Skip to main content

Plot Keras model losses and metrics live in Jupyter

Project description

PyPI version Build Status Coverage Status

History plot for Keras model training

history plot with batches

Plots training losses and metrics in Jupyter while training your keras model.

This project started as an attempt to extend livelossplot, but I realized that a rewrite was easier. Key improvements are:

  • keeps full history, including losses for each batch in each epoch
  • plots scatter of batch losses
  • plots arbitrary log entries, such a lr (learning rate)
  • simplifies codebase extendibility

Basic Usage

Install package from source:

pip install git+https://github.com/ig248/kerashistoryplot

Or from PyPi:

pip install kerashistoryplot

Import function:

from livehistoryplot.callbacks import PlotHistory

Use in callbacks:

model.fit(
    X, y,
    epochs=20,
    callbacks=[
        ReduceLROnPlateau(),
        PlotHistory(batches=True, n_cols=3, figsize=(15, 7))
    ]
)

More Examples

See notebooks/

Tests

Run

make dev-install
make test

References

Original livelossplot: https://github.com/stared/livelossplot

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

kerashistoryplot-0.0.10.tar.gz (70.2 kB view hashes)

Uploaded Source

Built Distribution

kerashistoryplot-0.0.10-py2.py3-none-any.whl (5.7 kB view hashes)

Uploaded Python 2 Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page