-
|
Hi, When I try to override hydra logging behaviour the path for Folder structure: ./
├── hydra
│ └── job_logging
│ └── custom_default.yaml
└── train
├── config.yaml
├── datamodule
│ ├── monai_holdout.yaml
│ ├── patch_func
│ │ └── label.yaml
│ ├── splitter
│ │ └── hold_out.yaml
│ ├── test_stage_transformations
│ │ └── dummy.yaml
│ └── train_stage_transformations
│ └── dummy.yaml
├── lightningmodule
│ ├── architecture
│ │ └── unet.yaml
│ ├── loss
│ │ └── generalized_dice.yaml
│ ├── metrics
│ │ └── dice.yaml
│ ├── optimizer_configuration
│ │ └── adam.yaml
│ ├── scheduler_configuration
│ │ └── step_lr.yaml
│ └── unet_like.yaml
└── trainer
└── standard.yamlI want to override the logging behaviour from defaults:
- _self_
- datamodule: monai_holdout
- lightningmodule: unet_like
- trainer: standard
- override hydra/job_logging: custom_default
hydra:
searchpath:
- file:///home/yere/Repositories/brain_tumor/configs
datamodule_type: monai_hold_out
lightningmodule_type: unet_likeMy main script: import hydra
from omegaconf import DictConfig, OmegaConf
from utils import datamodule_factory, lightningmodule_factory
@hydra.main(config_path="./configs/train", config_name="config")
def main(cfg: DictConfig):
"""
Perform training.
"""
OmegaConf.register_new_resolver("get_method", hydra.utils.get_method)
cfg = hydra.utils.instantiate(cfg, _recursive_=True)
datamodule_type = cfg.pop("datamodule_type")
lightningmodule_type = cfg.pop("lightningmodule_type")
datamodule = datamodule_factory(datamodule_type)(**cfg["datamodule"])
lightningmodule = lightningmodule_factory(lightningmodule_type)(
**cfg["lightningmodule"]
)
main()Then, I got the following error: |
Beta Was this translation helpful? Give feedback.
Answered by
Jasha10
May 31, 2022
Replies: 1 comment 1 reply
-
I think the problem is that |
Beta Was this translation helpful? Give feedback.
1 reply
-
|
Thanks for having so much patience hehe, and sorry for wasting your time. Yes that was the problem. I've modified the defaults:
- datamodule: brats2021
- lightningmodule: semantic_segmentator_unet
- trainer: config
- override hydra/job_logging: custom
- override lightningmodule/scheduler_configuration: null
- _self_
hydra:
searchpath:
- pkg://configs
datamodule_type: monai_hold_out
lightningmodule_type: semantic_segmentator
lightningmodule:
scheduler: null
datamodule:
batch_size: 1
configs/
├── data
│ ├── __init__.py
│ ├── __pycache__
│ │ ├── cfg_structure_folders.cpython-38.pyc
│ │ ├── __init__.cpython-38.pyc
│ │ └── structure_folders.cpython-38.pyc
│ └── structure_folders.py
├── hydra
│ └── job_logging
│ └── custom.yaml
└── train
├── config.yaml
├── datamodule
│ ├── brats2021.yaml
│ ├── patch_fn
│ │ ├── label_classed.yaml
│ │ └── testing.yaml
│ ├── split_fn
│ │ ├── hold_out.yaml
│ │ └── testing.yaml
│ ├── testing.yaml
│ ├── test_stage_transformations
│ │ ├── brats2021.yaml
│ │ └── testing.yaml
│ └── train_stage_transformations
│ ├── brats2021.yaml
│ └── testing.yaml
├── lightningmodule
│ ├── architecture
│ │ └── unet.yaml
│ ├── loss
│ │ └── generalized_dice.yaml
│ ├── loss_evaluator
│ │ └── simple_unet_like.yaml
│ ├── metrics
│ │ └── semantic_segmentation.yaml
│ ├── metrics_evaluator
│ │ └── simple_unet_like.yaml
│ ├── optimizer_configuration
│ │ └── adam.yaml
│ ├── postprocessing
│ │ └── dummy.yaml
│ ├── scheduler_configuration
│ │ └── step_lr.yaml
│ └── semantic_segmentator_unet.yaml
└── trainer
├── config.yaml
└── logger
└── wandb.yaml
|
Beta Was this translation helpful? Give feedback.
Answer selected by
YerePhy
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
I think the problem is that
config_pathis set to the directoryconfigs/train, but thehydra/job_logging/custom_default.yamlfile is not inside of theconfigs/traindirectory.