Picsellia
  • Picsellia
  • Getting started
    • Start using Picsellia
    • Create, annotate, review a Dataset
    • Create a new Dataset Version with merged labels
    • Train a custom Object Detection model
    • Train a custom Classification model
    • Deploy model in production (Tensorflow only)
    • Feedback loop - Send predictions from models to Datalake or Datasets
  • Data Management
    • Upload assets to your Lake
    • Edit Tags for Pictures
    • Create a Dataset
    • Add data to a Dataset
    • Create a new Dataset version
    • Configure your Labels
    • Import annotation from other Dataset version
  • Experiment Tracking
    • Initialize an experiment
    • Checkout an experiment
    • Log your results to Picsell.ia
    • Store your files to Picsell.ia
    • Evaluate your models
    • Retrieve and update your assets
    • Publish your model
    • Namespace
  • Hyperparameter tuning
    • Overview
    • Install Picsell CLI
    • Config
    • Launch your Hyperparameters tuning
  • Models
    • Model HUB
  • References
    • API Reference
    • Python SDK Reference
    • Python Training Reference
  • Organization
  • Website
Powered by GitBook
On this page

Was this helpful?

Models

One important part of Picsellia is the ability to store, share, re-use and deploy your trained models, here we will cover this topic.

PreviousLaunch your Hyperparameters tuningNextModel HUB

Last updated 3 years ago

Was this helpful?

The whole point of training AI algorithms is to obtain a powerful model that you will need to deploy to solve your problems.

But what you will discover during your many AI projects (or that you already discovered if you are already advanced in the field) is that you will never deploy only one model and be done with it.

Indeed, your models must evolve with the changes that occurs in real-world data over time.

This means that, to be efficient in training and deploying up-to-date models there are several steps that you need to master :

  • Store and version your models

  • Be able to reproduce results, meaning you must always know how your model was trained

  • Deploy in a scalable way

  • Monitor the model's predictions over time and avoid drift and bias

  • Share your models with your team/organization

It seems like a lot of work 🥵

Hopefully with Picsellia you are in good hands because we provide you with all the tools needed to perform all those steps seamlessly !

To know more, here are the pages/tutorials you need to check in order to leverage all the features we have developed to support your model development :

For example, our model HUB (and your Organization HUB) allows you to store, document, and share your trained models with all their files with your team 👇

To help you train and evaluate your models properly, you can use our experiment tracking system 👇

Deploy your models in only one-click and get your API endpoint 👇

Monitor your model predictions and also send them directly to your datasets for further exploitation 👇

Model HUB
Evaluate your models
Deploy model in production (Tensorflow only)
Feedback loop - Send predictions from models to Datalake or Datasets