Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Updates README.md to include instructions for launching Triton #289

Merged
5 commits merged into from
Jul 25, 2022
Merged

Updates README.md to include instructions for launching Triton #289

Show file tree
Hide file tree
Changes from 3 commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
4562c62
Move LFS info from CONTRIBUTING.md to README.md
dagardner-nv Jul 19, 2022
4bc5727
Add Triton to requirements sectionm, and move triton launching instru…
dagardner-nv Jul 19, 2022
c13fd01
Clean up requirements section:
dagardner-nv Jul 19, 2022
7f828c5
Switch to tritonserver 22.06 per feedback from Pete, along with a few…
dagardner-nv Jul 25, 2022
4eecb94
Update copyright
dagardner-nv Jul 25, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
51 changes: 3 additions & 48 deletions CONTRIBUTING.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -65,44 +65,15 @@ cd $MORPHEUS_ROOT
```
The large model and data files in this repo are stored using [Git Large File Storage (LFS)](https://git-lfs.github.com/). These files will be required for running the training/validation scripts and example pipelines for the Morpheus pre-trained models.

By default only those files stored in LFS strictly needed for running Morpheus are included when the Morpheus repository is cloned. Additional datasets can be downloaded using the `scripts/fetch_data.py` script. Usage of the script is as follows:
```bash
scripts/fetch_data.py fetch <dataset> [<dataset>...]
```

At time of writing the defined datasets are:
* all - Metaset includes all others
* examples - Data needed by scripts in the `examples` subdir
* models - Morpheus models (largest dataset)
* tests - Data used by unittests
* validation - Subset of the models dataset needed by some unittests

To download just the examples and models:
```bash
scripts/fetch_data.py fetch examples models
```

To download the data needed for unittests:
```bash
scripts/fetch_data.py fetch tests validation
```

If `Git LFS` is not installed before cloning the repository, the large files will not be pulled. If this is the case, follow the instructions for installing `Git LFS` from [here](https://git-lfs.github.com/), and then run the following command.
```bash
scripts/fetch_data.py fetch all
```
By default only those files stored in LFS strictly needed for running Morpheus are included when the Morpheus repository is cloned. Additional datasets can be downloaded using the `scripts/fetch_data.py` script. See the section [Git LFS](README.md#git-lfs) of the [README.md](README.md) file for details on this.

### Build in Docker Container

This workflow utilizes a docker container to set up most dependencies ensuring a consistent environment.

#### Prerequisites

- Pascal architecture or better
- NVIDIA driver `450.80.02` or higher
- [Docker](https://docs.docker.com/get-docker/)
- [The NVIDIA container toolkit](https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/install-guide.html#docker)

1. Ensure all [requirements](README.md#requirements) from [README.md](README.md) are met.
1. Build the development container
```bash
./docker/build_container_dev.sh
Expand Down Expand Up @@ -179,7 +150,7 @@ Note: These instructions assume the user is using `mamba` instead of `conda` sin

- Pascal architecture or better
dagardner-nv marked this conversation as resolved.
Show resolved Hide resolved
- NVIDIA driver `450.80.02` or higher
- [CUDA 11.0+](https://developer.nvidia.com/cuda-downloads)
- [Cuda 11.5](https://developer.nvidia.com/cuda-11-5-2-download-archive)
dagardner-nv marked this conversation as resolved.
Show resolved Hide resolved
- `conda` and `mamba`
- See the [Getting Started Guide](https://conda.io/projects/conda/en/latest/user-guide/install/index.html) if `conda` is not already installed
- Install `mamba`:
Expand Down Expand Up @@ -329,22 +300,6 @@ Launching a full production Kafka cluster is outside the scope of this project.

**Note:** This will consume messages.

### Launching Triton Server

Many of the validation tests and example workflows require a Triton server to function. To launch Triton server, use the following command:

```bash
docker run --rm -ti --gpus=all -p8000:8000 -p8001:8001 -p8002:8002 -v $PWD/models:/models \
nvcr.io/nvidia/tritonserver:21.12-py3 \
tritonserver --model-repository=/models/triton-model-repo \
--exit-on-error=false \
--model-control-mode=explicit \
--load-model abp-nvsmi-xgb \
--load-model sid-minibert-onnx \
--load-model phishing-bert-onnx
```
This will launch Triton using port 8001 for the GRPC server. This needs to match the Morpheus configuration.

### Pipeline Validation

To verify that all pipelines are working correctly, validation scripts have been added at `${MORPHEUS_ROOT}/scripts/validation`. There are scripts for each of the main workflows: Anomalous Behavior Profiling (ABP), Humans-as-Machines-Machines-as-Humans (HAMMAH), Phishing Detection (Phishing), and Sensitive Information Detection (SID).
Expand Down
56 changes: 46 additions & 10 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,33 +21,53 @@ More advanced users, or those who are interested in using the latest pre-release
The following sections must be followed prior to building the Morpheus container or building Morpheus from source.

#### Requirements
- Pascal architecture or better
- Pascal architecture GPU or better
- NVIDIA driver `450.80.02` or higher
- [Docker](https://docs.docker.com/get-docker/)
- [The NVIDIA container toolkit](https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/install-guide.html#docker)
- [NVIDIA Triton Inference Server](https://catalog.ngc.nvidia.com/orgs/nvidia/containers/tritonserver) `22.02` or higher
dagardner-nv marked this conversation as resolved.
Show resolved Hide resolved
- [Git LFS](https://git-lfs.github.com/)


#### Clone the Repository

```bash
MORPHEUS_ROOT=$(pwd)/morpheus
git clone https://github.com/NVIDIA/Morpheus.git $MORPHEUS_ROOT
cd $MORPHEUS_ROOT
```

#### Git LFS

The large model and data files in this repo are stored using [Git Large File Storage (LFS)](https://git-lfs.github.com/). These files will be required for running the training/validation scripts and example pipelines for the Morpheus pre-trained models.
The large model and data files in this repo are stored using [Git Large File Storage (LFS)](https://git-lfs.github.com/). Only those files which are strictly needed to run Morpheus are downloaded by default when the repository is cloned.

If `Git LFS` is not installed before cloning the repository, the large files will not be pulled. If this is the case, follow the instructions for installing `Git LFS` from [here](https://git-lfs.github.com/), and then run the following command:
The `scripts/fetch_data.py` script can be used to fetch the Morpheus pre-trained models, and other files required for running the training/validation scripts and example pipelines.

Usage of the script is as follows:
```bash
git lfs install
scripts/fetch_data.py fetch <dataset> [<dataset>...]
```

#### Clone the Repository
At time of writing the defined datasets are:
* all - Metaset includes all others
* examples - Data needed by scripts in the `examples` subdir
* models - Morpheus models (largest dataset)
* tests - Data used by unittests
* validation - Subset of the models dataset needed by some unittests

To download just the examples and models:
```bash
MORPHEUS_ROOT=$(pwd)/morpheus
git clone https://github.com/NVIDIA/Morpheus.git $MORPHEUS_ROOT
cd $MORPHEUS_ROOT
scripts/fetch_data.py fetch examples models
```

**Note:** If the repository was cloned before `Git LFS` was installed, you can ensure you have downloaded the LFS files with the command:
To download the data needed for unittests:
```bash
scripts/fetch_data.py fetch tests validation
```

If `Git LFS` is not installed the before cloning the repository, the `scripts/fetch_data.py` script will fail. If this is the case follow the instructions for installing `Git LFS` from [here](https://git-lfs.github.com/), and then run the following command:
```bash
git lfs pull
git lfs install
```

### Build Morpheus Container
Expand Down Expand Up @@ -76,6 +96,22 @@ DOCKER_IMAGE_TAG="v22.06.00a-runtime" ./docker/run_container_release.sh

It's possible to build from source outside of a container. However, due to the large number of dependencies, this can be complex and is only necessary for developers. Instructions for developers and contributors can be found in [CONTRIBUTING.md](./CONTRIBUTING.md).

## Launching Triton Server

Many of the validation tests and example workflows require a Triton server to function.
Use the following command to launch a Docker container for Triton loading all of the included pre-trained models:

```bash
docker run --rm -ti --gpus=all -p8000:8000 -p8001:8001 -p8002:8002 \
-v $PWD/models:/models \
nvcr.io/nvidia/tritonserver:22.02-py3 \
dagardner-nv marked this conversation as resolved.
Show resolved Hide resolved
tritonserver --model-repository=/models/triton-model-repo \
--exit-on-error=false \
--log-info=true \
--strict-readiness=false
```
This will launch Triton using the default network ports (8000 for HTTP, 8001 for GRPC, and 8002 for metrics).

## Running Morpheus

To run Morpheus, users will need to choose from the Morpheus Command Line Interface (CLI) or Python interface. Which interface to use depends on the user's needs, amount of customization, and operating environment. More information on each interface can be found below.
Expand Down