Sagemaker Estimators with Dockerfile

[celsofranssa]

I am using Sagemaker Estimators, such as shown in the code snippet below:

from sagemaker.pytorch import PyTorch
est = PyTorch(
    entry_point="train.py",
    source_dir="code",  # directory of your training script
    role=role,
    framework_version="1.5.0",
    py_version="py3",
    instance_type=instance_type,
    instance_count=1,
    volume_size=250,
    output_path=output_path,
    hyperparameters={"batch-size": 128, "epochs": 1, "learning-rate": 1e-3, "log-interval": 100},
)

However, it is still unclear how it handles (i) the Python dependencies (requirements.txt); (ii) the system dependencies (apt install ..). Is it possible to customize how the docker image is created (using a Dockerfile, for instance)?