ViuMini-Dense-360M / tokenizer

Commit History

  1. refactor: consolidate pure 32L Dense architecture, purge MoE files, audit hardening

    Verified signature
    3a5d3e5
  2. Audit fix: Update tokenizer/configs/tokenizer_config.json for ViuMini-Dense-360M

    Verified signature
    8ab3701
  3. Audit fix: Update tokenizer/configs/tokenizer_config.yaml for ViuMini-Dense-360M

    Verified signature
    91f527b

  1. fix: DDP auto_tune probe-before-wrap + broadcast micro/accum, R0-only logs, empty-shard guard, safety import path

    Verified signature
    804855f
  2. remove v1 tokenizer (stale, mask typo) — use v2 from dataset repo tokenizer/tokenizer.json

    Verified signature
    c146389
  3. sync local: enforce_mix sampler, tokenizer v2 (mask fix + 35/30/35), mix optional docs, repair scripts

    Verified signature
    991af26

  1. FINAL 48k tokenizer (47756, unk0)

    kaggle
    adde337

  1. docs: rename + push log

    Verified signature
    6266564
  2. init: 28L MoE 241M audit-fixed, smoke pass

    Verified signature
    140f9f6