH}.deb
sudo dpkg -i libtensorflow-cc_${TF_VERSION}_${ARCH}.deb
ldconfig
GPU
TF_VERSION=2.11.0
ARCH=$(dpkg --print-architecture)
wget https://github.com/ika-rwth-aachen/libtensorflow_cc/releases/download/v${TF_VERSION}/libtensorflow-cc_${TF_VERSION}-gpu_${ARCH}.deb
sudo dpkg -i libtensorflow-cc_${TF_VERSION}-gpu_${ARCH}.deb
ldconfig
Multi-Architecture Docker Images
CPU |
GPU |
rwthika/tensorflow-cc:2.11.0 |
rwthika/tensorflow-cc:2.11.0-gpu |
Version Information
TensorFlow |
Architecture |
Ubuntu |
GCC |
Bazel |
Python |
protobuf |
CUDA |
cuDNN |
TensorRT |
GPU Compute Capability |
2.11.0 |
amd64 |
20.04 |
9.4.0 |
6.0.0 |
3.8.10 |
3.9.2 |
11.2.152 |
8.1.0 |
7.2.2 |
8.6, 8.0, 7.5, 7.2, 7.0, 6.1, 6.0, 5.3 |
2.11.0 |
arm64 |
20.04 |
9.4.0 |
6.0.0 |
3.8.10 |
3.9.2 |
11.4.239 |
8.4.1 |
8.4.1 |
8.7, 8.6, 8.0, 7.5, 7.2, 7.0, 6.1, 6.0, 5.3 |
2.11.0 |
amd64 |
20.04 |
9.4.0 |
6.0.0 |
3.8.10 |
3.9.2 |
11.8.89 |
8.6.0 |
8.5.3 |
9.0, 8.9, 8.6, 8.0, 7.5, 7.0, 6.1, 6.0 |
2.11.0 |
arm64 |
20.04 |
9.4.0 |
6.0.0 |
3.8.10 |
3.9.2 |
11.8.89 |
8.6.0 |
8.5.3 |
8.7, 7.2, 6.2, 5.3 |