For detailed technical documentation of Kronfluence, please refer to the Technical Documentation page.
To run all examples, install the necessary packages:
pip install -r requirements.txt
Alternatively, navigate to each example folder and run pip install -r requirements.txt
.
Our examples cover the following tasks:
Task | Example Datasets |
---|---|
Regression | UCI |
Image Classification | CIFAR-10 & ImageNet |
Text Classification | GLUE |
Multiple-Choice | SWAG |
Summarization | CNN/DailyMail |
Language Modeling | WikiText-2 & OpenWebText |
These examples demonstrate various use cases of Kronfluence, including the usage of AMP (Automatic Mixed Precision) and DDP (Distributed Data Parallel). Many examples aim to replicate the settings used in our paper. If you would like to see more examples added to this repository, please leave an issue.