nbox 0.1.6

ML Inference 🥶

Nbox

A library that makes using a host of models provided by the opensource community a lot more easier.

The entire purpose of this package is to make inference chill.

pip install nbox

Usage

import nbox

# As all these models come from the popular frameworks you use such as 
# torchvision, efficient_pytorch or hf.transformers
model = nbox.load("torchvision/mobilenetv2", pretrained = True)

# nbox makes inference the priority so you can use it
# pass it a list for batch inference 
out_single = model('cat.jpg')
out = model([Image.open('cat.jpg'), np.array(Image.open('cat.jpg'))])
tuple(out.shape) == (2, 1000)

# deploy the model to a managed kubernetes cluster on NimbleBox.ai
url_endpoint, nbx_api_key = model.deploy()

# or load a cloud infer model and use seamlessly
model = nbox.load(
  model_key_or_url = url_endpoint,
  nbx_api_key = nbx_api_key,
  category = "image"
)

# Deja-Vu!
out_single = model('cat.jpg')
out = model([Image.open('cat.jpg'), np.array(Image.open('cat.jpg'))])
tuple(out.shape) == (2, 1000)

Things for Repo

• Using poetry for proper package management as @cshubhamrao says.
  • Add new packages with poetry add <name>. Do not add torch, tensorflow and others, useless burden to manage those.
  • When pushing to pypi just do poetry build && poetry publish this manages all the things around
• Install pytest and then run pytest tests/ -v.
• Using black for formatting, VSCode to the moon.

License

The code in thist repo is licensed as BSD 3-Clause. Please check for individual repositories for licenses. Here are some of them:

Requirements

• rich

Model Sources

99% of the credit for nbox goes to the amazing people behind these projects:

• torch
• transformers
• efficientnet-pytorch