-
bd349c6193
Add byte-level BPE tokenizer trained from scratch
main
jzzliu
2026-11-07 20:48:09 -07:00
-
046b09c88f
Add speculative decoding with exact-distribution tests; expand README
jzzliu
2026-11-07 20:33:28 -07:00
-
73ff04d77f
Add continuous-batching engine with variable-length KV cache
jzzliu
2026-11-07 20:27:21 -07:00
-
7f194aba23
Add grouped-query attention with parity tests against multi-head
jzzliu
2026-11-07 20:25:52 -07:00
-
3657ef88a8
README: correct model line count
jzzliu
2026-11-07 20:19:18 -07:00
-
6b7927c16c
Add microgpt: from-scratch GPT with RoPE, KV cache, sampling, tests and benchmark
jzzliu
2026-11-07 20:19:10 -07:00
-
5c1eb15b0a
Initial commit
jzzliu
2026-11-08 02:55:44 +00:00