Compactbot
/

char-gpt-1.2m

small-language-model

Model card Files Files and versions

17.7 MB

Ctrl+K

Ctrl+K

1 contributor

History: 11 commits

Compactbot's picture

Correct held-out perplexity to full-split test 1.4369 / 4.21 (was 60-batch log numbers)

fcf3599 verified 7 days ago

.gitattributes

1.52 kB
initial commit 7 days ago
LICENSE

87 Bytes
Add license. (#3) 7 days ago
README.md

2.9 kB
Correct held-out perplexity to full-split test 1.4369 / 4.21 (was 60-batch log numbers) 7 days ago
config.json

240 Bytes
Fix config: lm_head is NOT tied (checkpoint stores two distinct 65x128 tensors; model.py never ties). Set tie_word_embeddings=false to match the artifact. 7 days ago
model.py

4.05 kB
Add model code (CharGPT). (#1) 7 days ago
model.safetensors

8.85 MB
xet

Fix corrupted model.safetensors (re-trained clean weights, 53 tensors F32, 1,216,000 params) 7 days ago
model_clean.safetensors

8.85 MB
xet

Diagnostic: upload clean checkpoint to a fresh path to verify landed size 7 days ago
sample.txt

401 Bytes
Add sample generation. (#4) 7 days ago
tokenizer_config.json

713 Bytes
Add char tokenizer config. (#2) 7 days ago