Skip to content

Commit

Permalink
Update README for 22.08 release
Browse files Browse the repository at this point in the history
mc-nv committed Aug 26, 2022

Partially verified

This commit is signed with the committer’s verified signature.
gsmet’s contribution has been verified via GPG key.
We cannot verify signatures from co-authors, and some of the co-authors attributed to this commit require their commits to be signed.
1 parent b3d7a33 commit 98ee6a1
Showing 2 changed files with 338 additions and 2 deletions.
224 changes: 222 additions & 2 deletions README.md
Original file line number Diff line number Diff line change
@@ -26,5 +26,225 @@
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
-->

**NOTE: You are currently on the r22.08 branch which tracks stabilization
towards the next release. This branch is not usable during stabilization.**
# Triton Inference Server

[![License](https://img.shields.io/badge/License-BSD3-lightgrey.svg)](https://opensource.org/licenses/BSD-3-Clause)

**LATEST RELEASE: You are currently on the main branch which tracks
under-development progress towards the next release. The current release is
version [2.25.0](https://github.com/triton-inference-server/server/tree/r22.08)
and corresponds to the 22.08 container release on
[NVIDIA GPU Cloud (NGC)](https://catalog.ngc.nvidia.com/orgs/nvidia/containers/tritonserver).**

----
Triton Inference Server is an open source inference serving software that
streamlines AI inferencing. Triton enables teams to deploy any AI model from
multiple deep learning and machine learning frameworks, including TensorRT,
TensorFlow, PyTorch, ONNX, OpenVINO, Python, RAPIDS FIL, and more. Triton
supports inference across cloud, data center,edge and embedded devices on NVIDIA
GPUs, x86 and ARM CPU, or AWS Inferentia. Triton delivers optimized performance
for many query types, including real time, batched, ensembles and audio/video
streaming.

Major features include:

- [Supports multiple deep learning
frameworks](https://github.com/triton-inference-server/backend#where-can-i-find-all-the-backends-that-are-available-for-triton)
- [Supports multiple machine learning
frameworks](https://github.com/triton-inference-server/fil_backend)
- [Concurrent model
execution](docs/architecture.md#concurrent-model-execution)
- [Dynamic batching](docs/model_configuration.md#dynamic-batcher)
- [Sequence batching](docs/model_configuration.md#sequence-batcher) and
[implicit state management](docs/architecture.md#implicit-state-management)
for stateful models
- Provides [Backend API](https://github.com/triton-inference-server/backend) that
allows adding custom backends and pre/post processing operations
- Model pipelines using
[Ensembling](docs/architecture.md#ensemble-models) or [Business
Logic Scripting
(BLS)](https://github.com/triton-inference-server/python_backend#business-logic-scripting)
- [HTTP/REST and GRPC inference
protocols](docs/inference_protocols.md) based on the community
developed [KServe
protocol](https://github.com/kserve/kserve/tree/master/docs/predict-api/v2)
- A [C API](docs/inference_protocols.md#in-process-triton-server-api) and
[Java API](docs/inference_protocols.md#java-bindings-for-in-process-triton-server-api)
allow Triton to link directly into your application for edge and other in-process use cases
- [Metrics](docs/metrics.md) indicating GPU utilization, server
throughput, server latency, and more

Need enterprise support? NVIDIA global support is available for Triton
Inference Server with the
[NVIDIA AI Enterprise software suite](https://www.nvidia.com/en-us/data-center/products/ai-enterprise/).

## Serve a Model in 3 Easy Steps

```bash
# Step 1: Create the example model repository
git clone -b r22.08 https://github.com/triton-inference-server/server.git

cd server/docs/examples

./fetch_models.sh

# Step 2: Launch triton from the NGC Triton container
docker run --gpus=1 --rm --net=host -v /full/path/to/docs/examples/model_repository:/models nvcr.io/nvidia/tritonserver:22.08-py3 tritonserver --model-repository=/models

# Step 3: In a separate console, launch the image_client example from the NGC Triton SDK container
docker run -it --rm --net=host nvcr.io/nvidia/tritonserver:22.08-py3-sdk

/workspace/install/bin/image_client -m densenet_onnx -c 3 -s INCEPTION /workspace/images/mug.jpg

# Inference should return the following
Image '/workspace/images/mug.jpg':
15.346230 (504) = COFFEE MUG
13.224326 (968) = CUP
10.422965 (505) = COFFEEPOT
```
Please read the [QuickStart](docs/quickstart.md) guide for additional information
regarding this example. The quickstart guide also contains an example of how to launch Triton on [CPU-only systems](docs/quickstart.md#run-on-cpu-only-system).

## Examples and Tutorials

Check out [NVIDIA LaunchPad](https://www.nvidia.com/en-us/data-center/products/ai-enterprise-suite/trial/)
for free access to a set of hands-on labs with Triton Inference Server hosted on
NVIDIA infrastructure.

Specific end-to-end examples for popular models, such as ResNet, BERT, and DLRM
are located in the
[NVIDIA Deep Learning Examples](https://github.com/NVIDIA/DeepLearningExamples)
page on GitHub. The
[NVIDIA Developer Zone](https://developer.nvidia.com/nvidia-triton-inference-server)
contains additional documentation, presentations, and examples.

## Documentation

### Build and Deploy

The recommended way to build and use Triton Inference Server is with Docker
images.

- [Install Triton Inference Server with Docker containers](docs/build.md#building-triton-with-docker) (*Recommended*)
- [Install Triton Inference Server without Docker containers](docs/build.md#building-triton-without-docker)
- [Build a custom Triton Inference Server Docker container](docs/compose.md)
- [Build Triton Inference Server from source](docs/build.md#building-on-unsupported-platforms)
- [Build Triton Inference Server for Windows 10](docs/build.md#building-for-windows-10)
- Examples for deploying Triton Inference Server with Kubernetes and Helm on [GCP](deploy/gcp/README.md),
[AWS](deploy/aws/README.md), and [NVIDIA FleetCommand](deploy/fleetcommand/README.md)

### Using Triton

#### Preparing Models for Triton Inference Server

The first step in using Triton to serve your models is to place one or
more models into a [model repository](docs/model_repository.md). Depending on
the type of the model and on what Triton capabilities you want to enable for
the model, you may need to create a [model
configuration](docs/model_configuration.md) for the model.

- [Add custom operations to Triton if needed by your model](docs/custom_operations.md)
- Enable model pipelining with [Model Ensemble](docs/architecture.md#ensemble-models)
and [Business Logic Scripting (BLS)](https://github.com/triton-inference-server/python_backend#business-logic-scripting)
- Optimize your models setting [scheduling and batching](docs/architecture.md#models-and-schedulers)
parameters and [model instances](docs/model_configuration.md#instance-groups).
- Use the [Model Analyzer tool](https://github.com/triton-inference-server/model_analyzer)
to help optimize your model configuration with profiling
- Learn how to [explicitly manage what models are available by loading and
unloading models](docs/model_management.md)

#### Configure and Use Triton Inference Server

- Read the [Quick Start Guide](docs/quickstart.md) to run Triton Inference
Server on both GPU and CPU
- Triton supports multiple execution engines, called
[backends](https://github.com/triton-inference-server/backend#where-can-i-find-all-the-backends-that-are-available-for-triton), including
[TensorRT](https://github.com/triton-inference-server/tensorrt_backend),
[TensorFlow](https://github.com/triton-inference-server/tensorflow_backend),
[PyTorch](https://github.com/triton-inference-server/pytorch_backend),
[ONNX](https://github.com/triton-inference-server/onnxruntime_backend),
[OpenVINO](https://github.com/triton-inference-server/openvino_backend),
[Python](https://github.com/triton-inference-server/python_backend), and more
- Not all the above backends are supported on every platform supported by Triton.
Look at the
[Backend-Platform Support Matrix](https://github.com/triton-inference-server/backend/blob/main/docs/backend_platform_support_matrix.md)
to learn which backends are supported on your target platform.
- Learn how to [optimize performance](docs/optimization.md) using the
[Performance Analyzer](docs/perf_analyzer.md) and
[Model Analyzer](https://github.com/triton-inference-server/model_analyzer)
- Learn how to [manage loading and unloading models](docs/model_management.md) in
Triton
- Send requests directly to Triton with the [HTTP/REST JSON-based
or gRPC protocols](docs/inference_protocols.md#httprest-and-grpc-protocols)

#### Client Support and Examples

A Triton *client* application sends inference and other requests to Triton. The
[Python and C++ client libraries](https://github.com/triton-inference-server/client)
provide APIs to simplify this communication.

- Review client examples for [C++](https://github.com/triton-inference-server/client/blob/main/src/c%2B%2B/examples),
[Python](https://github.com/triton-inference-server/client/blob/main/src/python/examples),
and [Java](https://github.com/triton-inference-server/client/blob/main/src/java/src/main/java/triton/client/examples)
- Configure [HTTP](https://github.com/triton-inference-server/client#http-options)
and [gRPC](https://github.com/triton-inference-server/client#grpc-options)
client options
- Send input data (e.g. a jpeg image) directly to Triton in the [body of an HTTP
request without any additional metadata](https://github.com/triton-inference-server/server/blob/main/docs/protocol/extension_binary_data.md#raw-binary-request)

### Extend Triton

[Triton Inference Server's architecture](docs/architecture.md) is specifically
designed for modularity and flexibility

- [Customize Triton Inference Server container](docs/compose.md) for your use case
- [Create custom backends](https://github.com/triton-inference-server/backend)
in either [C/C++](https://github.com/triton-inference-server/backend/blob/main/README.md#triton-backend-api)
or [Python](https://github.com/triton-inference-server/python_backend)
- Create [decouple backends and models](docs/decoupled_models.md) that can send
multiple responses for a request or not send any responses for a request
- Use a [Triton repository agent](docs/repository_agents.md) to add functionality
that operates when a model is loaded and unloaded, such as authentication,
decryption, or conversion
- Deploy Triton on [Jetson and JetPack](docs/jetson.md)
- [Use Triton on AWS
Inferentia](https://github.com/triton-inference-server/python_backend/tree/main/inferentia)

### Additional Documentation

- [FAQ](docs/faq.md)
- [User Guide](docs#user-guide)
- [Developer Guide](docs#developer-guide)
- [Release Notes](https://docs.nvidia.com/deeplearning/triton-inference-server/release-notes/index.html)
- [GPU, Driver, and CUDA Support
Matrix](https://docs.nvidia.com/deeplearning/dgx/support-matrix/index.html)

## Contributing

Contributions to Triton Inference Server are more than welcome. To
contribute please review the [contribution
guidelines](CONTRIBUTING.md). If you have a backend, client,
example or similar contribution that is not modifying the core of
Triton, then you should file a PR in the [contrib
repo](https://github.com/triton-inference-server/contrib).

## Reporting problems, asking questions

We appreciate any feedback, questions or bug reporting regarding this project.
When posting [issues in GitHub](https://github.com/triton-inference-server/server/issues),
follow the process outlined in the [Stack Overflow document](https://stackoverflow.com/help/mcve).
Ensure posted examples are:
- minimal – use as little code as possible that still produces the
same problem
- complete – provide all parts needed to reproduce the problem. Check
if you can strip external dependencies and still show the problem. The
less time we spend on reproducing problems the more time we have to
fix it
- verifiable – test the code you're about to provide to make sure it
reproduces the problem. Remove all other problems that are not
related to your request/question.

## For more information

Please refer to the [NVIDIA Developer Triton page](https://developer.nvidia.com/nvidia-triton-inference-server)
for more information.
116 changes: 116 additions & 0 deletions RELEASE.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,116 @@
<!--
# Copyright 2018-2022, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions
# are met:
# * Redistributions of source code must retain the above copyright
# notice, this list of conditions and the following disclaimer.
# * Redistributions in binary form must reproduce the above copyright
# notice, this list of conditions and the following disclaimer in the
# documentation and/or other materials provided with the distribution.
# * Neither the name of NVIDIA CORPORATION nor the names of its
# contributors may be used to endorse or promote products derived
# from this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
# EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
# PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
# CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
# EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
# PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
# PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
# OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
-->

# Release Notes for 2.25.0

## New Freatures and Improvements

* New
[support for multiple cloud credentials](https://github.com/triton-inference-server/server/blob/main/docs/model_repository.md#cloud-storage-with-credential-file-beta)
has been enabled. This feature is in beta and is subject to change.

* Models using custom backends which implement
[auto-complete configuration](https://github.com/triton-inference-server/server/blob/main/docs/model_configuration.md#auto-generated-model-configuration),
can be loaded without explicit config.pbtxt file if they are named in form
`<model_name>.<backend_name>`.

* Users can specify a maximum memory limit when loading models onto the GPU
with the new
[--model-load-gpu-limit](https://github.com/triton-inference-server/server/blob/b3d7a3375e7adb1341724c0ac34661b4cde23cd2/src/main.cc#L629-L635)
tritonserver option and the
[TRITONSERVER_ServerOptionsSetModelLoadDeviceLimit](https://github.com/triton-inference-server/core/blob/c9cd6630ecb04bb26e2110cd65a37f23aec8153b/include/triton/core/tritonserver.h#L1861-L1872) C API function

* Added new documentation,
[Performance Tuning](https://github.com/triton-inference-server/server/blob/main/docs/performance_tuning.md), with a step by step guide to optimize models for
production

* From this release onwards Triton will default to
[TensorFlow version 2.X.](https://github.com/triton-inference-server/tensorflow_backend/tree/main#--backend-configtensorflowversionint)
TensorFlow version 1.X can still be manually specified via backend config.

* PyTorch backend has improved performance by using a separate CUDA Stream for
each model instance when the instance kind is GPU.

* Refer to the 22.08 column of the
[Frameworks Support Matrix](https://docs.nvidia.com/deeplearning/frameworks/support-matrix/index.html)
for container image versions on which the 22.08 inference server container is
based.

* Model Analyzer's profile subcommand now analyzes the results after Profile is
completed. Usage of the Analyze subcommand is deprecated. See
[Model Analyzer's documentation](https://github.com/triton-inference-server/model_analyzer/blob/main/docs/cli.md#subcommand-profile)
for further details.

## Known Issues

* There is no Jetpack release for 22.08, the latest release is 22.07.

* 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`.

* When auto-completing some model configs, backends may generate a model config
even though there is not enough metadata (ex. Graphdef models for TensorFlow
Backend). The user will see the model successfully load but fail to inference.
In this case the user should provide the full model configuration for these
models or use the `--disable-auto-complete-config` CLI option to show which
models fail to load.

* 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

* Auto-complete is not supported in the OpenVINO backend

* 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.

* Model Analyzer reported values for GPU utilization and GPU power are known to
be inaccurate and generally lower than reality.

0 comments on commit 98ee6a1

Please sign in to comment.