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

[Common]Fix Test for SDPA and Concat Unified Scales Test #3207

Merged
merged 1 commit into from
Jan 23, 2025
Merged

[Common]Fix Test for SDPA and Concat Unified Scales Test #3207

Show file tree
Hide file tree
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
10 changes: 7 additions & 3 deletions tests/openvino/native/test_unified_scales.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -17,12 +17,16 @@

class TestUnifiedScales(TemplateTestUnifiedScales):
def get_backend_specific_model(self, model: torch.nn.Module) -> ov.Model:
input_shape = model.INPUT_SHAPE
q_input_shape = model.Q_INPUT_SHAPE
kv_input_shape = model.KV_INPUT_SHAPE

backend_model = ov.convert_model(
model,
example_input=(
torch.randn(input_shape),
torch.randn(input_shape),
torch.ones(q_input_shape),
torch.ones(q_input_shape),
torch.ones(kv_input_shape),
torch.ones(kv_input_shape),
),
)

Expand Down
12 changes: 7 additions & 5 deletions tests/torch/fx/test_unified_scales.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,19 +11,21 @@

import torch

from nncf.torch.nncf_network import NNCFNetwork
from tests.cross_fw.test_templates.test_unified_scales import TemplateTestUnifiedScales
from tests.torch.fx.helpers import get_torch_fx_model_q_transformed


class TestUnifiedScales(TemplateTestUnifiedScales):
def get_backend_specific_model(self, model: torch.nn.Module) -> NNCFNetwork:
input_shape = model.INPUT_SHAPE
def get_backend_specific_model(self, model: torch.nn.Module) -> torch.fx.GraphModule:
q_input_shape = model.Q_INPUT_SHAPE
kv_input_shape = model.KV_INPUT_SHAPE
backend_model = get_torch_fx_model_q_transformed(
model,
(
torch.randn(input_shape),
torch.randn(input_shape),
torch.ones(q_input_shape),
torch.ones(q_input_shape),
torch.ones(kv_input_shape),
torch.ones(kv_input_shape),
),
)

Expand Down
9 changes: 6 additions & 3 deletions tests/torch/quantization/test_unified_scales.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -718,12 +718,15 @@ def test_unified_scales_with_shared_nodes():

class TestUnifiedScales(TemplateTestUnifiedScales):
def get_backend_specific_model(self, model: torch.nn.Module) -> NNCFNetwork:
input_shape = model.INPUT_SHAPE
q_input_shape = model.Q_INPUT_SHAPE
kv_input_shape = model.KV_INPUT_SHAPE
backend_model = wrap_model(
model,
(
torch.randn(input_shape),
torch.randn(input_shape),
torch.ones(q_input_shape),
torch.ones(q_input_shape),
torch.ones(kv_input_shape),
torch.ones(kv_input_shape),
),
trace_parameters=True,
)
Expand Down
11 changes: 6 additions & 5 deletions tests/torch/test_models/synthetic.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -665,16 +665,17 @@ def forward(self, x):


class ConcatSDPABlock(torch.nn.Module):
INPUT_SHAPE = (2, 10, 6)
Q_INPUT_SHAPE = [2, 10, 6]
KV_INPUT_SHAPE = [2, 10, 12]

def __init__(self):
super().__init__()

def forward(self, x, y):
def forward(self, x, y, z, w):
concatenated_input = torch.cat((x, y), dim=-1)
query = concatenated_input
key = concatenated_input
value = concatenated_input
attn_output = torch.nn.functional.scaled_dot_product_attention(query, key, value, dropout_p=0.2)
key = z
value = w
attn_output = torch.nn.functional.scaled_dot_product_attention(query, key, value)

return attn_output