AimLogger error on new Pytorch-Lightning Release
See original GitHub issue🐛 Bug
Removed _convert_params
function in new pytorch-lightning release (1.6.0).
To reproduce
from aim.pytorch_lightning import AimLogger
trainer = pl.Trainer(logger=AimLogger(experiment='experiment_name'))
trainer.logger.log_hyperparams(hparams)
Expected behavior
- Successfully log the hyperparameters with no errors
Environment
- Aim Version (3.8.0)
- Python version (3.8.12)
- pip version (21.2.4)
- OS (e.g., Linux)
- Pytorch-lightning (1.6.0)
Additional context
Error output
Error executing job with overrides: ['experiment=dma_net', 'debug=default', 'log_dir=debug', 'data_dir=/media/haritsahm/DataStorage/dataset/cityscapes/cityscape_fo_segmentation/', 'logger=aim']
Traceback (most recent call last):
File "train.py", line 22, in main
return train(config)
File "/home/haritsahm/Documents/projects/deeplearning/pytorch-DMANet/src/training_pipeline.py", line 74, in train
utils.log_hyperparameters(
File "/opt/conda/lib/python3.8/site-packages/pytorch_lightning/utilities/rank_zero.py", line 32, in wrapped_fn
return fn(*args, **kwargs)
File "/home/haritsahm/Documents/projects/deeplearning/pytorch-DMANet/src/utils/__init__.py", line 143, in log_hyperparameters
trainer.logger.log_hyperparams(hparams)
File "/opt/conda/lib/python3.8/site-packages/pytorch_lightning/utilities/rank_zero.py", line 32, in wrapped_fn
return fn(*args, **kwargs)
File "/opt/conda/lib/python3.8/site-packages/aim/sdk/adapters/pytorch_lightning.py", line 70, in log_hyperparams
params = self._convert_params(params)
AttributeError: 'AimLogger' object has no attribute '_convert_params'
Set the environment variable HYDRA_FULL_ERROR=1 for a complete stack trace.
root@0e330cd4b21f:/home/haritsahm/Documents/projects/deeplearning/pytorch-DMANet#
Issue Analytics
- State:
- Created a year ago
- Reactions:1
- Comments:8 (6 by maintainers)
Top Results From Across the Web
Release 3.15.1 Gev Sogomonian, Gor Arakelyan et al. - Aim
Aim easily integrates with your favourite ML frameworks. Aim loggers give access to the aim.Run object instance via the experiment property.
Read more >Production - PyTorch Lightning
We're happy to release PyTorch Lightning 0.9.0 today, which contains many great new features, more bug fixes than any release we ever had,...
Read more >PyTorch Lightning and Aim - AimStack
When the training data is limited, transformations on the available data are used to synthesize new data. Mostly you can randomly choose the ......
Read more >Apple Silicon, Multi-GPU and more - Lightning 1.7
We're excited to announce the release of PyTorch Lightning 1.7 ⚡️ ... fault-tolerant training enabled, it is useful to load the latest ...
Read more >Top Related Medium Post
No results found
Top Related StackOverflow Question
No results found
Troubleshoot Live Code
Lightrun enables developers to add logs, metrics and snapshots to live code - no restarts or redeploys required.
Start FreeTop Related Reddit Thread
No results found
Top Related Hackernoon Post
No results found
Top Related Tweet
No results found
Top Related Dev.to Post
No results found
Top Related Hashnode Post
No results found
Top GitHub Comments
hey @mpds! Sorry for inconvenience, but that’s actually a different
dev
release that doesn’t have the fix included in it. Can I ask you to try out this one and let me know if it works properly?Hey folks! Glad to inform that the fix was shipped with the new
aim v3.9
release, please have a try, and let me know if everything works as expected. @mpds @haritsahm @djwessel