reading tensorboard data

[menashe soffer]

Hello

I am trying to get tensorboard data using the EventAccumulator API (for which from tensorboard.backend.event_processing.event_accumulator import EventAccumulator)

event_acc = EventAccumulator(path)

event_acc.Reload()

curve_name = event_acc.Tags()['tensors'][0]

ts, step, val = zip(*event_acc.Tensors(curve_name))

the problems is that no matter how many epochs have elapsed, I am getting no more than 10 points (not sure how the data is pruned).
Is there a default of returning only 10 data points, and is there a way to override this?

thanks