Skip to content

Fix/turbo4 wht dequant #43

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
TheTom merged 1 commit into from
Apr 2, 2026
Merged

Fix/turbo4 wht dequant #43

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
16 changes: 9 additions & 7 deletions ggml/src/ggml-turbo-quant.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -451,9 +451,10 @@ void quantize_row_turbo4_0_ref(const float * GGML_RESTRICT x, block_turbo4_0 * G
memset(normalized, 0, d * sizeof(float));
}

/* Step 2: Rotate */
/* Step 2: Forward WHT rotation (matches CUDA set_rows) */
float rotated[TURBO_D];
matvec(turbo_rotation, normalized, rotated, d);
memcpy(rotated, normalized, d * sizeof(float));
turbo_cpu_fwht(rotated, d);

#if TURBO4_USE_4BIT
/* Step 3: 4-bit quantization (16 centroids) */
Expand Down Expand Up @@ -554,14 +555,15 @@ void dequantize_row_turbo4_0(const block_turbo4_0 * GGML_RESTRICT x, float * GGM
};
for (int block = 0; block < nb; block++) {
float norm = GGML_FP16_TO_FP32(x[block].norm);
float rotated[QK_TURBO4];
float * dst = y + block * d;
for (int i = 0; i < d; i++) {
uint8_t idx = (x[block].qs[i / 2] >> ((i % 2) * 4)) & 0xF;
rotated[i] = CENTROIDS_4BIT[idx];
dst[i] = CENTROIDS_4BIT[idx] * norm;
}
float * dst = y + block * d;
matvec(turbo_rotation_t, rotated, dst, d);
for (int i = 0; i < d; i++) dst[i] *= norm;
/* No inverse WHT, dequant stays in the rotated domain.
* Q is WHT-rotated by the graph, so <Q_rot, K_rot> gives correct attention scores.
* The inverse WHT is applied to the attention output via GGML_OP_TURBO_WHT (direction=1) in the graph.
*/
}
#else
/* Legacy 3-bit + QJL dequant */
Expand Down