ViperEkura a33ca04f60 fix: synchronize final decode async copy ...

- wait for the final split-KV tile before reading shared memory
- cover long decode with production context capacity

2026-08-09 00:31:47 +08:00

..

attn_common.h

refactor: unify paged and contiguous attention kernels via KVSource policy

2026-08-05 14:06:13 +08:00

attn_decode_split_kv_mma.cuh

fix: synchronize final decode async copy

2026-08-09 00:31:47 +08:00

attn_decode_split_kv.cuh

perf: preload V in decode split-kv shared mem and cache partial tensors

2026-08-06 18:27:00 +08:00

attn_decode.cu

fix: use c10::optional for o_part_buf/ml_part_buf decode kernel params

2026-08-06 20:50:48 +08:00

attn_dispatchers.cuh

perf: preload V in decode split-kv shared mem and cache partial tensors

2026-08-06 18:27:00 +08:00

attn_entry_utils.cuh

refactor: unify paged and contiguous attention kernels via KVSource policy

2026-08-05 14:06:13 +08:00

attn_kv_source.cuh

fix: use max_context_len for stable num_splits in paged decode

2026-08-07 14:42:53 +08:00

attn_mma_utils.cuh

build: parametrize CUDA version for wheels and docker

2026-07-31 14:10:55 +08:00

attn_paged_decode.cu

fix: add out_buf to attn_paged_decode for CUDA graph capture compatibility

2026-08-07 19:45:59 +08:00

attn_paged_prefill.cu

refactor: unify paged and contiguous attention kernels via KVSource policy

2026-08-05 14:06:13 +08:00

attn_prefill_split_q_mma.cuh

refactor: unify paged and contiguous attention kernels via KVSource policy

2026-08-05 14:06:13 +08:00

attn_prefill_split_q.cuh

refactor: unify paged and contiguous attention kernels via KVSource policy

2026-08-05 14:06:13 +08:00

attn_prefill.cu

perf: launch CUDA kernels on torch's current stream

2026-08-02 13:20:14 +08:00

attn_warp_utils.cuh

perf: extend MMA decode to arbitrary GQA ratio, add launch bounds, vectorize combine

2026-07-27 00:35:34 +08:00

rotary_emb.cu

perf: launch CUDA kernels on torch's current stream

2026-08-02 13:20:14 +08:00