Training Monitor Python Packages

epokio

Phone alerts when an ML training run stalls, hits NaN or finishes, from the logs it already writes. No code changes, no account.

delve

PyTorch model training and layer saturation monitor

runboard

Live, self-hosted W&B-style training dashboards for any cluster or VM, open in your browser without SSH. Zero dependencies.

messenger-logger-callback

messenger-logger-callback — Send ML training logs to Telegram. Standalone logger and Hugging Face Trainer callback. pip install messenger-logger-callback

ltui

TensorBoard-like terminal monitor for PyTorch Lightning metrics, CSVLogger runs, and SSH training.

trainpulse

Lightweight training health monitor. Detect loss spikes, gradient explosions, NaN, and plateaus — 2 lines of code, no server, no signup.

lightning-metrics-tui

TensorBoard-like terminal monitor for PyTorch Lightning metrics, CSVLogger runs, and SSH training.