Train and Deploy AutoGluon in the Cloud

AutoGluon-Cloud Documentation | AutoGluon Documentation

AutoGluon-Cloud lets you train and deploy state-of-the-art ML models for classification, regression, and time series forecasting on Amazon SageMaker. All it takes is a few lines of code; AutoGluon-Cloud handles the infrastructure, dependencies, and glue code for you.

💡 Why AutoGluon-Cloud?

• Works like local AutoGluon. Pass in DataFrames, get predictions back — as convenient as working locally, with the compute handled by AWS.
• No boilerplate. No training scripts, inference handlers, or serialization code to write and maintain.
• Official AWS containers. Everything runs in the AutoGluon Deep Learning Containers, maintained and security-patched by AWS.
• Sensible defaults, fully configurable. Under the hood it's just SageMaker running in your AWS account, so you stay in full control.

💾 Installation

pip install autogluon.cloud
autogluon-cloud bootstrap  # one-time setup for IAM role and S3 bucket

See the Setup tutorial for more details.

🚀 Foundation models

Zero-shot forecasts with a pretrained model like Chronos-2 — no training required. Full walkthrough.

from autogluon.cloud import TimeSeriesFoundationModel

# `data` can be a local path, S3 URL, or pandas DataFrame
data = "https://autogluon.s3.amazonaws.com/datasets/timeseries/m4_hourly_tiny/train.csv"

model = TimeSeriesFoundationModel("chronos-2")

# Batch prediction
predictions = model.predict(data=data, target="target", prediction_length=24)

# Real-time inference endpoint
endpoint = model.deploy()
predictions = endpoint.predict(data=data, target="target", prediction_length=24)
endpoint.delete_endpoint()

⚙️ Train your own predictor

Train an AutoGluon predictor on your data and serve it from SageMaker. Full walkthrough: time series, tabular.

from autogluon.cloud import TabularCloudPredictor

# `train_data` and `test_data` can be a local path, S3 URL, or pandas DataFrame
train_data = "https://autogluon.s3.amazonaws.com/datasets/Inc/train.csv"
test_data = "https://autogluon.s3.amazonaws.com/datasets/Inc/test.csv"

# Train
cloud_predictor = TabularCloudPredictor()
cloud_predictor.fit(
    train_data=train_data,
    predictor_init_args={"label": "class"},  # passed to TabularPredictor()
    predictor_fit_args={"time_limit": 120},  # passed to TabularPredictor.fit()
)

# Batch prediction
result = cloud_predictor.predict(test_data)

# Real-time inference endpoint
endpoint = cloud_predictor.deploy()
result = endpoint.predict(test_data)
endpoint.delete_endpoint()