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
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))
]
)
See notebooks/
Run
make dev-install
make test
Original livelossplot
: https://github.com/stared/livelossplot