Triton Inference Server Release 23.02
The Triton Inference Server container image, release 23.02, is available on NGC and is open source on GitHub.
Contents of the Triton Inference Server container
The Triton Inference Server Docker image contains the inference server executable and related shared libraries in /opt/tritonserver.
For the list of what the container includes, refer to Deep Learning Frameworks Support Matrix.
Driver Requirements
Release 23.02 is based on CUDA 12.0.1, which requires NVIDIA Driver release 525 or later. However, if you are running on a data center GPU (for example, T4 or any other data center GPU), you can use NVIDIA driver release 450.51 (or later R450), 470.57 (or later R470), 510.47 (or later R510), 515.65 (or later R515), or 525.85 (or later R525).
The CUDA driver's compatibility package only supports particular drivers. Thus, users should upgrade from all R418, R440, R460, and R520 drivers, which are not forward-compatible with CUDA 12.0. For a complete list of supported drivers, see the CUDA Application Compatibility topic. For more information, see CUDA Compatibility and Upgrades.GPU Requirements
Release 23.02 supports CUDA compute capability 6.0 and later. This corresponds to GPUs in the NVIDIA Pascal, NVIDIA Volta™, NVIDIA Turing™, NVIDIA Ampere architecture, and NVIDIA Hopper™ architecture families. For a list of GPUs to which this compute capability corresponds, see CUDA GPUs. For additional support details, see Deep Learning Frameworks Support Matrix.
Key Features and Enhancements
This Inference Server release includes the following key features and enhancements.
- Support for ensemble models in Model Analyzer.
- Support for GRPC Standard Health Check Protocol.
- Fixed intermittent hangs during model loading for Python backend.
- Refer to the 23.02 column of the Frameworks Support Matrix for container image versions on which the 23.02 inference server container is based.
NVIDIA Triton Inference Server Container Versions
The following table shows what versions of Ubuntu, CUDA, Triton Inference Server, and NVIDIA TensorRT™ are supported in each of the NVIDIA containers for Triton Inference Server. For older container versions, refer to the Frameworks Support Matrix.
Known Issues
- In some rare cases Triton's TensorRT backend might overwrite input tensors while they are still in use which leads to corrupt input data being used for inference with TensorRT models. If you encounter accuracy issues with your TensorRT model, you can work-around the issue byenabling the output_copy_stream option in your model's configuration.
- Some systems which implement malloc() may not release memory back to the operating system right away causing a false memory leak. This can be mitigated by using a different malloc implementation. Tcmalloc is installed in the Triton container and can be used by specifying the library in LD_PRELOAD.
- When using a custom operator for the PyTorch backend, the operator may not be loaded due to undefined Python library symbols. This can be work-around by specifying Python library in LD_PRELOAD.
- Auto-complete may cause an increase in server start time. To avoid a start time increase, users can provide the full model configuration and launch the server with --disable-auto-complete-config.
- Auto-complete does not support PyTorch models due to lack of metadata in the model. It can only verify that the number of inputs and the input names matches what is specified in the model configuration. There is no model metadata about the number of outputs and datatypes. Related PyTorch bug:https://github.com/pytorch/pytorch/issues/38273
- Perf Analyzer stability criteria has been changed which may result in reporting instability for scenarios that were previously considered stable. This change has been made to improve the accuracy of Perf Analyzer results. If you observe this message, it can be resolved by increasing the --measurement-interval in the time windows mode or --measurement-request-count in the count windows mode.
- Triton Client PIP wheels for Arm SBSA are not available
from PyPI and pip will install an incorrect Jetson
version of Triton Client library for Arm SBSA.
The correct client wheel file can be pulled directly from the Arm SBSA SDK image and manually installed.
- Traced models in PyTorch seem to create overflows when
int8 tensor values are transformed to int32 on the
GPU.
Refer to https://github.com/pytorch/pytorch/issues/66930 for more information.
- Triton cannot retrieve GPU metrics with MIG-enabled GPU devices (A100 and A30).
- Triton metrics might not work if the host machine is running a separate DCGM agent on bare-metal or in a container.