| Age | Commit message (Expand) | Author |
| 24 hours | prepare_data: hard-exit after DONE — HF streaming prefetch threads hang int...HEADmain | YurenHao0426 |
| 26 hours | prepare_data: write .partial then rename, so interrupted prep can't leave a h... | YurenHao0426 |
| 30 hours | revert collaborator auto-push flow; results come back manually (scp/drive), n... | YurenHao0426 |
| 30 hours | zero-token collaborator flow: deploy-key quickstart + auto results push (PUSH... | YurenHao0426 |
| 30 hours | HF upload tooling: env-only auth, scoped-token guidance, optional ladder auto... | yurenh |
| 30 hours | results flow: collect.py (node-side bundle) + plot_ladder.py (gap-vs-scale an... | yurenh |
| 30 hours | env bootstrap (optional venv) + self-check in the ladder runner | yurenh |
| 30 hours | ladder runner: SET override passthrough; validated end-to-end | yurenh |
| 30 hours | add zbp_n64 train config; guard missing configs in the ladder runner | yurenh |
| 30 hours | one-click resume-safe ladder runner (env-configurable) | yurenh |
| 30 hours | package: pyproject, Apache-2.0, install docs | yurenh |
| 30 hours | gitignore data/runs; data prep documented as on-node | yurenh |
| 30 hours | data prep (FineWeb-Edu->GPT2 BPE), rho diagnostics, tests, measured-cost notes | yurenh |
| 30 hours | cap score-core probe chunk (memory ~ chunk*B*H*T^2) | yurenh |
| 30 hours | scaffold: model (OLMo2-ish + ZBP partition), trainer (DDP/config), data shard... | yurenh |