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, When I Create A Task Using

Hi,
When i create a task using clearml-task specifying some hydra arguments the task correclty shows up in the web app with its arguments. But when i enqueue it (from the CLI or from the web app) no arguments are passed to the script.
For example running clearml-task --script train.py --args log.clearml=True train.epochs=200 clearml.save=True --project test --name test --skip-task-init (skip-task-init beacose i create the task if log.clearml=True ) in the web app it shows up like the screenshot but if I enqueue it and run it no arguments are passed to the script ( sys.argv is just the filename). I do understand that from clearml-task --help only argparse is supported
--args [ARGS [ARGS ...]] Arguments to pass to the remote execution, list of <argument>=<value> strings.Currently only argparse arguments are supported. Example: --args lr=0.003 batch_size=64but is there any workaround? Can you point me to the relevant files in clearml so i can add the support for hydra myself?

  
  
Posted one year ago
Votes Newest

Answers 5


This is the full print(cfg)
{'dataset': {'name': '<my-dataset-name>', 'path': '', 'target_shape': [128, 128]}, 'model': 'unet', 'models': {'unet': {'dim': 8, 'dim_mults': [1, 2, 4], 'num_blocks_per_stage': [2, 2, 2], 'num_self_attn_per_stage': [0, 0, 1], 'nested_unet_depths': [0, 0, 0], 'nested_unet_dim': 16, 'use_convnext': False, 'resnet_groups': 2, 'consolidate_upsample_fmaps': True, 'weight_standardize': False, 'attn_heads': 2, 'attn_dim_head': 16}}, 'train': {'accelerator': 'auto', 'gpus': 1, 'epochs': 100, 'batch_size': 8, 'lr': 0.001, 'seed': -1, 'log_every_n_steps': 1, 'checkpoints': False, 'enable_checkpoints': False, 'save': ''}, 'clearml': {'project': '<project-name>', 'task': '<task-name>', 'output_uri': 's3://<myhost>/<mybucket>', 'save': False, 'tags': None}, 'log': {'clearml': False, 'log_images_every_steps': 100}}

  
  
Posted one year ago

But not argument is passed to the scripts. Here i am printing sys.argv

  
  
Posted one year ago

When enqueued the configuration tab still shows the correct arguments

  
  
Posted one year ago

Yes it uses hydra and everything works fine without clearml. The script is similar to this one https://github.com/galatolofederico/lightning-template/blob/main/train.py

  
  
Posted one year ago

Hi TrickyFox41 , does you script contains hydra integration? do you import hydra? using some configuration file?

  
  
Posted one year ago