Skip to content

Fix gated delta kernel precision #1066

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

Draft
kernelpool wants to merge 1 commit into
base: main
Choose a base branch
from
+37 −30
Draft

Fix gated delta kernel precision #1066

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
67 changes: 37 additions & 30 deletions mlx_lm/models/gated_delta.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -59,37 +59,44 @@ def _make_gated_delta_kernel(has_mask=False, vectorized=False):
{g_setup}
auto beta_ = beta + b_idx * T * Hv;

for (int t = 0; t < T; ++t) {{
if ({mask_source}) {{
float kv_mem = 0.0f;
for (int i = 0; i < n_per_t; ++i) {{
auto s_idx = n_per_t * dk_idx + i;
state[i] = state[i] * {g_access};
kv_mem += state[i] * k_[s_idx];
}}
kv_mem = simd_sum(kv_mem);

auto delta = (v_[dv_idx] - kv_mem) * beta_[hv_idx];

float out = 0.0f;
for (int i = 0; i < n_per_t; ++i) {{
auto s_idx = n_per_t * dk_idx + i;
state[i] = state[i] + k_[s_idx] * delta;
out += state[i] * q_[s_idx];
}}
out = simd_sum(out);
if (thread_index_in_simdgroup == 0) {{
y[dv_idx] = static_cast<InT>(out);
}}
}}
// Increment data pointers to next time step
q_ += Hk * Dk;
k_ += Hk * Dk;
v_ += Hv * Dv;
y += Hv * Dv;
{g_advance}
beta_ += Hv;
#define BODY() {{ \
float kv_mem = 0.0f, kv_c = 0.0f; \
for (int i = 0; i < n_per_t; ++i) {{ \
auto s_idx = n_per_t * dk_idx + i; \
state[i] *= {g_access}; \
auto p = state[i] * k_[s_idx]; \
auto a = p - kv_c; \
auto b = kv_mem + a; \
kv_c = (b - kv_mem) - a; \
kv_mem = b; \
}} \
kv_mem = simd_sum(kv_mem); \
auto delta = (v_[dv_idx] - kv_mem) * beta_[hv_idx]; \
float out = 0.0f; \
for (int i = 0; i < n_per_t; ++i) {{ \
auto s_idx = n_per_t * dk_idx + i; \
state[i] += k_[s_idx] * delta; \
out += state[i] * q_[s_idx]; \
}} \
out = simd_sum(out); \
if (thread_index_in_simdgroup == 0) \
y[dv_idx] = static_cast<InT>(out); \
}}
#define ADV() q_ += Hk * Dk; k_ += Hk * Dk; v_ += Hv * Dv; y += Hv * Dv; {g_advance} beta_ += Hv;

int t = 0;
for (; t + 3 < T; t += 4) {{
if ({mask_source}) BODY() ADV()
if ({"mask[b_idx * T + t + 1]" if has_mask else "true"}) BODY() ADV()
if ({"mask[b_idx * T + t + 2]" if has_mask else "true"}) BODY() ADV()
if ({"mask[b_idx * T + t + 3]" if has_mask else "true"}) BODY() ADV()
}}
for (; t < T; ++t) {{
if ({mask_source}) BODY()
ADV()
}}
#undef BODY
#undef ADV
for (int i = 0; i < n_per_t; ++i) {{
auto s_idx = n_per_t * dk_idx + i;
o_state[s_idx] = static_cast<StT>(state[i]);
Expand Down