Commit History

Autor SHA1 Mensaxe Data
  Tri Dao 6e8b25e426 Refactor hai 2 meses
  Ying Zhang 1c9717d699 address comments hai 2 meses
  Ying Zhang dff976a84a fixes hai 3 meses
  Ying Zhang 7b4e68e04f hopper local attention hai 3 meses
  Ying Zhang db80387343 Add seqused_q in fwd / bwd and seqused_k in bwd. hai 3 meses
  jayhshah c92ca63268 FA3 FP8 qkv descales + restore max offset for h128 causal + added sync for producer WG (#1173) hai 3 meses
  Tri Dao bafe253042 [FA3] Bwd hai 4 meses
  jayhshah 5018ac6ac5 Fp8 kernel with "in-kernel" transpose of V in producer (#1100) hai 4 meses
  Tri Dao 3aae9c18c1 Revert "Changes For FP8 (#1075)" hai 4 meses
  ganeshcolfax 1899c970c8 Changes For FP8 (#1075) hai 4 meses
  Ying Zhang dfe1a59e4b Add var-seq-len to FA3 fp16 / bf16 fwd (#1072) hai 4 meses
  Cameron Shinn cb516f855b Remove torchlib dependency from cpp files (#1083) hai 4 meses
  Tri Dao 74b0761ff7 [FA3] BF16 forward hai 5 meses
  Tri Dao 7f67966cc7 FA3 initial code release hai 5 meses