Skip to main content

​
Overview

The login functions provide the primary interface for authenticating users and creating SyftboxManager instances in different environments. All functions support both Data Owner (DO) and Data Scientist (DS) modes.

​
login()

Wrapper function that delegates to login_ds() for backwards compatibility. 
import syft_client as sc

client = sc.login(
    email="[email protected]",
    sync=True,
    load_peers=True
)
​
email
str | None
default:"None"
User email address. For Colab environments, automatically retrieved if not provided.
​
sync
bool
default:"True"
Whether to sync with Google Drive after login.
​
load_peers
bool
default:"True"
Whether to load peer list after login.
​
token_path
str | Path | None
default:"None"
Path to Google Drive token file. Required for Jupyter environments.
​
return
SyftboxManager
Configured SyftboxManager instance for interacting with the Syftbox ecosystem.

​
login_ds()

Login as a Data Scientist to access shared datasets and submit jobs. 
import syft_client as sc

# Colab environment
client = sc.login_ds(email="[email protected]")

# Jupyter environment
client = sc.login_ds(
    email="[email protected]",
    token_path="/path/to/token.json"
)

​
Parameters

​
email
str | None
default:"None"
User email address. Required for Jupyter, auto-detected in Colab.
​
sync
bool
default:"True"
Whether to perform initial sync with Google Drive after login.
​
load_peers
bool
default:"True"
Whether to load the peer list after login.
​
token_path
str | Path | None
default:"None"
Path to Google Drive OAuth token file. Required for Jupyter environments. Can also be set via settings.token_path configuration.

​
Returns

​
return
SyftboxManager
Configured SyftboxManager instance in Data Scientist mode with:
  • Access to shared datasets via client.datasets
  • Ability to submit jobs via client.submit_bash_job() and client.submit_python_job()
  • Peer management via client.peers

​
Behavior

Environment Detection:
  • Automatically detects Colab or Jupyter environment
  • In Colab: Uses /content/SyftBox_{email} as default folder
  • In Jupyter: Uses ~/SyftBox_{email} as default folder
Initialization:
  1. Creates SyftboxManager with only_ds=True flag
  2. Optionally syncs to download latest files
  3. Optionally loads peer connections
  4. Prints connection status

​
Raises

  • ValueError: If email is required but not provided
  • NotImplementedError: If Jupyter login attempted without token_path
  • ValueError: If environment is not supported (not Colab or Jupyter)

​
Example

import syft_client as sc

# Login as Data Scientist
client = sc.login_ds(email="[email protected]")

# Access shared datasets
for dataset in client.datasets.get_all():
    print(f"Dataset: {dataset.name} from {dataset.owner}")

# Submit a job to a Data Owner
client.submit_python_job(
    user="[email protected]",
    script="analysis.py",
    description="Run analysis on private data"
)

​
login_do()

Login as a Data Owner to manage datasets, approve jobs, and control access. 
import syft_client as sc

# Colab environment
client = sc.login_do(email="[email protected]")

# Jupyter environment  
client = sc.login_do(
    email="[email protected]",
    token_path="/path/to/token.json"
)

​
Parameters

​
email
str | None
default:"None"
User email address. Required for Jupyter, auto-detected in Colab.
​
sync
bool
default:"True"
Whether to perform initial sync with Google Drive after login.
​
load_peers
bool
default:"True"
Whether to load peer list and requests after login.
​
token_path
str | Path | None
default:"None"
Path to Google Drive OAuth token file. Required for Jupyter environments.

​
Returns

​
return
SyftboxManager
Configured SyftboxManager instance in Data Owner mode with:
  • Dataset creation and management via client.create_dataset(), client.delete_dataset()
  • Job approval and execution via client.process_approved_jobs()
  • Peer request management via client.approve_peer_request(), client.reject_peer_request()
  • Checkpoint creation for efficient syncing via client.create_checkpoint()

​
Behavior

Environment Detection:
  • Automatically detects Colab or Jupyter environment
  • In Colab: Uses /content/SyftBox_{email} as default folder
  • In Jupyter: Uses ~/SyftBox_{email} as default folder
Initialization:
  1. Creates SyftboxManager with only_datasite_owner=True flag
  2. Optionally syncs to process pending requests and events
  3. Optionally loads peer list including pending requests
  4. Prints connection status

​
Raises

  • ValueError: If email is required but not provided
  • NotImplementedError: If Jupyter login attempted without token_path
  • ValueError: If environment is not supported (not Colab or Jupyter)

​
Example

import syft_client as sc

# Login as Data Owner
client = sc.login_do(email="[email protected]")

# Create and share a dataset
dataset = client.create_dataset(
    name="medical_data",
    mock_path="mock_data.csv",
    private_path="private_data.csv",
    users="any"
)

# Review and approve peer requests
for peer in client.peers:
    if peer.state == "pending":
        print(f"Peer request from: {peer.email}")
        client.approve_peer_request(peer.email)

# Process approved jobs
client.process_approved_jobs(stream_output=True)

​
Environment Configuration

​
Colab

  • Email auto-detected from Google account
  • SyftBox folder: /content/SyftBox_{email}
  • OAuth tokens managed automatically
  • No token_path required

​
Jupyter

  • Email must be provided
  • SyftBox folder: ~/SyftBox_{email}
  • Requires token_path to OAuth credentials
  • Token can be set via:
    • Function parameter: token_path="/path/to/token.json"
    • Config setting: settings.token_path = "/path/to/token.json"

​
See Also

Build docs developers (and LLMs) love

Get started for freeTalk to us