Skip to content

Expand TestLinenConsistency parameterized.product in attention_test #5279

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

Expand TestLinenConsistency parameterized.product in attention_test #5279

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
45 changes: 35 additions & 10 deletions tests/nnx/nn/attention_test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -206,7 +206,7 @@ class DummyModule(nnx.Module):

np.testing.assert_allclose(attn_jax, attn_manual, atol=1e-6)

# TODO: add all possible constructor argument values to parameterized.product

class TestLinenConsistency(parameterized.TestCase):
@parameterized.product(
use_bias=[True, False],
Expand All @@ -215,6 +215,8 @@ class TestLinenConsistency(parameterized.TestCase):
precision=[Precision.DEFAULT, Precision.HIGH, Precision.HIGHEST],
decode=[True, False],
normalize_qk=[True, False],
qkv_features=[None, 8],
out_features=[None, 6],
)
def test_nnx_attention_equivalence(
self,
Expand All @@ -224,16 +226,16 @@ def test_nnx_attention_equivalence(
precision: PrecisionLike,
decode: bool,
normalize_qk: bool,
qkv_features: tp.Optional[int],
out_features: tp.Optional[int],
):
key = jax.random.key(42)
rngs = nnx.Rngs(42)

num_heads = 2
in_features = 3
qkv_features = 6
out_features = 6
in_features = 4

x = jax.numpy.ones((1, in_features))
x = jnp.ones((1, in_features))
model_nnx = nnx.MultiHeadAttention(
num_heads=num_heads,
in_features=in_features,
Expand Down Expand Up @@ -264,12 +266,37 @@ def test_nnx_attention_equivalence(
getattr(model_nnx, qkvo).kernel[...] = variables['params'][qkvo]['kernel']
if use_bias:
getattr(model_nnx, qkvo).bias[...] = variables['params'][qkvo]['bias']
if normalize_qk:
model_nnx.query_ln.scale[...] = variables['params']['query_ln']['scale']
model_nnx.key_ln.scale[...] = variables['params']['key_ln']['scale']

# Guard: verify params were copied correctly
for name in ('query', 'key', 'value', 'out'):
np.testing.assert_array_equal(
variables['params'][name]['kernel'],
getattr(model_nnx, name).kernel[...],
)
if use_bias:
np.testing.assert_array_equal(
variables['params'][name]['bias'],
getattr(model_nnx, name).bias[...],
)
if normalize_qk:
np.testing.assert_array_equal(
variables['params']['query_ln']['scale'],
model_nnx.query_ln.scale[...],
)
np.testing.assert_array_equal(
variables['params']['key_ln']['scale'],
model_nnx.key_ln.scale[...],
)
if decode:
model_nnx.init_cache(x.shape, dtype=dtype)

out_nnx = model_nnx(x)
out, cache = model.apply(variables, x, mutable=['cache'])
np.testing.assert_array_equal(out, out_nnx)
out, _ = model.apply(variables, x, mutable=['cache'])
rtol = 1e-3 if dtype == jnp.float16 or param_dtype == jnp.float16 else 1e-6
np.testing.assert_allclose(out, out_nnx, rtol=rtol)


class TestKVFeatures(parameterized.TestCase):
Expand All @@ -284,7 +311,7 @@ def test_varying_num_features(self):
qkv_features = 6
out_features = 6

x = jax.numpy.ones((1, in_features))
x = jnp.ones((1, in_features))
y = jax.random.normal(key, (1, in_kv_features))
layer = nnx.MultiHeadAttention(
num_heads=num_heads,
Expand Down Expand Up @@ -354,5 +381,3 @@ class DummyModule(nnx.Module):

if __name__ == '__main__':
absltest.main()