Name | azureml-ngc-tools JSON |
Version |
1.5.1
JSON |
| download |
home_page | |
Summary | AzureML integration with NGC |
upload_time | 2022-12-07 17:04:01 |
maintainer | |
docs_url | None |
author | |
requires_python | >=3.5 |
license | BSD |
keywords |
azureml
ngc
gpu
|
VCS |
|
bugtrack_url |
|
requirements |
No requirements were recorded.
|
Travis-CI |
No Travis.
|
coveralls test coverage |
No coveralls.
|
# AzureML NVIDIA GPU Cloud tools
The code contained within this repository allows pulling the images from NVIDIA GPU Cloud (NGC).
## Installation
To install this package type
`pip install azureml-ngc-tools`
Alternatively, clone this repository and use python to install.
```
git clone https://github.com/.../azureml-ngc-tools.git
python setup.py install
```
## Configuration
Two configuration files are required:
1. A `json` file that contains the parameters to log in to AzureML Workspace and create Compute Target. **All the parameters shown below need to be provided.**
```
{
"azureml_user":
{
"subscription_id": "<YOUR-SUBSCRIPTION-ID>",
"resource_group": "<YOUR-RESOURCE-GROUP>",
"workspace_name": "<YOUR-WORKSPACE-NAME>",
"telemetry_opt_out": <true|false>
},
"aml_compute":
{
"ct_name":"<NAME-OF-YOUR-COMPUTE-TARGET>",
"exp_name":"<NAME-OF-YOUR-EXPERIMENT>",
"vm_name":"<SIZE-OF-THE-AZUREML-VM>",
"admin_name":"<ADMINISTRATOR-NAME>",
"min_nodes":<MINIMUM-NUMBER-OF-NODES>,
"max_nodes":<MAXIMUM-NUMBER-OF-NODES>,
"vm_priority": "<dedicated|lowpriority>",
"idle_seconds_before_scaledown":<NUMBER-OF-SECONDS-TO-SCALE-DOWN>,
"python_interpreter":"<PATH-TO-PYTHON-INTERPRETER>",
"conda_packages":[<LIST-OF-ADDITIONAL-CONDA-OR-PIP-PACKAGES>],
"environment_name":"<NAME-OF-ENVIRONMENT>",
"docker_enabled":<true|false>,
"user_managed_dependencies":<true|false>,
"jupyter_port":<JUPYTER-PORT-FOR-FORWARDING>
}
}
```
An example (fictitious):
```
{
"azureml_user":
{
"subscription_id": "ef4455fa-3e35-433c-a410-76d7a8a9e793",
"resource_group": "sample-rg",
"workspace_name": "sample-ws",
"telemetry_opt_out": false
},
"aml_compute":
{
"ct_name":"sample-ct",
"exp_name":"sample-exp",
"vm_name":"Standard_NC6s_v3",
"admin_name": "sample",
"min_nodes":0,
"max_nodes":1,
"vm_priority": "dedicated",
"idle_seconds_before_scaledown":300,
"python_interpreter":"/usr/bin/python",
"conda_packages":["matplotlib","jupyterlab"],
"environment_name":"sample_env",
"docker_enabled":true,
"user_managed_dependencies":true,
"jupyter_port":9000
}
}
```
2. A `json` file that contains information about the content you want to download from NGC. **The `base_dockerfile` parameter shown below needs to be provided.**
```
{
"base_dockerfile":"<URI-TO-NGC-CONTAINER>",
"additional_content": {
"download_content": false,
"unzip_content": false,
"upload_content": false,
"list":[
{
"url": <URL_TO_CONTENT>,
"filename": <FILENAME_TO_SAVE_TO>,
"localdirectory": <DIRECTORY_TO_EXTRACT_CONTENTS>,
"computedirectory": <DIRECTORY_TO_UPLOAD_CONTENTS>,
"zipped": <false|true>
},
...
]
}
}
```
An example:
```
{
"base_dockerfile":"nvcr.io/nvidia/clara-train-sdk:v3.0",
"additional_content": {
"download_content": true,
"unzip_content": true,
"upload_content": true,
"list":[
{
"url":"https://api.ngc.nvidia.com/v2/resources/nvidia/med/getting_started/versions/1/zip",
"filename":"clarasdk.zip",
"localdirectory":"clara",
"computedirectory":"clara",
"zipped":true
}
]
}
}
```
## Usage
Assuming that the AzureML config file is `user_config.json` and the NGC config file is `ngc_app.json`, and both of the files are located in the same folder, to create the cluster run the following code
`azureml-ngc-tools --login user_config.json --app ngc_app.json`
Raw data
{
"_id": null,
"home_page": "",
"name": "azureml-ngc-tools",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.5",
"maintainer_email": "",
"keywords": "azureml,ngc,gpu",
"author": "",
"author_email": "",
"download_url": "https://files.pythonhosted.org/packages/be/ef/d239318060014ca7c1b9317e034eb9e55af31d8cc69933bb049aa0e3b699/azureml_ngc_tools-1.5.1.tar.gz",
"platform": null,
"description": "# AzureML NVIDIA GPU Cloud tools\n\nThe code contained within this repository allows pulling the images from NVIDIA GPU Cloud (NGC).\n\n## Installation\n\nTo install this package type \n\n`pip install azureml-ngc-tools`\n\nAlternatively, clone this repository and use python to install.\n\n```\ngit clone https://github.com/.../azureml-ngc-tools.git\npython setup.py install\n```\n\n## Configuration\n\nTwo configuration files are required:\n\n1. A `json` file that contains the parameters to log in to AzureML Workspace and create Compute Target. **All the parameters shown below need to be provided.**\n\n```\n{\n \"azureml_user\":\n {\n \"subscription_id\": \"<YOUR-SUBSCRIPTION-ID>\",\n \"resource_group\": \"<YOUR-RESOURCE-GROUP>\",\n \"workspace_name\": \"<YOUR-WORKSPACE-NAME>\",\n \"telemetry_opt_out\": <true|false>\n },\n \"aml_compute\":\n {\n \"ct_name\":\"<NAME-OF-YOUR-COMPUTE-TARGET>\",\n \"exp_name\":\"<NAME-OF-YOUR-EXPERIMENT>\",\n \"vm_name\":\"<SIZE-OF-THE-AZUREML-VM>\",\n \"admin_name\":\"<ADMINISTRATOR-NAME>\",\n \"min_nodes\":<MINIMUM-NUMBER-OF-NODES>,\n \"max_nodes\":<MAXIMUM-NUMBER-OF-NODES>,\n \"vm_priority\": \"<dedicated|lowpriority>\",\n \"idle_seconds_before_scaledown\":<NUMBER-OF-SECONDS-TO-SCALE-DOWN>,\n \"python_interpreter\":\"<PATH-TO-PYTHON-INTERPRETER>\",\n \"conda_packages\":[<LIST-OF-ADDITIONAL-CONDA-OR-PIP-PACKAGES>],\n \"environment_name\":\"<NAME-OF-ENVIRONMENT>\",\n \"docker_enabled\":<true|false>,\n \"user_managed_dependencies\":<true|false>,\n \"jupyter_port\":<JUPYTER-PORT-FOR-FORWARDING>\n }\n}\n\n```\n\nAn example (fictitious):\n\n```\n{\n \"azureml_user\":\n {\n \"subscription_id\": \"ef4455fa-3e35-433c-a410-76d7a8a9e793\",\n \"resource_group\": \"sample-rg\",\n \"workspace_name\": \"sample-ws\",\n \"telemetry_opt_out\": false\n },\n \"aml_compute\":\n {\n \"ct_name\":\"sample-ct\",\n \"exp_name\":\"sample-exp\",\n \"vm_name\":\"Standard_NC6s_v3\",\n \"admin_name\": \"sample\",\n \"min_nodes\":0,\n \"max_nodes\":1,\n \"vm_priority\": \"dedicated\",\n \"idle_seconds_before_scaledown\":300,\n \"python_interpreter\":\"/usr/bin/python\",\n \"conda_packages\":[\"matplotlib\",\"jupyterlab\"],\n \"environment_name\":\"sample_env\",\n \"docker_enabled\":true,\n \"user_managed_dependencies\":true,\n \"jupyter_port\":9000\n }\n}\n\n```\n\n2. A `json` file that contains information about the content you want to download from NGC. **The `base_dockerfile` parameter shown below needs to be provided.**\n\n```\n{\n \"base_dockerfile\":\"<URI-TO-NGC-CONTAINER>\",\n \"additional_content\": {\n \"download_content\": false,\n \"unzip_content\": false,\n \"upload_content\": false,\n \"list\":[\n {\n \"url\": <URL_TO_CONTENT>,\n \"filename\": <FILENAME_TO_SAVE_TO>,\n \"localdirectory\": <DIRECTORY_TO_EXTRACT_CONTENTS>,\n \"computedirectory\": <DIRECTORY_TO_UPLOAD_CONTENTS>,\n \"zipped\": <false|true>\n },\n ...\n ]\n }\n}\n```\n\nAn example:\n\n```\n{\n \"base_dockerfile\":\"nvcr.io/nvidia/clara-train-sdk:v3.0\",\n \"additional_content\": {\n \"download_content\": true,\n \"unzip_content\": true,\n \"upload_content\": true,\n \"list\":[\n {\n \"url\":\"https://api.ngc.nvidia.com/v2/resources/nvidia/med/getting_started/versions/1/zip\",\n \"filename\":\"clarasdk.zip\",\n \"localdirectory\":\"clara\",\n \"computedirectory\":\"clara\",\n \"zipped\":true\n }\n ]\n }\n}\n```\n\n## Usage\nAssuming that the AzureML config file is `user_config.json` and the NGC config file is `ngc_app.json`, and both of the files are located in the same folder, to create the cluster run the following code\n\n`azureml-ngc-tools --login user_config.json --app ngc_app.json`\n",
"bugtrack_url": null,
"license": "BSD",
"summary": "AzureML integration with NGC",
"version": "1.5.1",
"split_keywords": [
"azureml",
"ngc",
"gpu"
],
"urls": [
{
"comment_text": "",
"digests": {
"md5": "c0756a2113d2559dc82643009d3cca1d",
"sha256": "cd97924a9781a7b724a31e881ba912f35b1fe41b51d162a1a1693e7fc98d2b18"
},
"downloads": -1,
"filename": "azureml_ngc_tools-1.5.1-py3-none-any.whl",
"has_sig": false,
"md5_digest": "c0756a2113d2559dc82643009d3cca1d",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.5",
"size": 21747,
"upload_time": "2022-12-07T17:03:59",
"upload_time_iso_8601": "2022-12-07T17:03:59.245659Z",
"url": "https://files.pythonhosted.org/packages/27/8e/00afcc9d4ea2fe7411c9fab7955e67b26811480a838ccd783bdc069c3cf4/azureml_ngc_tools-1.5.1-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"md5": "e6b5ab115eaa356fabe863e4dce7e28b",
"sha256": "a7a3dd39e2b225ca623c1d5a9fdb98a5ebe47bafe647fb31256e6069fa11c55f"
},
"downloads": -1,
"filename": "azureml_ngc_tools-1.5.1.tar.gz",
"has_sig": false,
"md5_digest": "e6b5ab115eaa356fabe863e4dce7e28b",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.5",
"size": 40191,
"upload_time": "2022-12-07T17:04:01",
"upload_time_iso_8601": "2022-12-07T17:04:01.025035Z",
"url": "https://files.pythonhosted.org/packages/be/ef/d239318060014ca7c1b9317e034eb9e55af31d8cc69933bb049aa0e3b699/azureml_ngc_tools-1.5.1.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2022-12-07 17:04:01",
"github": false,
"gitlab": false,
"bitbucket": false,
"lcname": "azureml-ngc-tools"
}