Getting Started

Building your first model

It's time to build your first ML model on JFrog ML!

A model build is a trained, serialized and tested model instance, bundled with necessary dependencies that can be later deployed to production.

JFrog ML builds models on a scalable infrastructure that allows you to customize resources, whether using a pre-trained model or running live training of any size.

Creating a model

To get started, we need to create a new model and a new project on JFrog ML. Projects allow us to group and organize our models under one place.

Creating models and projects can be done either through the user interface or by using the Qwak CLI.

In this tutorial, we will be using the Qwak CLI.

1. Create the Credit Risk model under the Credit Risk Modeling project:

qwak models create "Credit Risk" --project "Credit Risk Modeling"

2. Generate a local example model with all the required files:

qwak models init --example credit_risk .

3. Remotely build the model on JFrog ML by pointing the qwak client to your local model directory and with the remote model ID. JFrog ML generates model IDs automatically by converting the model name to lowercase and removing any spaces

qwak models build ./credit_risk --model-id credit_risk --name "credit_risk_build_v1"

Which will show the following in the terminal:

# ✅ Fetching Model Code (0:00:00.16)
# ✅ Registering Qwak Build - 100% (0:00:07.79)
# 
# Build ID d12c188e-c6ad-4f91-9a74-9d18c1b18018 triggered remotely
# 
########### To follow build logs using CLI
# qwak models builds logs -b d12c188e-c6ad-4f91-9a74-9d18c1b18018 --follow
#
########### To follow build logs using Qwak platform
# https://app.qwak.ai/projects/b731b293-405a-491e-a17d-8c63c3d03017/credit_risk/build/d12c188e-c6ad-4f91-9a74-9d18c1b18018

Viewing build logs

To monitor the progress of your build on JFrog ML, view logs via the CLI or UI.

Logs provide real-time updates on the process, including any errors or warnings or debug information.

CLI: Run qwak models builds logs -b <build-id> --follow and replace build-id with your build id.

App: To view build logs via the UI, follow the links provided in the terminal when building a model.

qwak models build ./credit_risk --model-id “credit_risk” --name "credit_risk_build_v1"
 
# ✅ Fetching Model Code (0:00:00.16)
# ✅ Registering Qwak Build - 100% (0:00:07.79)

# Build ID d12c188e-c6ad-4f91-9a74-9d18c1b18018 triggered remotely

########### To follow build logs using CLI
# qwak models builds logs -b d12c188e-c6ad-4f91-9a74-9d18c1b18018 --follow

########### To follow build logs using Qwak platform
# https://app.qwak.ai/projects/b731b293-405a-491e-a17d-8c63c3d03017/credit_risk/build/d12c188e-c6ad-4f91-9a74-9d18c1b18018

👍

Using the --deploy flag will build and automatically deploy your model.

qwak models build ./credit_risk --model-id credit_risk --name "credit_risk_build_v1" --deploy

Which will show the following in the terminal:

# ✅ Fetching Model Code (0:00:00.19)
# ✅ Registering Qwak Build - 100% (0:00:06.17)
# ✅ Deploying - Waiting for build to finish (0:03:14.53)
# 
# Build ID a08faef3-dbb8-483d-8017-94b35f259c9c finished successfully and deployed
# 
########### To view the model using Qwak platform
# https://app.qwak.ai/projects/b731b293-405a-491e-a17d-8c63c3d03017/credit_risk

Deploying your model

After a successful build, our model can be deployed as a real-time inference endpoint on JFrog ML, ready to handle predictions. Copy the build ID from the previous build step and replace it with YOUR_BUILD_ID

qwak models deploy realtime --model-id credit_risk --build-id {YOUR_BUILD_ID}

After running the deployment command, you can expect to see the following output:

╒═══════════════╤══════════════════════════════════════╕
│ Environment   │ qwak_demo                            │
├───────────────┼──────────────────────────────────────┤
│ Model ID      │ credit_risk                          │
├───────────────┼──────────────────────────────────────┤
│ Build ID      │ f42af8a7-2942-459f-b768-a981b7098cb7 │
├───────────────┼──────────────────────────────────────┤
│ Deployment ID │ e2d9a66f-26be-4201-b1ea-f1ccf312d5d4 │
╘═══════════════╧══════════════════════════════════════╛

Deployment initiated successfully, Use --sync to wait for deployment to be ready.

Testing your model

After a successful model deployment, you can test your live inference endpoint.

The Qwak Python SDK includes a real-time client module which you have to separately install:

pip install qwak-inference

You can use it to run inference and predictions using your deployed real-time model:

from qwak_inference import RealTimeClient

# Update this variable with your account name (lower cased)
ENVIRONMENT = "your-account-name"

QWAK_MODEL_ID = 'credit_risk'

if __name__ == '__main__':
  feature_vector = [
    {
      "UserId" : "male",
      "Age" : 3,
      "Sex" : "male",
      "Job" : 2,
      "Housing" : "male",
      "Saving accounts" : "male",
      "Checking account" : "male",
      "Credit amount" : 54.2,
      "Duration" : 4,
      "Purpose" : "male",
      "Age_cat" : "male",
    }]

  client = RealTimeClient(model_id=QWAK_MODEL_ID,
                          environment=ENVIRONMENT)
  response = client.predict(feature_vector)
  print(response)  

Once you begin making predictions using the deployed model, you will be able to view relevant metrics in the Health dashboard on the Model Overview tab.

Querying model predictions

Querying model predictions is an essential step in the machine learning development process. With JFrog ML, it's easy to query your model's predictions and view relevant metrics.

1. Open your model page on the Qwak Application
2. Choose the Analytics tab
3. Click Run, and you'll see a table containing a row for every prediction made against the model.