vulkan: fix assertion when qy_needs_dequant (llama/12068)

Looks like a copy/paste bug from qx_needs_dequant.
This commit is contained in:
Jeff Bolz 2025-02-25 09:30:21 -06:00 committed by Georgi Gerganov
parent 394768c48b
commit a0f76b2da7

View File

@ -4183,7 +4183,7 @@ static void ggml_vk_mul_mat_q_f16(ggml_backend_vk_context * ctx, vk_context& sub
} }
if (qy_needs_dequant) { if (qy_needs_dequant) {
d_Y = ctx->prealloc_y; d_Y = ctx->prealloc_y;
GGML_ASSERT(d_Y->size >= y_sz * ne02 * ne03); GGML_ASSERT(d_Y->size >= y_sz * ne12 * ne13);
} else { } else {
d_Y = d_Qy; d_Y = d_Qy;
y_buf_offset = qy_buf_offset; y_buf_offset = qy_buf_offset;
@ -4760,7 +4760,7 @@ static void ggml_vk_mul_mat_id_q_f16(ggml_backend_vk_context * ctx, vk_context&
} }
if (qy_needs_dequant) { if (qy_needs_dequant) {
d_Y = ctx->prealloc_y; d_Y = ctx->prealloc_y;
GGML_ASSERT(d_Y->size >= y_sz * ne02 * ne03); GGML_ASSERT(d_Y->size >= y_sz * ne12 * ne13);
} else { } else {
d_Y = d_Qy; d_Y = d_Qy;
y_buf_offset = qy_buf_offset; y_buf_offset = qy_buf_offset;