Skip to content
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

[ROCM] adjust test_flash_attn_rocm test tolerance #21379

Merged
merged 1 commit into from
Jul 17, 2024
Merged

[ROCM] adjust test_flash_attn_rocm test tolerance #21379

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
16 changes: 8 additions & 8 deletions onnxruntime/test/python/transformers/test_flash_attn_rocm.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -35,8 +35,8 @@ def test_gqa_no_past_flash_attention(self, _, config, local, rotary, rotary_inte
rotary=rotary,
rotary_interleaved=rotary_interleaved,
packed=packed,
rtol=0.002,
atol=0.002,
rtol=0.001,
atol=0.005,
)
parity_check_gqa_prompt_no_buff(
config,
Expand All @@ -45,8 +45,8 @@ def test_gqa_no_past_flash_attention(self, _, config, local, rotary, rotary_inte
rotary=rotary,
rotary_interleaved=rotary_interleaved,
packed=packed,
rtol=0.002,
atol=0.002,
rtol=0.001,
atol=0.005,
)

@parameterized.expand(gqa_past_flash_attention_test_cases())
Expand All @@ -67,8 +67,8 @@ def test_gqa_past_flash_attention(self, _, config, local, rotary, rotary_interle
rotary=rotary,
rotary_interleaved=rotary_interleaved,
packed=packed,
rtol=0.002,
atol=0.002,
rtol=0.001,
atol=0.005,
)
parity_check_gqa_past_no_buff(
config,
Expand All @@ -77,8 +77,8 @@ def test_gqa_past_flash_attention(self, _, config, local, rotary, rotary_interle
rotary=rotary,
rotary_interleaved=rotary_interleaved,
packed=packed,
rtol=0.002,
atol=0.002,
rtol=0.001,
atol=0.005,
)


Expand Down
Loading