Deep Learning Studio 1.1.0 with Squeezenet, Wideresnet and more Features

Deep Learning Studio 1.1.0 with Squeezenet, Wideresnet and more Features

Code to GUI in seconds

We are pleased to introduce the new feature for importing your existing Keras project (based on Keras 1.2.2) in Deep Learning Studio for further development:

a) Import your project including model, dataset and hyperparameters; from single python file or a zip file containing multiple python files.
b) Open the imported project on Deep Learning Studio, configure the dataset properties and switch to Model Tab to continue further development of your project.


Model as Layer

You can import your Keras models into Deep Learning Studio as Layer. These layers can be used just like regular layers.

Pre-Trained Models

State of the art deep learning models are now available including SqueezeNet and WideResNet; in addition to existing ResNet, VGG and InceptionV3 pre-trained networks to give you big boost in your AI tasks.

We also have Densenet coming soon.

Copy Paste Layers

Now you can easily duplicate a layer by simple CTRL+C and CTRL+V key press. The copy operation also maintains the original layer properties in the new one.

Easy Searching

Now you can easily search the layers and drag them on to the canvas area.

Auto Checkpoints

This feature allows you to save the weights of training run on Best Accuracy or Lowest Loss on validation.

Resume Training

While training, you can start by loading the weights from previous runs, rather than training it from the scratch. Training runs can be compared in “Results Tab”.



Deep Cognition’s Announcements

Deep Cognition will be releasing Falcon TensorApp Server and TensorApp MarketPlace in Q4 2018.



  • 8330, Sterling Street, Irving, TX, USA 75063
  • +1-214-441-3517