Skip to content

fix: simplify qkv_format checks using a membership tuple #3365

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
andrewwhitecdw wants to merge 1 commit into
base: main
Choose a base branch
from
+2 −2
Open

fix: simplify qkv_format checks using a membership tuple #3365

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
4 changes: 2 additions & 2 deletions tests/pytorch/attention/run_attention_with_cp.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -451,7 +451,7 @@ def run_dpa_with_cp(
for x in [q_orig, k_orig, v_orig, dout_orig] + ([] if bias is None else [bias])
]
bias_ = rest[0] if len(rest) else None
if qkv_format == "bshd" or qkv_format == "sbhd":
if qkv_format in ("bshd", "sbhd"):
seq_dim = qkv_format.index("s")
q_, k_, v_, dout_ = [
x.view(
Expand Down Expand Up @@ -584,7 +584,7 @@ def run_dpa_with_cp(
out, dq, dk, dv, dbias, out_, dq_, dk_, dv_, dbias_ = tensors

############ compare results between CP and no-CP ############
if qkv_format == "bshd" or qkv_format == "sbhd":
if qkv_format in ("bshd", "sbhd"):
if is_training:
dq, dk, dv, out = [
x.view(
Expand Down