Welcome to modelz’s documentation!¶

modelz is a Python package for calling model prediction and management APIs that are on the modelz platform. We provide both Python SDK and CLI interface.

Installation¶

Install from PyPI

pip install modelz-py

Install from source

pip install git+https://github.com/tensorchord/modelz-py

Examples¶

Check modelz_docs for more details.

stable diffusion example¶

import modelz

APIKey = "mzi-abcdefg..."

cli = modelz.ModelzClient(deployment="abcdgefg...", key=APIKey, serde="msgpack")
cli.inference(params="A dog is running in the grass.").save_to_file("dog.jpg")

image bind example¶

import modelz


APIKey = "mzi-abcdefg..."

cli = modelz.ModelzClient(deployment="imagebind-XXX", key=APIKey)

input = {"model": "imagebind-text", "input": ["A dog", "doggery", "puppy"]},
resp = cli.inference(params=data, serde="msgpack")
embeddings = resp["data"]
for emb in embeddings:
   print(emb['embedding'])