Examples: query, "exact match", wildcard*, wild?ard, wild*rd
Fuzzy search: cake~ (finds cakes, bake)
Term boost: "red velvet"^4, chocolate^2
Field grouping: tags:(+work -"fun-stuff")
Escaping: Escape characters +-&|!(){}[]^"~*?:\ with \, e.g. \+
Range search: properties.timestamp:[1587729413488 TO *] (inclusive), properties.title:{A TO Z}(excluding A and Z)
Combinations: chocolate AND vanilla, chocolate OR vanilla, (chocolate OR vanilla) NOT "vanilla pudding"
Field search: properties.title:"The Title" AND text
Answered
Hi All, Want To Ask, How I Can Debugg Logger Class. I Have Problems With Displaying Graphs On The Ui, Logger.Report_Text Works, But Any Scalar Or Plot Is Not Displayed, I Integrated Tensorboard, But There Are No Logs From It Either. There Are No Warnings

Hi all,
want to ask, How I can debugg Logger class. I have problems with displaying graphs on the UI, logger.report_text works, but any scalar or plot is not displayed, I integrated tensorboard, but there are no logs from it either. There are no warnings or errors from the logger, is there a way to manually send a request to UI? There is no such problem in other services, so I discarded everything related to devops because they are all pushed through one autoscaler. Also try task.get_logger() / Logger.current_logger()

  
  
Posted one year ago
Votes Newest

Answers 3


@<1523701070390366208:profile|CostlyOstrich36> self hosted and
class ClearmlLogger:

def __init__(self, task):

self.task = task
self.task_logger = task.get_logger()
self.task_logger.set_default_upload_destination(
' None ')
self.writer = SummaryWriter('runs')

def log_training(self, reconstruction_loss, learning_rate, iteration):
self.task.get_logger().report_scalar(
title='learning.rate',
series='learning.rate',
value=learning_rate,
iteration=iteration
)
self.writer.add_scalar('training learning.rate',
learning_rate,
iteration)

self.writer.add_scalar('training reconstruction_loss',
reconstruction_loss,
iteration)
self.writer.close()

self.task_logger.report_scalar(
title='training.reconstruction_loss',
series='training.reconstruction_loss',
value=reconstruction_loss,
iteration=iteration
)
I try different options and no reporting on the UI, but logger.task_logger.report_text works good with string. I think it can be lib version problem or some network.

  
  
Posted one year ago

Hi @<1603198163143888896:profile|LonelyKangaroo55> , how are you currently reporting, do you have a code snippet? Are you using the community server or a self hosted one?

  
  
Posted one year ago

Hey @<1603198163143888896:profile|LonelyKangaroo55> If you only use the summary writer, does it report properly to both TB and ClearML?

  
  
Posted one year ago