How to Generate Experiment Tracking Run IDs in Python

Generate unique experiment run IDs with timestamps and random suffixes for tracking ML pipeline executions.

Easy Python 3.9+ Aug 9, 2026 ML engineering pipelines 13 views 0 copies

run-ids experiment-tracking ml-pipelines unique-ids timestamps

Python code

14 lines

Python 3.9+

import random
import string
import time

def generate_run_id(prefix="exp"):
    timestamp = time.strftime("%Y%m%d_%H%M%S")
    suffix = "".join(random.choices(string.ascii_lowercase + string.digits, k=6))
    return f"{prefix}_{timestamp}_{suffix}"

if __name__ == "__main__":
    # Simulate tracking three experiment runs
    for i in range(3):
        run_id = generate_run_id()
        print(f"Run {i + 1}: {run_id}")

Output

stdout

Run 1: exp_20240315_143022_ab3x9z
Run 2: exp_20240315_143022_k7p2mq
Run 3: exp_20240315_143022_t5w8jr

How it works

The generate_run_id function combines a timestamp from time.strftime with a random 6-character alphanumeric suffix. This creates human-readable, chronologically sortable IDs unique enough for most experiment tracking needs. The timestamp ensures IDs are roughly ordered by creation time, while the random suffix reduces collision probability. Using random.choices with string.ascii_lowercase + string.digits avoids ambiguous characters and keeps IDs compact.

Common mistakes

Forgetting that two runs in the same second can collide without the random suffix
Using `time.time()` directly instead of `strftime` which produces ugly unreadable IDs
Not including a prefix to distinguish experiment types in large tracking systems

Variations

Use `uuid.uuid4().hex[:6]` instead of `random.choices` for stronger uniqueness guarantees
Format timestamp as `%Y%m%d%H%M%S%f` (with microseconds) to reduce collision risk further

Real-world use cases

Logging each training run into MLflow or Weights & Biases with a unique identifier for later comparison.
Storing model artifacts and checkpoints in cloud storage with run-specific keys for easy retrieval.
Tracking hyperparameter sweeps where each configuration needs a distinct ID to link results back.

How to Generate Experiment Tracking Run IDs in Python

Python code

Output

How it works

Common mistakes

Variations

Real-world use cases

More from ML engineering pipelines

Tutorials

Quizzes

Python code

Output

How it works

Common mistakes

Variations

Real-world use cases

More from ML engineering pipelines

Keep learning

Tutorials

Quizzes