train: preflight -- static-review fixes: absolute-step cursor, hub_listing pagination, cursor required on resume, repo_type everywhere, rank-0 repo create 658b0a8 verified Cion-lab commited on 13 days ago
preflight: P2 fast-fails after the first dead cell (each repeat bills a GPU session) 51eb041 verified Cion-lab commited on 13 days ago
preflight: P2 parses RUN_JSON per cell and requires the frozen 22L/2048 cell to report 7aab8c2 verified Cion-lab commited on 13 days ago
preflight: assert the trapezoid SHAPE (off-by-one at the last step is expected, not a bug) e687c2c verified Cion-lab commited on 13 days ago
train: preflight -- readback counts as verification; GQA kv must divide heads; pargs reports before optional probe 48c0e43 verified Cion-lab commited on 13 days ago
train: preflight -- download_checkpoint prefix/landing-dir fix (found by P1 v2) 9a7dcc5 verified Cion-lab commited on 13 days ago
preflight: stage pargs -- construct TrainingArguments/model/schedule on the Kaggle image, cross-check 22L=106,194,240 and 20L=99,114,048 87df298 verified Cion-lab commited on 13 days ago
train: preflight -- hubckpt.ensure_repo writes LFS attributes before the first checkpoint push (found by P1) 1279da6 verified Cion-lab commited on 13 days ago
train: Gate 3 preflight driver (reader, hub cycle, T1 throughput, cold resume) e7ee32a verified Cion-lab commited on 13 days ago