Skip to content

Experiments commands

List experiments

ml experiments list 

List the experiments under specified project.

External metrics

ml lets you send metrics after your experiment has run!

There are 3 levels for metrics: experiment, epoch and iteration metrics. To send experiment metrics, you can specify the project ID and experiment ID.

ml experiments update_metrics  \
    --experimentId myExperimentID \
    --metrics '{"ex_cost": 99}'

If you have a model's weights hash at a certain epoch or iteration of the experiment, you can give the hash and ml will find the experiment for you. Let say your model's weights hash is 324e16b5e, you can send the experiment metrics with.

ml experiments update_metrics \
    --weightsHash 324e16b5e --metrics '{"ex_cost": 99}'

Similarly, if you wish to attach the metric to a specific iteration you can use the update_model_metrics command. ml will automatiicaly identify the right iteration from the weights hash and inject the metrics to all relevant iterations.

ml experiments update_model_metrics \
    --weightsHash 324e16b5e --metrics '{"ex_cost": 99}'

External charts

ml provides a convenient way to submit external charts for experiments.

The charts can be single line charts (having single Y value per X point) or multi point chart (having multiple Y values per X point) Experiment can be identified by the project and experiment Ids or by model weights hash (see above).

Send a single point chart for experiment 1 in project 123.

ml experiments updateChart  --experimentId 1 \
    --chartName "my chart name" --chartScope train \
    --chartX '[0.1,0.5,0.8]' --chartY '[0.9,0.5,0.2]' \
    --chartXName "X" --chartYName "Y"

Send a multi point chart for the experiment assisated with weightsHash aa918d88a21eedbf6.

ml experiments updateChart --weightsHash aa918d88a21eedbf6  \
    --chartName ililil --chartScope train  \
    --chartX '[4,8,12]' --chartY '[[1,2,4],[2,4,8],[3,6,12]]' \
    --chartXName "Precision" --chartYName "Recall"