-
Fix train.py argparse: Support --max_steps alias and --resume flag
-
feat: wipe-proof HF persistence (full ckpt + live log + newest-2 cleanup)
-
feat: full-ckpt HF push every 10k steps (resume safety net) + docs
-
feat: hf_round corpus-slice rotation (Option A full-data: file rotate + row skip + per-round seed)
-
fix: exclude non-5col schema paths from HF stream (wikipedia/grammar/gsm8k CastError on shuffle)
-
fix: probe OOM reserve 4GB + expandable_segments + natural-mode stream shuffle (Hindi-only trap)
-
fix: DDP auto_tune probe-before-wrap + broadcast micro/accum, R0-only logs, empty-shard guard, safety import path
-
DDP multi-GPU support (torchrun, stride-shard, rank0 ckpt, ckpt without module prefix)
-
sync local: enforce_mix sampler, tokenizer v2 (mask fix + 35/30/35), mix optional docs, repair scripts
-
docs: add official comprehensive data audit report (145.93GB/105.1B tokens - 100% PASS)
-
feat: SDPA FlashAttention, language tags conditioning, probe safety, slim checkpoints, and streaming clean
-
fix: audit fixes for multi-token alignment, weight decay groups, and MoE monitoring
-
auto-GPU train + progress display (sync from rental)
-
step3 CLOSE: lock 0.05 (yaml+json+code)
-
step3: add F(0.10/50) + G(0.05/100)
-
step3: fair-50 matrix D+E, lock logic
-
step3: toy add Config C (0.05, 50 steps)
-
step3: toy_router_test.py for Kaggle (no local run per policy)
-
fix: smoke fallback for Kaggle clone without data/raw
-
init: 28L MoE 241M audit-fixed, smoke pass