Skip to content

fix(sparse-attn): correct block_size computation in backward kernel #1

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
Chuge0335 wants to merge 1 commit into
base: main
Choose a base branch
from
Open

fix(sparse-attn): correct block_size computation in backward kernel #1

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 3 additions & 2 deletions csrc/attn/video_sparse_attn/vsa/block_sparse_attn_triton.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -247,11 +247,12 @@ def _attn_bwd_dq(dq, q, K, V, #

kv_blocks = tl.load(q2k_num + meta_base) # int32
kv_ptr = q2k_index + meta_base * max_kv_blks # ptr to list
block_size = tl.load(variable_block_sizes + q_blk)


for blk_idx in range(kv_blocks*2):
block_sparse_offset = (tl.load(kv_ptr + blk_idx//2).to(tl.int32)*2 + blk_idx%2) *step_n * stride_tok
kv_idx = tl.load(kv_ptr + blk_idx//2).to(tl.int32)
block_size = tl.load(variable_block_sizes + kv_idx) - (blk_idx % 2) * step_n
block_sparse_offset = (kv_idx*2 + blk_idx%2) * step_n * stride_tok
kT = tl.load(kT_ptrs + block_sparse_offset)
vT = tl.load(vT_ptrs + block_sparse_offset)
qk = tl.dot(q, kT)
Expand Down