Yesterday
-
feat: replace DGL with shared packed events and native PyTorch graphs
-
feat: add initial root transformer model
-
ci: support macOS core development (#8)
- ci: support macOS core development (3efae5d9927adbcf425c4e070c43e49f3914f1cf)
- test: sort validation artifact imports (09017da62dfb13b4c6cdc34fe4542b857cb89a6c)
- test: isolate legacy DGL shim (0b48d9cb6ff536f40a7e271d3be43d5d7b215039)
- deps: enable DGL for Apple Silicon (64efda9acf92ec91d505718393c2b4ca7b516e2c)
- feat: add portable PyTorch ROOT-GNN backend (63628a2cc4f52362bbac279aa2262a18e62bd488)
- test: exercise trainer optimizer on MPS (d831167ddf326a2f33d1e43750c18a8841ebf13a)
- docs: document Hugging Face data samples (5f3dd145a990babba47d4a7408ea7a62931c3891) -
rewriting codebase (#7)
- Set up migration project structure (082be018b82c5d36082abd11348e15cc0e2c7a80)
- Document root GNN architecture and migration (a76eccda4456b334cc1ec74fa34c06b59a6b3851)
- chore: set up development environment and tooling (05bbd9c0946750c685169e934fd9d344e2561d85)
- chore: target Python 3.12 (3c84acce43ec0cd168654b6a82103a76eecae313)
- feat: implement shared collider feature construction (a61ae40286983e03233e6ae632886d40a87f4910)
- feat: implement ROOT-GNN graph construction (df4e400136516462fa824f06103788969621ead0)
- feat: implement ROOT event data ingestion (565354b95f235844990bdd8f47487f4abfa92603)
- feat: add Hydra configuration and CLI workflows (7997c78b672579ae008f362ea582e464e6f23ae0)
- perf: profile and optimize ROOT-GNN execution (9dcd2b7f88e94da0e57539affd58521e5c6306fd)
- feat: make Hydra configs installable (cbbff50f401079c69990a7590edda79d7aff15e3)
- build: harden packaging and release validation (d5cfbf068eda872a325955f30a5f82d49a4e023b)
- ci: add package and runtime sm…
