https://github.com/tsenst/lightning-experiments-logger
A SageMaker Experiment logger class for PyTorch Lightning
https://github.com/tsenst/lightning-experiments-logger
deep-learning experiment-tracking logger machine-learning mlops python pytorch pytorch-lightning
Last synced: 3 months ago
JSON representation
A SageMaker Experiment logger class for PyTorch Lightning
- Host: GitHub
- URL: https://github.com/tsenst/lightning-experiments-logger
- Owner: tsenst
- License: other
- Created: 2023-10-08T12:14:28.000Z (over 2 years ago)
- Default Branch: main
- Last Pushed: 2023-11-03T19:43:58.000Z (over 2 years ago)
- Last Synced: 2025-10-01T17:44:04.861Z (8 months ago)
- Topics: deep-learning, experiment-tracking, logger, machine-learning, mlops, python, pytorch, pytorch-lightning
- Language: Python
- Homepage: https://medium.com/idealo-tech-blog/experiment-tracking-with-aws-sagemaker-and-pytorch-lightning-68b22fd4deee
- Size: 2.98 MB
- Stars: 6
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
[](https://pypi.org/project/sagemaker-experiments-logger/)
[](https://tsenst.github.io/lightning-experiments-logger/index.html)
# SagemakerExperimentsLogger
SagemakerExperimentsLogger provides a simple way to log experimental data such as hyperparameter settings and evaluation metrics via [AWS SageMaker Experiments API](https://aws.amazon.com/blogs/aws/amazon-sagemaker-experiments-organize-track-and-compare-your-machine-learning-trainings/). It can be easily integration into the concept of [Pytorch Lightning Trainer class](https://lightning.ai/docs/pytorch/stable/common/trainer.html).
For detailed documentation, including the API reference, see [Read the Docs](https://tsenst.github.io/lightning-experiments-logger/index.html)
## Installation
You can install the latest (nightly) version with pip using ssh with
```bash
pip install sagemaker-experiments-logger
```
## Quickstart
The SageMaker Experiments logger can be easily applied by setup an own run context:
```Python
from pytorch_lightning import Trainer
from sagemaker.experiments.run import Run
from experiments_addon.logger import SagemakerExperimentsLogger
with Run(experiment_name="testExperiment", run_name="testRun1"):
logger = SagemakerExperimentsLogger()
trainer = Trainer(
logger=logger,
...
)
...
```
or by using an existing run context. For example in a SageMaker Training Step
```Python
from pytorch_lightning import Trainer
from experiments_addon.logger import SagemakerExperimentsLogger
logger = SagemakerExperimentsLogger(experiment_name="TestExp", run_name="TestRun")
trainer = Trainer(
logger=logger,
...
)
...
```
## Usage
Try [Tutorial Notebook](https://github.com/tsenst/lightning-experiments-logger/blob/main/example/tutorial.ipynb) to learn more about the usage of the [SagemakerExperimentsLogger](https://tsenst.github.io/lightning-experiments-logger/api_logger.html) class.
It is also worth to read the blog post: [Experiment Tracking With AWS SageMaker and PyTorch Lightning](https://medium.com/idealo-tech-blog/experiment-tracking-with-aws-sagemaker-and-pytorch-lightning-68b22fd4deee)
## Contributing
I welcome all contributions!
To file a bug or request a feature, please file a GitHub issue. Pull requests are welcome.
## License
This library is licensed under the Apache 2.0 License.