CorfuMV - fast, light and clear framework to versioning your TensorFlow models

Install now

• pip install https://github.com/averagepythonfan/corfumv/archive/main.zip

What is that all about?

1. We develope machine learning models
2. We need light and fast framework to versioning models
3. We use CorfuMV

Firstly, we need to up our CorfuMV server and MongoDB

# docker-compose.yml
version: "3.3"

services:
  mongodb:
    image: mongo:latest
    container_name: mongodb
    environment:
      MONGO_INITDB_ROOT_USERNAME: root
      MONGO_INITDB_ROOT_PASSWORD: secret
  corfumv:
    image: zhenyaover9000/corfumv:0.2.2.post0
    container_name: corfumv
    environment:
      MONGO: mongodb://root:secret@mongodb:27017
      SYNC: True
    ports:
      - "11000:11000"
    entrypoint: uvicorn corfumv.server:app --host 0.0.0.0 --port 11000

and then up services:

~$: docker compose up -d

Code example:

>>> from corfumv.client import CorfuClient
>>> client = CorfuClient("http://localhost:11000")
>>> exp = client.create_experiment(name="test_del", tags=["test", "delete"])
>>> exp.id
... '658432fe394f866bc0096605'
>>> md = client.create_model(name="test_md", tags=["test", "model"])
>>> md.id
... '6584339e394f866bc0096607'