jzzliu 7f194aba23 Add grouped-query attention with parity tests against multi-head ...

GQA/MQA via GPTConfig.n_kv_head; the cache stores only the KV heads. Attention
also accepts per-row positions and a key padding mask, used by batched decoding.

2026-11-07 20:25:52 -07:00

..

bench_gqa.py

Add grouped-query attention with parity tests against multi-head

2026-11-07 20:25:52 -07:00

benchmark.py

Add microgpt: from-scratch GPT with RoPE, KV cache, sampling, tests and benchmark

2026-11-07 20:19:10 -07:00