sieve.function

def function(
  name: str,
  gpu: bool = False,
  python_packages: List[str] = [],
  python_version: str = "3.8",
  system_packages: List[str] = [],
  cuda_version: str = None,
  machine_type: str = None,
  iterator_input: bool = False,
  persist_output: bool = False,
  run_commands: List[str] = []
)

A sieve.function is a Python function that can be deployed to Sieve. The only diference between a sieve.function and a sieve.Model is that a sieve.Model contains both a __setup__ and a __predict__ function, while a sieve.function only contains a single function. Models are typically used for heavy machine learning models that require a longer setup step.

​

Dummy Example

import sieve

@sieve.function(name="sample_function")
def sample_function(echo: str) -> str:
  return echo

​

Video Splitting Example

The example below shows how to use sieve.function to split a video into frames.

import sieve

@sieve.function(
  name="video-splitter",
  gpu = False,
  python_packages=[
    "ffmpeg-python==0.2.0"
  ],
  system_packages=["libgl1-mesa-glx", "libglib2.0-0", "ffmpeg"],
  python_version="3.8"
)
def VideoSplitter(video: sieve.Video) -> sieve.Image:
  # use ffmpeg to extract all frames in video as bmp files and return the path to the folder
  import tempfile
  temp_dir = tempfile.mkdtemp()

  import subprocess
  subprocess.call([
    'ffmpeg',
    '-i', video.path,
    f'{temp_dir}/%09d.jpg'
  ])
  import os
  filenames = os.listdir(temp_dir)
  filenames.sort()
  for i, filename in enumerate(filenames):
    print(os.path.join(temp_dir, filename), i)
    yield sieve.Image(path=os.path.join(temp_dir, filename), frame_number=i, fps=video.fps)

​

Arguments

• name Name of the function.
• gpu Whether the function should run on a GPU. Defaults to False.
• python_packages List of Python packages to install. Defaults to [].
• python_version Python version to use. Defaults to "3.8".
• system_packages List of system packages to install. Defaults to [].
• cuda_version CUDA version to use. Defaults to None.
• machine_type Machine type to use. Defaults to None.
• iterator_input Whether the function takes in an iterator. Defaults to False.
• persist_output Whether the function’s output should be persisted. Defaults to False.
• run_commands List of commands to run when the function is built. Defaults to [].

​

name

def name(self) -> str:

Name of the function.

​

gpu

def gpu(self) -> bool:

Whether the function should run on a GPU.

​

python_packages

def python_packages(self) -> List[str]:

List of Python packages to install.

​

python_version

def python_version(self) -> str:

Python version to use.

​

system_packages

def system_packages(self) -> List[str]:

List of system packages to install.

​

cuda_version

def cuda_version(self) -> str:

CUDA version to use.

​

machine_type

def machine_type(self) -> str:

Machine type to use.

​

iterator_input

def iterator_input(self) -> bool:

Whether the function takes in an iterator.

​

persist_output

def persist_output(self) -> bool:

Whether the function’s output should be persisted to permanent storage for later retrieval.

​

run_commands

def run_commands(self) -> List[str]:

List of commands to run when the function is built.