backend.ai-client 1.0.2

Backend.AI API Client Library

The official API client library for Backend.AI

Usage

You should set the access key and secret key as environment variables to use the API. Grab your keypair from cloud.backend.ai or your cluster admin.

export BACKEND_ACCESS_KEY=...
export BACKEND_SECRET_KEY=...

# optional (for local clusters)
export BACKEND_ENDPOINT="https://my-precious-cluster/"

Command-line Interface

Use ai.backend.client.cli module with run command.

To run the code specified in the command line directly, use -c option to pass the code string.

$ python -m ai.backend.client.cli run python3 -c "print('hello world')"
∙ Client session token: d3694dda6e5a9f1e5c718e07bba291a9
✔ Kernel (ID: zuF1OzMIhFknyjUl7Apbvg) is ready.
hello world
✔ Cleaned up the kernel.

You can even run a C code on-the-fly. (Note that we put a dollar sign before the single-quoted code argument so that the shell to interpret '\n' as actual newlines.)

$ python -m ai.backend.client.cli run c -c $'#include <stdio.h>\nint main() {printf("hello world\\n");}'
∙ Client session token: abc06ee5e03fce60c51148c6d2dd6126
✔ Kernel (ID: d1YXvee-uAJTx4AKYyeksA) is ready.
hello world
✔ Cleaned up the kernel.

For larger programs, you may upload multiple files and then build & execute them. The below is a simple example to run a sample C program.

$ git clone https://gist.github.com/achimnol/df464c6a3fe05b21e9b06d5b80e986c5 c-example
Cloning into 'c-example'...
Unpacking objects: 100% (5/5), done.
$ cd c-example
$ python -m ai.backend.client.cli run c main.c mylib.c mylib.h
∙ Client session token: 1c352a572bc751a81d1f812186093c47
✔ Kernel (ID: kJ6CgWR7Tz3_v2WsDHOwLQ) is ready.
✔ Uploading done.
✔ Build finished.
myvalue is 42
your name? LABLUP
hello, LABLUP!
✔ Cleaned up the kernel.

Please refer the --help manual provided by the run command.

You may use a shortcut command lcc and lpython instead of typing the full Python module path like:

$ lcc main.c mylib.c mylib.h

To use API development tools such as GraphiQL for the admin API, run an insecure local API proxy. This will attach all the necessary authorization headers to your vanilla HTTP API requests.

$ python -m ai.backend.client.cli proxy
∙ Starting an insecure API proxy at http://localhost:8084

Synchronous API

from ai.backend.client import Kernel

kern = Kernel.get_or_create('lua5', client_token='abc')
result = kern.execute('print("hello world")', mode='query')
print(result['console'])
kern.destroy()

You need to take care of client_token because it determines whether to reuse kernel sessions or not. Sorna cloud has a timeout so that it terminates long-idle kernel sessions, but within the timeout, any kernel creation requests with the same client_token let Sorna cloud to reuse the kernel.

Asynchronous API

import asyncio
from ai.backend.client.asyncio import AsyncKernel

async def main():
    kern = await AsyncKernel.get_or_create('lua5', client_token='abc')
    result = await kern.execute('print("hello world")', mode='query')
    print(result['console'])
    await kern.destroy()

loop = asyncio.get_event_loop()
try:
    loop.run_until_complete(main())
finally:
    loop.close()

All the methods of AsyncKernel objects are exactly same to the synchronous version, except that they are coroutines.

Additionally, AsyncKernel offers async-only method stream_pty(). It returns a StreamPty object which allows you to access a pseudo-tty of the kernel. StreamPty works like an async-generator and provides methods to send stdin inputs as well as resize the terminal.