Commit History
Audit fix: Update tokenizer/configs/tokenizer_config.json for ViuMini-Dense-360M 8ab3701 verified
Audit fix: Update tokenizer/configs/tokenizer_config.yaml for ViuMini-Dense-360M 91f527b verified
fix: DDP auto_tune probe-before-wrap + broadcast micro/accum, R0-only logs, empty-shard guard, safety import path 804855f verified
remove v1 tokenizer (stale, mask typo) — use v2 from dataset repo tokenizer/tokenizer.json c146389 verified
sync local: enforce_mix sampler, tokenizer v2 (mask fix + 35/30/35), mix optional docs, repair scripts 991af26 verified
FINAL 48k tokenizer (47756, unk0) adde337
kaggle commited on