Skip to content

CUDA: faster Mixtral prompt processing for partial offloading #4553

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 1 commit into from
Dec 21, 2023
Merged

CUDA: faster Mixtral prompt processing for partial offloading #4553

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
29 changes: 22 additions & 7 deletions ggml-cuda.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8782,8 +8782,6 @@ static void ggml_cuda_mul_mat_id(const ggml_tensor * src0, const ggml_tensor * s
// TODO: mmq/mmv support
#endif

GGML_ASSERT(dst->backend == GGML_BACKEND_GPU);

const int64_t nb11 = src1->nb[1];
const int64_t nb1 = dst->nb[1];

Expand Down Expand Up @@ -8812,13 +8810,21 @@ static void ggml_cuda_mul_mat_id(const ggml_tensor * src0, const ggml_tensor * s
ggml_tensor src1_row = *src1;
ggml_tensor dst_row = *dst;

src1_row.backend = GGML_BACKEND_GPU;
dst_row.backend = GGML_BACKEND_GPU;

src1_row.extra = &src1_row_extra;
dst_row.extra = &dst_row_extra;

char * src1_original = (char *) src1_extra->data_device[g_main_device];
char * dst_original = (char *) dst_extra->data_device[g_main_device];
char * src1_original = src1->backend == GGML_BACKEND_CPU ?
(char *) src1->data : (char *) src1_extra->data_device[g_main_device];
char * dst_original = dst->backend == GGML_BACKEND_CPU ?
(char *) dst->data : (char *) dst_extra->data_device[g_main_device];

if (src1->ne[1] == 1) {
GGML_ASSERT(src1->backend == GGML_BACKEND_GPU);
GGML_ASSERT(dst->backend == GGML_BACKEND_GPU);

for (int64_t i01 = 0; i01 < ids->ne[1]; i01++) {
//int32_t row_id;
//CUDA_CHECK(cudaMemcpyAsync(&row_id, ids_dev + i01*ids->nb[1] + id*ids->nb[0], sizeof(int32_t), cudaMemcpyDeviceToHost, g_cudaStreams[g_main_device][0]));
Expand Down Expand Up @@ -8846,6 +8852,11 @@ static void ggml_cuda_mul_mat_id(const ggml_tensor * src0, const ggml_tensor * s
src1_row_extra.data_device[g_main_device] = src1_contiguous;
dst_row_extra.data_device[g_main_device] = dst_contiguous;

const cudaMemcpyKind src1_kind = src1->backend == GGML_BACKEND_CPU ?
cudaMemcpyHostToDevice : cudaMemcpyDeviceToDevice;
const cudaMemcpyKind dst_kind = dst->backend == GGML_BACKEND_CPU ?
cudaMemcpyHostToDevice : cudaMemcpyDeviceToDevice;

for (int32_t row_id = 0; row_id < n_as; ++row_id) {
const struct ggml_tensor * src0_row = dst->src[row_id + 2];

Expand All @@ -8860,7 +8871,7 @@ static void ggml_cuda_mul_mat_id(const ggml_tensor * src0, const ggml_tensor * s
GGML_ASSERT(row_id >= 0 && row_id < n_as);

CUDA_CHECK(cudaMemcpyAsync(src1_contiguous + num_src1_rows*nb11, src1_original + i01*nb11,
nb11, cudaMemcpyDeviceToDevice, stream));
nb11, src1_kind, stream));
num_src1_rows++;
}

Expand Down Expand Up @@ -8892,14 +8903,18 @@ static void ggml_cuda_mul_mat_id(const ggml_tensor * src0, const ggml_tensor * s
GGML_ASSERT(row_id >= 0 && row_id < n_as);

CUDA_CHECK(cudaMemcpyAsync(dst_original + i01*nb1, dst_contiguous + num_src1_rows*nb1,
nb1, cudaMemcpyDeviceToDevice, stream));
nb1, dst_kind, stream));
num_src1_rows++;
}
}

ggml_cuda_pool_free(src1_contiguous, as_src1);
ggml_cuda_pool_free(dst_contiguous, as_dst);
}

if (dst->backend == GGML_BACKEND_CPU) {
CUDA_CHECK(cudaStreamSynchronize(stream));
}
}

static void ggml_cuda_scale(const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
Expand Down Expand Up @@ -9298,7 +9313,7 @@ bool ggml_cuda_compute_forward(struct ggml_compute_params * params, struct ggml_
|| (tensor->src[0] != nullptr && (tensor->src[0]->backend == GGML_BACKEND_GPU || tensor->src[0]->backend == GGML_BACKEND_GPU_SPLIT))
|| (tensor->src[1] != nullptr && tensor->src[1]->backend == GGML_BACKEND_GPU);

if (!any_on_device && tensor->op != GGML_OP_MUL_MAT) {
if (!any_on_device && tensor->op != GGML_OP_MUL_MAT && tensor->op != GGML_OP_MUL_MAT_ID) {
return false;
}

Expand Down