Skip to content

Enable addmm low-precision cpu #2661

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

Merged
awni merged 1 commit into from
Oct 10, 2025
+9 −4
Merged

Enable addmm low-precision cpu #2661

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
4 changes: 0 additions & 4 deletions mlx/backend/cpu/matmul.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -131,10 +131,6 @@ void Matmul::eval_cpu(const std::vector<array>& inputs, array& out) {
}

void AddMM::eval_cpu(const std::vector<array>& inputs, array& out) {
if (out.dtype() != float32) {
throw std::runtime_error(
"[AddMM::eval_cpu] Currently only supports float32.");
}
if (out.size() == 0) {
out.set_data(allocator::malloc(out.nbytes()));
return;
Expand Down
9 changes: 9 additions & 0 deletions python/tests/test_blas.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -712,6 +712,15 @@ def test_addmm(self):
expected = beta * c + alpha * (a @ b)
self.assertTrue(mx.allclose(expected, out))

# Test half precision
for t, tol in [(mx.float16, 1e-3), (mx.bfloat16, 1e-2)]:
c = mx.ones((32, 32)).astype(t)
a = mx.random.uniform(shape=(32, 32)).astype(t)
b = mx.random.uniform(shape=(32, 32)).astype(t)
out = mx.addmm(c, a, b)
expected = a @ b + c
self.assertTrue(mx.allclose(out, expected, rtol=tol, atol=tol))

def test_addmm_grad(self):
def make_ref_addmm(alpha, beta):
return lambda c, a, b: alpha * (a @ b) + beta * c
Expand Down