Quick Start

Getting started with Deep Learning Studio!

Is a software tool that aims to simplify the creation of artificial intelligence and deep learning models. It is compatible with a number of open-source programming frameworks.
Below you'll find some helpful information to get you started. If you don’t see a topic covered here, try searching for it on our official website.
Note: If you have not installed the DLS 3.0 desktop version, please refer to the installation guide link.

Signing in

If you are an existing user of DLS, click on "Login with Deep Cognition". Enter username/email and password.
If you don't have account with Deep Cognition, "Sign up here" register your details with Deep Cognition. For detailed information of sign up refer this link.

Create project

In this guide, we will use a sample project of CIFAR-10 Object Recognition in Images
    Projects -> Sample projects
    Click on copy icon from sample projects tab
    Click on project title(it will take you to the data tab)
    Select Cifar10 - public dataset if not selected, by clicking on the drop-down of dataset
    Now go to the model tab
    Here you can see a neural network is already designed in model tab
Information
You can learn more about designing a neural network NN with DLS, which is super easy and one of a great feature of DLS.
    Go to Hyperparameters tab
    No of epochs is 10 here
    Now go to training tab and select the device (GPU/CPU) if not selected and click on Start Training button
    Observe the training progress and system usage
    Once training is finished, you can download/deploy your trained model. Here, we will deploy our trained model and then will do the inference
    Go to inference/deploy tab and enter the server name and click on deploy button. It will take a moment and after that deploy it by clicking on deploy button
    Your app is successfully deployed. Click on the link at bottom
    Here, drop an image and click on Submit to do inference

Congratulations!!
You have successfully completed the quick start. You can always explore more about the complete features.
Last modified 1yr ago