| Name | ovhai JSON |
| Version |
0.1a4
JSON |
| download |
| home_page | None |
| Summary | A client library for accessing OVHcloud's AI Solutions |
| upload_time | 2024-05-28 14:04:51 |
| maintainer | None |
| docs_url | None |
| author | None |
| requires_python | <4,>=3.8 |
| license | Apache-2.0 |
| keywords |
ovhai
ovh
ai
|
| VCS |
 |
| bugtrack_url |
|
| requirements |
No requirements were recorded.
|
| Travis-CI |
No Travis.
|
| coveralls test coverage |
No coveralls.
|
# ovhai - OVHcloud's AI Python SDK
The `ovhai` library is a Python client that allows developers to easily use the [OVHcloud AI API](https://gra.training.ai.cloud.ovh.net/#/). With this SDK, you can run, manage, automate your notebooks, training and deployments in the cloud using OVHcloud's AI products (AI Notebooks, AI Training, AI Deploy).
## ⚠️ Alpha Warning ⚠️
This package is currently in the alpha phase of development. The APIs and functionalities of the package may not be fully tested.
## Installation
To install the SDK, run the following command:
```bash
pip install ovhai
```
The SDK requires Python 3.8 or higher. For information about how to update your Python version, see the [official Python documentation](https://www.python.org/downloads/).
## Getting started - Example Usage
Once you've installed the AI SDK, you can import it to use OVHcloud's AI products using the [API](https://gra.training.ai.cloud.ovh.net/#/).
You can start by the client creation:
```python
from ovhai import AuthenticatedClient
client = AuthenticatedClient(
base_url="https://gra.training.ai.cloud.ovh.net",
token="YOUR_AI_TOKEN",
)
```
The token used to create the client can be created via the OVHcloud Control Panel (UI), from the AI Dashboard.
Once your client is defined, you can call an endpoint:
```python
from ovhai.models import Me
from ovhai.api.me import me
from ovhai.ovhai_types import Response
with client as client:
res: Me = me.sync(client=client)
print(res)
# or if you need more info (e.g. status_code)
#response: Response[Me] = me.sync_detailed(client=client)
#print(response)
```
Or do the same thing with an async version:
```python
import asyncio
from ovhai.models import Me
from ovhai.api.me import me
from ovhai.ovhai_types import Response
async def main(client):
res: Me = await me.asyncio(client=client)
print(res)
# or if you need more info (e.g. status_code)
#response: Response[Me] = await me.asyncio_detailed(client=client)
#print(response)
# Run the main function asynchronously
asyncio.run(main(client=client))
```
In the `ovhai/api` [folder](https://github.com/ovh/ovhai-python-sdk/tree/main/ovhai/api), you will find all the endpoints you can call up. They are grouped by folder according to their purpose (`notebook`, `job`, `app`, `...`).
For example, to launch a notebook, you need to import the [`notebook_new` file](https://github.com/ovh/ovhai-python-sdk/blob/master/ovhai/api/notebook/notebook_new.py), located at `/ovhai/api/notebook`. You will also need to import the objects linked to this endpoint (those mentioned in the python file), since you will manipulate them (`Notebook` and `NotebookSpec` here, in addition to the classic `AuthenticatedClient` and `Response`). This will allow you to launch your first notebook using the `ovhai` python library, based on your specifications:
```python
from ovhai import AuthenticatedClient
from ovhai.api.notebook import notebook_new
from ovhai.models import NotebookSpec, Notebook
from ovhai.ovhai_types import Response
client = AuthenticatedClient(
base_url="https://gra.training.ai.cloud.ovh.net",
token="YOUR_AI_TOKEN",
)
# Define notebook parameters
editor_id = "jupyterlab"
framework_id = "conda"
framework_version = "conda-py39-cudaDevel11.8-v22-4"
nb_cpu = 2
# Create the notebook creation request
notebook_specs = {
"env": {"editorId": editor_id, "frameworkId": framework_id, "frameworkVersion": framework_version},
"resources": {"cpu": nb_cpu},
}
with client as client:
response: Response[Notebook] = notebook_new.sync_detailed(
client=client, body=NotebookSpec.from_dict(notebook_specs)
)
print(response)
```
The Response object returned will contain various information, including your notebook UUID.
## Things to know
Every [OVHcloud's AI API](https://gra.training.ai.cloud.ovh.net/#/) endpoint has its dedicated Python module that comes with four functions:
1. `sync`: Blocking request that returns parsed data (if successful) or `None`
2. `sync_detailed`: Blocking request that always returns a `Request`, optionally with `parsed` set if the request was successful.
3. `asyncio`: Like `sync` but async instead of blocking
4. `asyncio_detailed`: Like `sync_detailed` but async instead of blocking
To implement the call you want, find the folder and then the file corresponding to your needs in the [api folder](https://github.com/ovh/ovhai-python-sdk/tree/master/ovhai/api). Then choose the method that suits you best from the four mentioned above. Then import the objects you need to use this method.
## Advanced customizations
There are more settings on the generated `AuthenticatedClient` class which let you control more runtime behavior, check out the [docstring on that class](https://github.com/ovh/ovhai-python-sdk/blob/master/ovhai/client.py) for more info. You can also customize the underlying `httpx.Client` or `httpx.AsyncClient` (depending on your use-case):
```python
from ovhai import AuthenticatedClient
def log_request(request):
print(f"Request event hook: {request.method} {request.url} - Waiting for response")
def log_response(response):
request = response.request
print(f"Response event hook: {request.method} {request.url} - Status {response.status_code}")
client = AuthenticatedClient(
base_url="https://gra.training.ai.cloud.ovh.net",
token="YOUR_AI_TOKEN",
httpx_args={"event_hooks": {"request": [log_request], "response": [log_response]}},
)
# Or get the underlying httpx client to modify directly with client.get_httpx_client() or client.get_async_httpx_client()
```
For example, the previous code snippet shows how to define a custom request and response event hook using the `httpx_args` parameter of the `AuthenticatedClient` class. The `httpx_args` parameter is used to pass additional arguments to the underlying `httpx.Client` or `httpx.AsyncClient`. In this case, the `event_hooks` argument is used to specify custom functions that will be called before and after each HTTP request and response.
You can even set the httpx client directly, but beware that this will override any existing settings (e.g., `base_url`).
Raw data
{
"_id": null,
"home_page": null,
"name": "ovhai",
"maintainer": null,
"docs_url": null,
"requires_python": "<4,>=3.8",
"maintainer_email": null,
"keywords": "ovhai ovh AI",
"author": null,
"author_email": null,
"download_url": "https://files.pythonhosted.org/packages/f5/fd/53798b5993b736b0c3142383f792c39b093c36edfd625a2087b1366c1269/ovhai-0.1a4.tar.gz",
"platform": null,
"description": "# ovhai - OVHcloud's AI Python SDK\n\nThe `ovhai` library is a Python client that allows developers to easily use the [OVHcloud AI API](https://gra.training.ai.cloud.ovh.net/#/). With this SDK, you can run, manage, automate your notebooks, training and deployments in the cloud using OVHcloud's AI products (AI Notebooks, AI Training, AI Deploy).\n\n## \u26a0\ufe0f Alpha Warning \u26a0\ufe0f\n\nThis package is currently in the alpha phase of development. The APIs and functionalities of the package may not be fully tested.\n\n## Installation\nTo install the SDK, run the following command:\n\n```bash\npip install ovhai\n```\n\nThe SDK requires Python 3.8 or higher. For information about how to update your Python version, see the [official Python documentation](https://www.python.org/downloads/).\n\n## Getting started - Example Usage\n\nOnce you've installed the AI SDK, you can import it to use OVHcloud's AI products using the [API](https://gra.training.ai.cloud.ovh.net/#/).\n\nYou can start by the client creation:\n\n```python\nfrom ovhai import AuthenticatedClient\n\nclient = AuthenticatedClient(\n base_url=\"https://gra.training.ai.cloud.ovh.net\", \n token=\"YOUR_AI_TOKEN\",\n)\n```\n\nThe token used to create the client can be created via the OVHcloud Control Panel (UI), from the AI Dashboard.\n\nOnce your client is defined, you can call an endpoint:\n\n```python\nfrom ovhai.models import Me\nfrom ovhai.api.me import me\nfrom ovhai.ovhai_types import Response\n\nwith client as client:\n res: Me = me.sync(client=client)\n print(res)\n \n # or if you need more info (e.g. status_code)\n #response: Response[Me] = me.sync_detailed(client=client)\n #print(response)\n```\n\nOr do the same thing with an async version:\n\n```python\nimport asyncio\nfrom ovhai.models import Me\nfrom ovhai.api.me import me\nfrom ovhai.ovhai_types import Response\n\nasync def main(client):\n res: Me = await me.asyncio(client=client)\n print(res)\n \n # or if you need more info (e.g. status_code)\n #response: Response[Me] = await me.asyncio_detailed(client=client)\n #print(response)\n \n# Run the main function asynchronously\nasyncio.run(main(client=client))\n```\n\nIn the `ovhai/api` [folder](https://github.com/ovh/ovhai-python-sdk/tree/main/ovhai/api), you will find all the endpoints you can call up. They are grouped by folder according to their purpose (`notebook`, `job`, `app`, `...`). \n\nFor example, to launch a notebook, you need to import the [`notebook_new` file](https://github.com/ovh/ovhai-python-sdk/blob/master/ovhai/api/notebook/notebook_new.py), located at `/ovhai/api/notebook`. You will also need to import the objects linked to this endpoint (those mentioned in the python file), since you will manipulate them (`Notebook` and `NotebookSpec` here, in addition to the classic `AuthenticatedClient` and `Response`). This will allow you to launch your first notebook using the `ovhai` python library, based on your specifications:\n\n```python\nfrom ovhai import AuthenticatedClient\nfrom ovhai.api.notebook import notebook_new\nfrom ovhai.models import NotebookSpec, Notebook\nfrom ovhai.ovhai_types import Response\n\nclient = AuthenticatedClient(\n base_url=\"https://gra.training.ai.cloud.ovh.net\",\n token=\"YOUR_AI_TOKEN\",\n)\n\n# Define notebook parameters\neditor_id = \"jupyterlab\"\nframework_id = \"conda\"\nframework_version = \"conda-py39-cudaDevel11.8-v22-4\"\nnb_cpu = 2\n\n# Create the notebook creation request\nnotebook_specs = {\n \"env\": {\"editorId\": editor_id, \"frameworkId\": framework_id, \"frameworkVersion\": framework_version},\n \"resources\": {\"cpu\": nb_cpu},\n}\n\nwith client as client:\n response: Response[Notebook] = notebook_new.sync_detailed(\n client=client, body=NotebookSpec.from_dict(notebook_specs)\n )\n print(response)\n```\n\nThe Response object returned will contain various information, including your notebook UUID.\n\n## Things to know\n\nEvery [OVHcloud's AI API](https://gra.training.ai.cloud.ovh.net/#/) endpoint has its dedicated Python module that comes with four functions:\n1. `sync`: Blocking request that returns parsed data (if successful) or `None`\n2. `sync_detailed`: Blocking request that always returns a `Request`, optionally with `parsed` set if the request was successful.\n3. `asyncio`: Like `sync` but async instead of blocking\n4. `asyncio_detailed`: Like `sync_detailed` but async instead of blocking\n\nTo implement the call you want, find the folder and then the file corresponding to your needs in the [api folder](https://github.com/ovh/ovhai-python-sdk/tree/master/ovhai/api). Then choose the method that suits you best from the four mentioned above. Then import the objects you need to use this method.\n\n## Advanced customizations\n\nThere are more settings on the generated `AuthenticatedClient` class which let you control more runtime behavior, check out the [docstring on that class](https://github.com/ovh/ovhai-python-sdk/blob/master/ovhai/client.py) for more info. You can also customize the underlying `httpx.Client` or `httpx.AsyncClient` (depending on your use-case):\n\n```python\nfrom ovhai import AuthenticatedClient\n\ndef log_request(request):\n print(f\"Request event hook: {request.method} {request.url} - Waiting for response\")\n\ndef log_response(response):\n request = response.request\n print(f\"Response event hook: {request.method} {request.url} - Status {response.status_code}\")\n\nclient = AuthenticatedClient(\n base_url=\"https://gra.training.ai.cloud.ovh.net\",\n token=\"YOUR_AI_TOKEN\",\n httpx_args={\"event_hooks\": {\"request\": [log_request], \"response\": [log_response]}},\n)\n\n# Or get the underlying httpx client to modify directly with client.get_httpx_client() or client.get_async_httpx_client()\n```\n\nFor example, the previous code snippet shows how to define a custom request and response event hook using the `httpx_args` parameter of the `AuthenticatedClient` class. The `httpx_args` parameter is used to pass additional arguments to the underlying `httpx.Client` or `httpx.AsyncClient`. In this case, the `event_hooks` argument is used to specify custom functions that will be called before and after each HTTP request and response.\n\nYou can even set the httpx client directly, but beware that this will override any existing settings (e.g., `base_url`).",
"bugtrack_url": null,
"license": "Apache-2.0",
"summary": "A client library for accessing OVHcloud's AI Solutions",
"version": "0.1a4",
"project_urls": {
"Repository": "https://github.com/ovh/ovhai-python-sdk"
},
"split_keywords": [
"ovhai",
"ovh",
"ai"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "f5fd53798b5993b736b0c3142383f792c39b093c36edfd625a2087b1366c1269",
"md5": "35be4b798974eb5f13057c113504c6aa",
"sha256": "07fd92c2f53f7d6d5912afb7fa48b40bf72d99d406dfff6da7df234eb0658508"
},
"downloads": -1,
"filename": "ovhai-0.1a4.tar.gz",
"has_sig": false,
"md5_digest": "35be4b798974eb5f13057c113504c6aa",
"packagetype": "sdist",
"python_version": "source",
"requires_python": "<4,>=3.8",
"size": 79670,
"upload_time": "2024-05-28T14:04:51",
"upload_time_iso_8601": "2024-05-28T14:04:51.443808Z",
"url": "https://files.pythonhosted.org/packages/f5/fd/53798b5993b736b0c3142383f792c39b093c36edfd625a2087b1366c1269/ovhai-0.1a4.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2024-05-28 14:04:51",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "ovh",
"github_project": "ovhai-python-sdk",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"lcname": "ovhai"
}
