diff --git a/.zenodo.json b/.zenodo.json new file mode 100644 index 0000000..9765aa3 --- /dev/null +++ b/.zenodo.json @@ -0,0 +1,10 @@ +{ + "title": "batch-probe: GPU memory probing and thermal-aware CPU thread control", + "upload_type": "software", + "description": "Probes the largest OOM-safe GPU batch size (binary search with OOM recovery) and runs a Kalman-filtered, thermal-aware controller for CPU thread / GPU load to keep a node within its thermal envelope. Works with PyTorch, CuPy, JAX, or any GPU framework.", + "creators": [ + {"name": "Bond, Andrew H.", "affiliation": "San Jose State University", "orcid": "0009-0003-2599-6158"} + ], + "license": {"id": "MIT"}, + "keywords": ["pytorch", "cupy", "jax", "gpu", "memory", "batch-size", "oom", "cuda", "thermal", "threads"] +} diff --git a/CITATION.cff b/CITATION.cff new file mode 100644 index 0000000..f202a83 --- /dev/null +++ b/CITATION.cff @@ -0,0 +1,26 @@ +cff-version: 1.2.0 +message: "If you use batch-probe in your research, please cite it as below." +title: "batch-probe: GPU memory probing and thermal-aware CPU thread control" +type: software +authors: + - family-names: Bond + given-names: Andrew H. + email: andrew.bond@sjsu.edu + affiliation: San Jose State University + orcid: "https://orcid.org/0009-0003-2599-6158" +version: "0.5.0" +date-released: "2026-06-30" +license: MIT +url: "https://github.com/ahb-sjsu/batch-probe" +repository-code: "https://github.com/ahb-sjsu/batch-probe" +keywords: + - pytorch + - cupy + - jax + - gpu + - memory + - batch-size + - oom + - cuda + - thermal + - threads