Commit History

Autor SHA1 Mensaxe Data
  Tri Dao 7bc3f031a4 Compile for both Sm80 and Sm90 hai 3 semanas
  Tri Dao 7a802796e1 Big refactor and update hai 4 semanas
  Kai Londenberg b443207c1f Paged Attention support for FA3 (#1268) hai 2 meses
  jayhshah a5a75274bc FA3 kvcache + split kv + gqa parallelization (#1236) hai 3 meses
  Ying Zhang 7b4e68e04f hopper local attention hai 5 meses
  Ying Zhang db80387343 Add seqused_q in fwd / bwd and seqused_k in bwd. hai 5 meses
  jayhshah c92ca63268 FA3 FP8 qkv descales + restore max offset for h128 causal + added sync for producer WG (#1173) hai 5 meses
  Tri Dao bafe253042 [FA3] Bwd hai 6 meses
  Ying Zhang dfe1a59e4b Add var-seq-len to FA3 fp16 / bf16 fwd (#1072) hai 6 meses
  Cameron Shinn cb516f855b Remove torchlib dependency from cpp files (#1083) hai 6 meses
  Tri Dao 74b0761ff7 [FA3] BF16 forward hai 6 meses
  Tri Dao 7f67966cc7 FA3 initial code release hai 6 meses