Popular repositories Loading

1. orion orion Public

   An interference-aware scheduler for fine-grained GPU sharing

   Python 164 32

2. fmengine fmengine Public

   Utilities for Training Very Large Models

   Python 58 9

3. modyn modyn Public

   Modyn is a research-platform for training ML models on growing datasets.

   Python 53 9

4. sailor sailor Public

   AI model training on heterogeneous, geo-distributed resources

   Python 46 14

5. dirigent dirigent Public

   Dirigent: Lightweight Serverless Orchestration

   Go 44 7

6. OpenTela OpenTela Public

   OpenTela is a decentralized compute fabric for running machine learning applications.

   Jupyter Notebook 42 12