v1.0 Now Available

Monitor ML Training
in Real-Time.

Track metrics, visualize system performance, and compare experiments with just 3 lines of code. Built for PyTorch, TensorFlow, and Scikit-Learn.

View on GitHub

train.py

import ml_monitor

# 1. Init Run
ml_monitor.init({
  api_key="sk_live_...",
  project_id="my-project-id"
})

# 2. Log Metrics
for epoch in range(20):
  ...
  ml_monitor.log({
    "loss": 0.5,
    "accuracy": 0.92
  })

Live Visualization

Watch your loss curves and accuracy metrics update in real-time as your model trains.

System Monitoring

Automatically track GPU, CPU, and RAM usage to identify bottlenecks in your pipeline.

Run Comparison

Overlay metrics from multiple runs to compare hyperparameters and find the best model.

How to get started

Install the Package

pip install git+https://github.com/B4K2/Ml-monitoring-package.git

Create a Project

Add 3 Lines of Code

Import the package, initialize it with your key, and log metrics inside your training loop.

Monitor ML Training in Real-Time.