SKIL for Data Scientists

You data science it. We handle the rest.

Model Import

The Skymind Intelligence Layer (SKIL) provides routines for importing machine learning models originally configured and trained using Tensorflow, Keras, scikit-learn, and various other libraries.

Supported Formats

Workflow Automation

With connectors to widely used data science and big data tools, SKIL eliminates the data science and IT gap, accelerating time-to-market.

Model Management

          import skil_client
          uploads = client.upload("tensorflow_rnn.pb")
          new_model = DeployModel(name="recommender_rnn", scale=30, file_location=uploads[0].path)
          model = client.deploy_model(deployment_id, new_model)
          ndarray = INDArray(array=base64.b64encode(x_in))
          input = Prediction(id=1234, prediction=ndarray, needsPreProcessing=false)
          result = client.predict(input, "production", "recommender_rnn")
        

Distributed Training

  • SKIL will provision and orchestrate model training across a CPU or GPU cluster.
  • Built-in resource scheduling to maximize compute resources.

Model History Server

  • Track model performance over time to monitor for signs of data shift.
  • A/B test, audit, and monitor performance to uncover best performing model.

Inference

  • Support for real-time inference via REST or batch inference using Spark.
  • 1-click deploy to update models on the fly.