Skip to content

cuda : fix 2-bit quants on amd hip #5105

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
merged 2 commits into from
Jan 24, 2024
Merged

cuda : fix 2-bit quants on amd hip #5105

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
5a69780
cuda : fix 2-bit quants on amd hip
Engininja2 Jan 24, 2024
249dfc0
use __low2float intrinsic function for new quants
Engininja2 Jan 24, 2024
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
6 changes: 3 additions & 3 deletions ggml-cuda.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4283,7 +4283,7 @@ static __device__ __forceinline__ float vec_dot_iq2_xxs_q8_1(
q8 += 8;
aux32 >>= 7;
}
const float d = (float)bq2->d * (0.5f + aux32) * (float)bq8_1[ib32].ds.x * 0.25f;
const float d = (float)bq2->d * (0.5f + aux32) * __low2float(bq8_1[ib32].ds) * 0.25f;
return d * sumi;
#else
// iqs is 0...15
Expand All @@ -4294,7 +4294,7 @@ static __device__ __forceinline__ float vec_dot_iq2_xxs_q8_1(
const uint8_t * grid1 = (const uint8_t *)(iq2xxs_grid + aux8[2*il+0]);
const uint8_t * grid2 = (const uint8_t *)(iq2xxs_grid + aux8[2*il+1]);
const uint32_t aux32 = q2[2] | (q2[3] << 16);
const float d = (float)bq2->d * (0.5f + (aux32 >> 28)) * (float)bq8_1[ib32].ds.x * 0.25f;
const float d = (float)bq2->d * (0.5f + (aux32 >> 28)) * __low2float(bq8_1[ib32].ds) * 0.25f;
const uint8_t signs1 = ksigns_iq2xs[(aux32 >> 14*il) & 127];
const uint8_t signs2 = ksigns_iq2xs[(aux32 >> (14*il + 7)) & 127];
const int8_t * q8 = bq8_1[ib32].qs + 16*il;
Expand Down Expand Up @@ -4339,7 +4339,7 @@ static __device__ __forceinline__ float vec_dot_iq2_xs_q8_1(
}
q8 += 8;
}
const float d = (float)bq2->d * (float)bq8_1[ib32].ds.x * 0.25f;
const float d = (float)bq2->d * __low2float(bq8_1[ib32].ds) * 0.25f;
return d * ((0.5f + ls1) * sumi1 + (0.5f + ls2) * sumi2);
#else
assert(false);
Expand Down