File operations - Weights & Biases Documentation

Serverless Sandboxes is in public preview.

Use file operations to share data between your local environment and a sandbox. You can read files from a sandbox, write files to it, or mount local files and directories into it. For example, you can write a Python script to a sandbox, run it, and read the output file back to your local environment. You can also mount a directory of training data into a sandbox for a machine learning job.

Choose the right file access methodMount files or directories when you want the sandbox to access local data without copying it.Read and write files when you want to transfer smaller files between your local environment and the sandbox, or when you want to save sandbox output locally.

Write a file to the sandbox

Write a file to the sandbox when you need to upload inputs your sandbox code uses, such as scripts, configs, or data files. Transfer a file from your local environment to the sandbox using the Sandbox.write_file() method.

from pathlib import Path
from wandb.sandbox import Sandbox

# Path of the local file you want to write to the sandbox
text_file = Path("hello.txt")

with Sandbox.run() as sandbox:
    # Write a file to the sandbox
    sandbox.write_file("hello.txt", text_file.read_bytes()).result()

See the Sandbox class reference documentation for a full list of parameters and options for Sandbox.write_file().

Read a file from the sandbox

Save a file from the sandbox to your local environment using the Sandbox.read_file() method.

from pathlib import Path
from wandb.sandbox import Sandbox

with Sandbox.run() as sandbox:
    # Demo creates a file in the sandbox to read back
    sandbox.exec(["sh", "-c", "echo 'Hello, world.' > hello.txt"]).result()

    # Save the file on local machine (not sandbox)
    content = sandbox.read_file("hello.txt").result()  # b"Hello, world.\n"
    Path("hello.txt").write_bytes(content)

See the Sandbox class reference documentation for a full list of parameters and options for Sandbox.read_file().

Mount a file or directory

Use mounted files to provide local files to the sandbox at creation time. Unlike Sandbox.write_file(), which transfers files to a running sandbox, mounted files are available as soon as the sandbox starts. Mounted files appear in the sandbox at the path you specify.

Mounted files are read-only in the sandbox. If you need to modify files in the sandbox, use Sandbox.write_file() instead.

The following code snippet mounts the local files train.py and requirements.txt to the sandbox root directory. The sandbox installs dependencies from requirements.txt and then runs train.py.

from pathlib import Path
from wandb.sandbox import Sandbox, NetworkOptions

mounted_files = [
    {"mount_path": "requirements.txt", "file_content": Path("requirements.txt").read_bytes()},
    {"mount_path": "train.py", "file_content": Path("train.py").read_bytes()},
        ] 

print("Starting sandbox...")
with Sandbox.run(mounted_files=mounted_files) as sandbox:

    # The mounted files are available in the sandbox at the specified mount paths
    print("Installing dependencies...")
    result = sandbox.exec(["pip", "install", "-r", "requirements.txt"], check=True).result()
    print(result.stdout)

    print("Running script...")
    result = sandbox.exec(["python", "train.py"]).result()
    print(result.stdout)

​Write a file to the sandbox

​Read a file from the sandbox

​Mount a file or directory

Write a file to the sandbox

Read a file from the sandbox

Mount a file or directory