Aug 1

1. 19:52

   tokenizer: regenerate token_bytes.pt with upstream 2ce972a raw-bytes fix (191/32768 ids corrected)

   ChrisMcCormick
   Verified signature

   47b3fec
2. 18:25

   Model card: step 5568 is now a ready-made nanochat checkpoint (+8-rank optimizer shards)

   ChrisMcCormick
   Verified signature

   c92b43b
3. 18:23

   meta_005568.json: add user_config so chat_sft inherits the run's LRs

   ChrisMcCormick
   Verified signature

   3768187
4. 18:23

   meta_001950.json: add user_config so chat_sft inherits the run's LRs

   ChrisMcCormick
   Verified signature

   1bfa5e5
5. 18:23

   Add nanochat-format d24 checkpoint (step 5568) + world_size=8 optimizer shards

   ChrisMcCormick
   Verified signature

   7b1ab30
6. 15:57

   Flag that val bpb is not comparable across the batched/varlen pipelines

   ChrisMcCormick
   Verified signature

   4c66da7
7. 15:24

   Converter now emits nanochat optimizer shards for an SFT warm-start

   ChrisMcCormick
   Verified signature

   412ea12
8. 14:46

   Correct optimizer-state notes: SFT needs only the weights; W_O axis is a no-op at d24

   ChrisMcCormick
   Verified signature

   162df34
9. 14:12

   Add model card, meta.json, logs, tokenizer, training source, and nanochat converter

   ChrisMcCormick
   Verified signature

   11f07f9
10. 04:30

    Upload folder using huggingface_hub

    ChrisMcCormick
    Verified signature

    a669958
11. 04:27

    initial commit

    ChrisMcCormick
    Verified signature

    f911c2f