Logging metrics#
A metric can be accuracy, MSE, or any numerical value. You can log scores and metrics as
- single values, with
=
assignment. - series of values, with the
append()
method. Eachappend()
call adds a value to the field.
append()
replaces log()
As of neptune-client 0.16.14
, append()
and extend()
are the preferred methods for logging series of values.
You can upgrade your installation with pip install -U neptune-client
or continue using log()
.
Neptune displays all value series as charts.
Example
# Log scores (single value)
run["score"] = 0.97
run["test/acc"] = 0.97
# Log metrics (series of values)
for epoch in range(100):
# your training loop
acc = ...
loss = ...
metric = ...
run["train/accuracy"].append(acc)
run["train/loss"].append(loss)
run["metric"].append(metric)
Tip
In the runs table, you can group and sort runs by the value of any field.
Setting custom index values#
You can specify custom index values with the step
argument:
The entries logged for step
must be strictly increasing Int or Float values.
This is effectively like setting custom values for the x axis of a chart.