Skip to content

GitLab

"benchmarks/vscode:/vscode.git/clone" did not exist on "66233af7b6e4217653f1a9952180d68376af7d2a"
Unverified Commit dec66d25 authored by Szymon Ożóg's avatar Szymon Ożóg Committed by GitHub
Browse files

[Kernel] GGUF MMVQ kernel for multiple input vectors (#18754) 


Signed-off-by: default avatarSzymonOzog <szymon.ozog@gmail.com>
parent 8d120701
Hide whitespace changes
Inline Side-by-side
Showing with 95 additions and 87 deletions
csrc/quantization/gguf/gguf_kernel.cu
View file @ dec66d25
...@@ -92,111 +92,112 @@ torch::Tensor ggml_mul_mat_vec_a8(torch::Tensor W, // quant weight ...@@ -92,111 +92,112 @@ torch::Tensor ggml_mul_mat_vec_a8(torch::Tensor W, // quant weight
torch::Tensor X, // input torch::Tensor X, // input
int64_t type, int64_t row) { int64_t type, int64_t row) {
int col = X.sizes()[1]; int col = X.sizes()[1];
int vecs = X.sizes()[0];
const int padded = (col + 512 - 1) / 512 * 512; const int padded = (col + 512 - 1) / 512 * 512;
const at::cuda::OptionalCUDAGuard device_guard(device_of(X)); const at::cuda::OptionalCUDAGuard device_guard(device_of(X));
auto options = torch::TensorOptions().dtype(X.dtype()).device(W.device()); auto options = torch::TensorOptions().dtype(X.dtype()).device(W.device());
at::Tensor Y = torch::empty({1, row}, options); at::Tensor Y = torch::empty({vecs, row}, options);
cudaStream_t stream = at::cuda::getCurrentCUDAStream().stream(); cudaStream_t stream = at::cuda::getCurrentCUDAStream().stream();
options = torch::TensorOptions().dtype(torch::kInt32).device(W.device()); options = torch::TensorOptions().dtype(torch::kInt32).device(W.device());
at::Tensor quant_X = torch::empty({1, padded / 32 * 9}, options); at::Tensor quant_X = torch::empty({vecs, padded / 32 * 9}, options);
VLLM_DISPATCH_FLOATING_TYPES(X.scalar_type(), "ggml_mul_mat_vec_a8", [&] { VLLM_DISPATCH_FLOATING_TYPES(X.scalar_type(), "ggml_mul_mat_vec_a8", [&] {
quantize_row_q8_1_cuda<scalar_t>((scalar_t*)X.data_ptr(), quantize_row_q8_1_cuda<scalar_t>(
(void*)quant_X.data_ptr(), col, 1, stream); (scalar_t*)X.data_ptr(), (void*)quant_X.data_ptr(), col, vecs, stream);
switch (type) { switch (type) {
case 2: case 2:
mul_mat_vec_q4_0_q8_1_cuda<scalar_t>( mul_mat_vec_q4_0_q8_1_cuda<scalar_t>(
(void*)W.data_ptr(), (void*)quant_X.data_ptr(), (void*)W.data_ptr(), (void*)quant_X.data_ptr(),
(scalar_t*)Y.data_ptr(), col, row, stream); (scalar_t*)Y.data_ptr(), col, row, vecs, stream);
break; break;
case 3: case 3:
mul_mat_vec_q4_1_q8_1_cuda<scalar_t>( mul_mat_vec_q4_1_q8_1_cuda<scalar_t>(
(void*)W.data_ptr(), (void*)quant_X.data_ptr(), (void*)W.data_ptr(), (void*)quant_X.data_ptr(),
(scalar_t*)Y.data_ptr(), col, row, stream); (scalar_t*)Y.data_ptr(), col, row, vecs, stream);
break; break;
case 6: case 6:
mul_mat_vec_q5_0_q8_1_cuda<scalar_t>( mul_mat_vec_q5_0_q8_1_cuda<scalar_t>(
(void*)W.data_ptr(), (void*)quant_X.data_ptr(), (void*)W.data_ptr(), (void*)quant_X.data_ptr(),
(scalar_t*)Y.data_ptr(), col, row, stream); (scalar_t*)Y.data_ptr(), col, row, vecs, stream);
break; break;
case 7: case 7:
mul_mat_vec_q5_1_q8_1_cuda<scalar_t>( mul_mat_vec_q5_1_q8_1_cuda<scalar_t>(
(void*)W.data_ptr(), (void*)quant_X.data_ptr(), (void*)W.data_ptr(), (void*)quant_X.data_ptr(),
(scalar_t*)Y.data_ptr(), col, row, stream); (scalar_t*)Y.data_ptr(), col, row, vecs, stream);
break; break;
case 8: case 8:
mul_mat_vec_q8_0_q8_1_cuda<scalar_t>( mul_mat_vec_q8_0_q8_1_cuda<scalar_t>(
(void*)W.data_ptr(), (void*)quant_X.data_ptr(), (void*)W.data_ptr(), (void*)quant_X.data_ptr(),
(scalar_t*)Y.data_ptr(), col, row, stream); (scalar_t*)Y.data_ptr(), col, row, vecs, stream);
break; break;
case 10: case 10:
mul_mat_vec_q2_K_q8_1_cuda<scalar_t>( mul_mat_vec_q2_K_q8_1_cuda<scalar_t>(
(void*)W.data_ptr(), (void*)quant_X.data_ptr(), (void*)W.data_ptr(), (void*)quant_X.data_ptr(),
(scalar_t*)Y.data_ptr(), col, row, stream); (scalar_t*)Y.data_ptr(), col, row, vecs, stream);
break; break;
case 11: case 11:
mul_mat_vec_q3_K_q8_1_cuda<scalar_t>( mul_mat_vec_q3_K_q8_1_cuda<scalar_t>(
(void*)W.data_ptr(), (void*)quant_X.data_ptr(), (void*)W.data_ptr(), (void*)quant_X.data_ptr(),
(scalar_t*)Y.data_ptr(), col, row, stream); (scalar_t*)Y.data_ptr(), col, row, vecs, stream);
break; break;
case 12: case 12:
mul_mat_vec_q4_K_q8_1_cuda<scalar_t>( mul_mat_vec_q4_K_q8_1_cuda<scalar_t>(
(void*)W.data_ptr(), (void*)quant_X.data_ptr(), (void*)W.data_ptr(), (void*)quant_X.data_ptr(),
(scalar_t*)Y.data_ptr(), col, row, stream); (scalar_t*)Y.data_ptr(), col, row, vecs, stream);
break; break;
case 13: case 13:
mul_mat_vec_q5_K_q8_1_cuda<scalar_t>( mul_mat_vec_q5_K_q8_1_cuda<scalar_t>(
(void*)W.data_ptr(), (void*)quant_X.data_ptr(), (void*)W.data_ptr(), (void*)quant_X.data_ptr(),
(scalar_t*)Y.data_ptr(), col, row, stream); (scalar_t*)Y.data_ptr(), col, row, vecs, stream);
break; break;
case 14: case 14:
mul_mat_vec_q6_K_q8_1_cuda<scalar_t>( mul_mat_vec_q6_K_q8_1_cuda<scalar_t>(
(void*)W.data_ptr(), (void*)quant_X.data_ptr(), (void*)W.data_ptr(), (void*)quant_X.data_ptr(),
(scalar_t*)Y.data_ptr(), col, row, stream); (scalar_t*)Y.data_ptr(), col, row, vecs, stream);
break; break;
case 16: case 16:
mul_mat_vec_iq2_xxs_q8_1_cuda<scalar_t>( mul_mat_vec_iq2_xxs_q8_1_cuda<scalar_t>(
(void*)W.data_ptr(), (void*)quant_X.data_ptr(), (void*)W.data_ptr(), (void*)quant_X.data_ptr(),
(scalar_t*)Y.data_ptr(), col, row, stream); (scalar_t*)Y.data_ptr(), col, row, vecs, stream);
break; break;
case 17: case 17:
mul_mat_vec_iq2_xs_q8_1_cuda<scalar_t>( mul_mat_vec_iq2_xs_q8_1_cuda<scalar_t>(
(void*)W.data_ptr(), (void*)quant_X.data_ptr(), (void*)W.data_ptr(), (void*)quant_X.data_ptr(),
(scalar_t*)Y.data_ptr(), col, row, stream); (scalar_t*)Y.data_ptr(), col, row, vecs, stream);
break; break;
case 18: case 18:
mul_mat_vec_iq3_xxs_q8_1_cuda<scalar_t>( mul_mat_vec_iq3_xxs_q8_1_cuda<scalar_t>(
(void*)W.data_ptr(), (void*)quant_X.data_ptr(), (void*)W.data_ptr(), (void*)quant_X.data_ptr(),
(scalar_t*)Y.data_ptr(), col, row, stream); (scalar_t*)Y.data_ptr(), col, row, vecs, stream);
break; break;
case 19: case 19:
mul_mat_vec_iq1_s_q8_1_cuda<scalar_t>( mul_mat_vec_iq1_s_q8_1_cuda<scalar_t>(
(void*)W.data_ptr(), (void*)quant_X.data_ptr(), (void*)W.data_ptr(), (void*)quant_X.data_ptr(),
(scalar_t*)Y.data_ptr(), col, row, stream); (scalar_t*)Y.data_ptr(), col, row, vecs, stream);
break; break;
case 20: case 20:
mul_mat_vec_iq4_nl_q8_1_cuda<scalar_t>( mul_mat_vec_iq4_nl_q8_1_cuda<scalar_t>(
(void*)W.data_ptr(), (void*)quant_X.data_ptr(), (void*)W.data_ptr(), (void*)quant_X.data_ptr(),
(scalar_t*)Y.data_ptr(), col, row, stream); (scalar_t*)Y.data_ptr(), col, row, vecs, stream);
break; break;
case 21: case 21:
mul_mat_vec_iq3_s_q8_1_cuda<scalar_t>( mul_mat_vec_iq3_s_q8_1_cuda<scalar_t>(
(void*)W.data_ptr(), (void*)quant_X.data_ptr(), (void*)W.data_ptr(), (void*)quant_X.data_ptr(),
(scalar_t*)Y.data_ptr(), col, row, stream); (scalar_t*)Y.data_ptr(), col, row, vecs, stream);
break; break;
case 22: case 22:
mul_mat_vec_iq2_s_q8_1_cuda<scalar_t>( mul_mat_vec_iq2_s_q8_1_cuda<scalar_t>(
(void*)W.data_ptr(), (void*)quant_X.data_ptr(), (void*)W.data_ptr(), (void*)quant_X.data_ptr(),
(scalar_t*)Y.data_ptr(), col, row, stream); (scalar_t*)Y.data_ptr(), col, row, vecs, stream);
break; break;
case 23: case 23:
mul_mat_vec_iq4_xs_q8_1_cuda<scalar_t>( mul_mat_vec_iq4_xs_q8_1_cuda<scalar_t>(
(void*)W.data_ptr(), (void*)quant_X.data_ptr(), (void*)W.data_ptr(), (void*)quant_X.data_ptr(),
(scalar_t*)Y.data_ptr(), col, row, stream); (scalar_t*)Y.data_ptr(), col, row, vecs, stream);
break; break;
case 29: case 29:
mul_mat_vec_iq1_m_q8_1_cuda<scalar_t>( mul_mat_vec_iq1_m_q8_1_cuda<scalar_t>(
(void*)W.data_ptr(), (void*)quant_X.data_ptr(), (void*)W.data_ptr(), (void*)quant_X.data_ptr(),
(scalar_t*)Y.data_ptr(), col, row, stream); (scalar_t*)Y.data_ptr(), col, row, vecs, stream);
break; break;
} }
}); });
......
csrc/quantization/gguf/mmvq.cuh
View file @ dec66d25
// copied and adapted from https://github.com/ggerganov/llama.cpp/blob/b2899/ggml-cuda/mmvq.cu // copied and adapted from https://github.com/ggerganov/llama.cpp/blob/b2899/ggml-cuda/mmvq.cu
template <typename scalar_t, int qk, int qi, typename block_q_t, int vdr, vec_dot_q_cuda_t vec_dot_q_cuda> template <typename scalar_t, int qk, int qi, typename block_q_t, int vdr, vec_dot_q_cuda_t vec_dot_q_cuda>
static __global__ void mul_mat_vec_q(const void * __restrict__ vx, const void * __restrict__ vy, scalar_t * __restrict__ dst, const int ncols, const int nrows) { static __global__ void mul_mat_vec_q(const void * __restrict__ vx, const void * __restrict__ vy, scalar_t * __restrict__ dst, const int ncols, const int nrows, const int nvecs) {
const auto row = blockIdx.x*blockDim.y + threadIdx.y; const auto row = blockIdx.x*blockDim.y + threadIdx.y;
const auto vec = blockIdx.y;
if (row >= nrows) { if (row >= nrows || vec >= nvecs) {
return; return;
} }
const int blocks_per_row = ncols / qk; const int blocks_per_row = ncols / qk;
const int blocks_per_warp = vdr * WARP_SIZE / qi; const int blocks_per_warp = vdr * WARP_SIZE / qi;
const int nrows_y = (ncols + 512 - 1) / 512 * 512;
// partial sum for each thread
// partial sum for each thread
float tmp = 0.0f; float tmp = 0.0f;
const block_q_t * x = (const block_q_t *) vx; const block_q_t * x = (const block_q_t *) vx;
...@@ -19,7 +22,7 @@ static __global__ void mul_mat_vec_q(const void * __restrict__ vx, const void * ...@@ -19,7 +22,7 @@ static __global__ void mul_mat_vec_q(const void * __restrict__ vx, const void *
for (auto i = threadIdx.x / (qi/vdr); i < blocks_per_row; i += blocks_per_warp) { for (auto i = threadIdx.x / (qi/vdr); i < blocks_per_row; i += blocks_per_warp) {
const int ibx = row*blocks_per_row + i; // x block index const int ibx = row*blocks_per_row + i; // x block index
const int iby = i * (qk/QK8_1); // y block index that aligns with ibx const int iby = vec*(nrows_y/QK8_1) + i * (qk/QK8_1); // y block index that aligns with ibx
const int iqs = vdr * (threadIdx.x % (qi/vdr)); // x block quant index when casting the quants to int const int iqs = vdr * (threadIdx.x % (qi/vdr)); // x block quant index when casting the quants to int
...@@ -33,177 +36,177 @@ static __global__ void mul_mat_vec_q(const void * __restrict__ vx, const void * ...@@ -33,177 +36,177 @@ static __global__ void mul_mat_vec_q(const void * __restrict__ vx, const void *
} }
if (threadIdx.x == 0) { if (threadIdx.x == 0) {
dst[row] = tmp; dst[vec*nrows + row] = tmp;
} }
} }
template<typename scalar_t> template<typename scalar_t>
static void mul_mat_vec_q4_0_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, cudaStream_t stream) { static void mul_mat_vec_q4_0_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, const int nvecs, cudaStream_t stream) {
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(block_num_y, 1, 1); const dim3 block_nums(block_num_y, nvecs, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<scalar_t, QK4_0, QI4_0, block_q4_0, VDR_Q4_0_Q8_1_MMVQ, vec_dot_q4_0_q8_1> mul_mat_vec_q<scalar_t, QK4_0, QI4_0, block_q4_0, VDR_Q4_0_Q8_1_MMVQ, vec_dot_q4_0_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows, nvecs);
} }
template<typename scalar_t> template<typename scalar_t>
static void mul_mat_vec_q4_1_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, cudaStream_t stream) { static void mul_mat_vec_q4_1_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, const int nvecs, cudaStream_t stream) {
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(block_num_y, 1, 1); const dim3 block_nums(block_num_y, nvecs, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<scalar_t, QK4_0, QI4_1, block_q4_1, VDR_Q4_1_Q8_1_MMVQ, vec_dot_q4_1_q8_1> mul_mat_vec_q<scalar_t, QK4_0, QI4_1, block_q4_1, VDR_Q4_1_Q8_1_MMVQ, vec_dot_q4_1_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows, nvecs);
} }
template<typename scalar_t> template<typename scalar_t>
static void mul_mat_vec_q5_0_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, cudaStream_t stream) { static void mul_mat_vec_q5_0_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, const int nvecs, cudaStream_t stream) {
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(block_num_y, 1, 1); const dim3 block_nums(block_num_y, nvecs, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<scalar_t, QK5_0, QI5_0, block_q5_0, VDR_Q5_0_Q8_1_MMVQ, vec_dot_q5_0_q8_1> mul_mat_vec_q<scalar_t, QK5_0, QI5_0, block_q5_0, VDR_Q5_0_Q8_1_MMVQ, vec_dot_q5_0_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows, nvecs);
} }
template<typename scalar_t> template<typename scalar_t>
static void mul_mat_vec_q5_1_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, cudaStream_t stream) { static void mul_mat_vec_q5_1_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, const int nvecs, cudaStream_t stream) {
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(block_num_y, 1, 1); const dim3 block_nums(block_num_y, nvecs, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<scalar_t, QK5_1, QI5_1, block_q5_1, VDR_Q5_1_Q8_1_MMVQ, vec_dot_q5_1_q8_1> mul_mat_vec_q<scalar_t, QK5_1, QI5_1, block_q5_1, VDR_Q5_1_Q8_1_MMVQ, vec_dot_q5_1_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows, nvecs);
} }
template<typename scalar_t> template<typename scalar_t>
static void mul_mat_vec_q8_0_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, cudaStream_t stream) { static void mul_mat_vec_q8_0_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, const int nvecs, cudaStream_t stream) {
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(block_num_y, 1, 1); const dim3 block_nums(block_num_y, nvecs, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<scalar_t, QK8_0, QI8_0, block_q8_0, VDR_Q8_0_Q8_1_MMVQ, vec_dot_q8_0_q8_1> mul_mat_vec_q<scalar_t, QK8_0, QI8_0, block_q8_0, VDR_Q8_0_Q8_1_MMVQ, vec_dot_q8_0_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows, nvecs);
} }
template<typename scalar_t> template<typename scalar_t>
static void mul_mat_vec_q2_K_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, cudaStream_t stream) { static void mul_mat_vec_q2_K_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, const int nvecs, cudaStream_t stream) {
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(block_num_y, 1, 1); const dim3 block_nums(block_num_y, nvecs, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<scalar_t, QK_K, QI2_K, block_q2_K, VDR_Q2_K_Q8_1_MMVQ, vec_dot_q2_K_q8_1> mul_mat_vec_q<scalar_t, QK_K, QI2_K, block_q2_K, VDR_Q2_K_Q8_1_MMVQ, vec_dot_q2_K_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows, nvecs);
} }
template<typename scalar_t> template<typename scalar_t>
static void mul_mat_vec_q3_K_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, cudaStream_t stream) { static void mul_mat_vec_q3_K_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, const int nvecs, cudaStream_t stream) {
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(block_num_y, 1, 1); const dim3 block_nums(block_num_y, nvecs, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<scalar_t, QK_K, QI3_K, block_q3_K, VDR_Q3_K_Q8_1_MMVQ, vec_dot_q3_K_q8_1> mul_mat_vec_q<scalar_t, QK_K, QI3_K, block_q3_K, VDR_Q3_K_Q8_1_MMVQ, vec_dot_q3_K_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows, nvecs);
} }
template<typename scalar_t> template<typename scalar_t>
static void mul_mat_vec_q4_K_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, cudaStream_t stream) { static void mul_mat_vec_q4_K_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, const int nvecs, cudaStream_t stream) {
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(block_num_y, 1, 1); const dim3 block_nums(block_num_y, nvecs, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<scalar_t, QK_K, QI4_K, block_q4_K, VDR_Q4_K_Q8_1_MMVQ, vec_dot_q4_K_q8_1> mul_mat_vec_q<scalar_t, QK_K, QI4_K, block_q4_K, VDR_Q4_K_Q8_1_MMVQ, vec_dot_q4_K_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows, nvecs);
} }
template<typename scalar_t> template<typename scalar_t>
static void mul_mat_vec_q5_K_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, cudaStream_t stream) { static void mul_mat_vec_q5_K_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, const int nvecs, cudaStream_t stream) {
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(block_num_y, 1, 1); const dim3 block_nums(block_num_y, nvecs, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<scalar_t, QK_K, QI5_K, block_q5_K, VDR_Q5_K_Q8_1_MMVQ, vec_dot_q5_K_q8_1> mul_mat_vec_q<scalar_t, QK_K, QI5_K, block_q5_K, VDR_Q5_K_Q8_1_MMVQ, vec_dot_q5_K_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows, nvecs);
} }
template<typename scalar_t> template<typename scalar_t>
static void mul_mat_vec_q6_K_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, cudaStream_t stream) { static void mul_mat_vec_q6_K_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, const int nvecs, cudaStream_t stream) {
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(block_num_y, 1, 1); const dim3 block_nums(block_num_y, nvecs, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<scalar_t, QK_K, QI6_K, block_q6_K, VDR_Q6_K_Q8_1_MMVQ, vec_dot_q6_K_q8_1> mul_mat_vec_q<scalar_t, QK_K, QI6_K, block_q6_K, VDR_Q6_K_Q8_1_MMVQ, vec_dot_q6_K_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows, nvecs);
} }
template<typename scalar_t> template<typename scalar_t>
static void mul_mat_vec_iq2_xxs_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, cudaStream_t stream) { static void mul_mat_vec_iq2_xxs_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, const int nvecs, cudaStream_t stream) {
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(block_num_y, 1, 1); const dim3 block_nums(block_num_y, nvecs, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<scalar_t, QK_K, QI2_XXS, block_iq2_xxs, 1, vec_dot_iq2_xxs_q8_1> mul_mat_vec_q<scalar_t, QK_K, QI2_XXS, block_iq2_xxs, 1, vec_dot_iq2_xxs_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows, nvecs);
} }
template<typename scalar_t> template<typename scalar_t>
static void mul_mat_vec_iq2_xs_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, cudaStream_t stream) { static void mul_mat_vec_iq2_xs_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, const int nvecs, cudaStream_t stream) {
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(block_num_y, 1, 1); const dim3 block_nums(block_num_y, nvecs, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<scalar_t, QK_K, QI2_XS, block_iq2_xs, 1, vec_dot_iq2_xs_q8_1> mul_mat_vec_q<scalar_t, QK_K, QI2_XS, block_iq2_xs, 1, vec_dot_iq2_xs_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows, nvecs);
} }
template<typename scalar_t> template<typename scalar_t>
static void mul_mat_vec_iq2_s_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, cudaStream_t stream) { static void mul_mat_vec_iq2_s_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, const int nvecs, cudaStream_t stream) {
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(block_num_y, 1, 1); const dim3 block_nums(block_num_y, nvecs, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<scalar_t, QK_K, QI2_S, block_iq2_s, 1, vec_dot_iq2_s_q8_1> mul_mat_vec_q<scalar_t, QK_K, QI2_S, block_iq2_s, 1, vec_dot_iq2_s_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows, nvecs);
} }
template<typename scalar_t> template<typename scalar_t>
static void mul_mat_vec_iq3_xxs_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, cudaStream_t stream) { static void mul_mat_vec_iq3_xxs_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, const int nvecs, cudaStream_t stream) {
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(block_num_y, 1, 1); const dim3 block_nums(block_num_y, nvecs, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<scalar_t, QK_K, QI3_XXS, block_iq3_xxs, 1, vec_dot_iq3_xxs_q8_1> mul_mat_vec_q<scalar_t, QK_K, QI3_XXS, block_iq3_xxs, 1, vec_dot_iq3_xxs_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows, nvecs);
} }
template<typename scalar_t> template<typename scalar_t>
static void mul_mat_vec_iq1_s_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, cudaStream_t stream) { static void mul_mat_vec_iq1_s_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, const int nvecs, cudaStream_t stream) {
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(block_num_y, 1, 1); const dim3 block_nums(block_num_y, nvecs, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<scalar_t, QK_K, QI1_S, block_iq1_s, 1, vec_dot_iq1_s_q8_1> mul_mat_vec_q<scalar_t, QK_K, QI1_S, block_iq1_s, 1, vec_dot_iq1_s_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows, nvecs);
} }
template<typename scalar_t> template<typename scalar_t>
static void mul_mat_vec_iq1_m_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, cudaStream_t stream) { static void mul_mat_vec_iq1_m_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, const int nvecs, cudaStream_t stream) {
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(block_num_y, 1, 1); const dim3 block_nums(block_num_y, nvecs, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<scalar_t, QK_K, QI1_M, block_iq1_m, 1, vec_dot_iq1_m_q8_1> mul_mat_vec_q<scalar_t, QK_K, QI1_M, block_iq1_m, 1, vec_dot_iq1_m_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows, nvecs);
} }
template<typename scalar_t> template<typename scalar_t>
static void mul_mat_vec_iq4_nl_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, cudaStream_t stream) { static void mul_mat_vec_iq4_nl_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, const int nvecs, cudaStream_t stream) {
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(block_num_y, 1, 1); const dim3 block_nums(block_num_y, nvecs, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<scalar_t, QK4_NL, QI4_NL, block_iq4_nl, VDR_Q4_0_Q8_1_MMVQ, vec_dot_iq4_nl_q8_1> mul_mat_vec_q<scalar_t, QK4_NL, QI4_NL, block_iq4_nl, VDR_Q4_0_Q8_1_MMVQ, vec_dot_iq4_nl_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows, nvecs);
} }
template<typename scalar_t> template<typename scalar_t>
static void mul_mat_vec_iq4_xs_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, cudaStream_t stream) { static void mul_mat_vec_iq4_xs_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, const int nvecs, cudaStream_t stream) {
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(block_num_y, 1, 1); const dim3 block_nums(block_num_y, nvecs, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<scalar_t, QK_K, QI4_XS, block_iq4_xs, 1, vec_dot_iq4_xs_q8_1> mul_mat_vec_q<scalar_t, QK_K, QI4_XS, block_iq4_xs, 1, vec_dot_iq4_xs_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows, nvecs);
} }
template<typename scalar_t> template<typename scalar_t>
static void mul_mat_vec_iq3_s_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, cudaStream_t stream) { static void mul_mat_vec_iq3_s_q8_1_cuda(const void * vx, const void * vy, scalar_t * dst, const int ncols, const int nrows, const int nvecs, cudaStream_t stream) {
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(block_num_y, 1, 1); const dim3 block_nums(block_num_y, nvecs, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<scalar_t, QK_K, QI3_XS, block_iq3_s, 1, vec_dot_iq3_s_q8_1> mul_mat_vec_q<scalar_t, QK_K, QI3_XS, block_iq3_s, 1, vec_dot_iq3_s_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows, nvecs);
} }
vllm/_custom_ops.py
View file @ dec66d25
...@@ -594,7 +594,7 @@ if hasattr(torch.ops._C, "ggml_dequantize"): ...@@ -594,7 +594,7 @@ if hasattr(torch.ops._C, "ggml_dequantize"):
quant_type: int, quant_type: int,
row: torch.SymInt, row: torch.SymInt,
) -> torch.Tensor: ) -> torch.Tensor:
return torch.empty((1, row), dtype=X.dtype, device=W.device) return torch.empty((X.shape[0], row), dtype=X.dtype, device=W.device)
@register_fake("_C::ggml_mul_mat_a8") @register_fake("_C::ggml_mul_mat_a8")
def _ggml_mul_mat_a8_fake( def _ggml_mul_mat_a8_fake(
......
vllm/model_executor/layers/quantization/gguf.py
View file @ dec66d25
...@@ -99,6 +99,10 @@ MMQ_QUANT_TYPES = STANDARD_QUANT_TYPES | KQUANT_TYPES ...@@ -99,6 +99,10 @@ MMQ_QUANT_TYPES = STANDARD_QUANT_TYPES | KQUANT_TYPES
def _fused_mul_mat_gguf(x: torch.Tensor, qweight: torch.Tensor, def _fused_mul_mat_gguf(x: torch.Tensor, qweight: torch.Tensor,
qweight_type: int) -> torch.Tensor: qweight_type: int) -> torch.Tensor:
if qweight_type in IMATRIX_QUANT_TYPES:
mmvq_safe = 8 if qweight.shape[0] > 5120 else 16
else:
mmvq_safe = 2 if qweight.shape[0] > 5120 else 6
# HACK: when doing chunked prefill we don't generate output tokens # HACK: when doing chunked prefill we don't generate output tokens
# so input to logits generator is empty which causes invalid parameter # so input to logits generator is empty which causes invalid parameter
if x.shape[0] == 0: if x.shape[0] == 0:
...@@ -110,7 +114,7 @@ def _fused_mul_mat_gguf(x: torch.Tensor, qweight: torch.Tensor, ...@@ -110,7 +114,7 @@ def _fused_mul_mat_gguf(x: torch.Tensor, qweight: torch.Tensor,
if qweight_type in UNQUANTIZED_TYPES: if qweight_type in UNQUANTIZED_TYPES:
return x @ qweight.T return x @ qweight.T
# enable MMVQ in contiguous batching with batch_size=1 # enable MMVQ in contiguous batching with batch_size=1
if x.shape[0] == 1 and qweight_type in MMVQ_QUANT_TYPES: if x.shape[0] <= mmvq_safe and qweight_type in MMVQ_QUANT_TYPES:
y = ops.ggml_mul_mat_vec_a8(qweight, x, qweight_type, qweight.shape[0]) y = ops.ggml_mul_mat_vec_a8(qweight, x, qweight_type, qweight.shape[0])
# Use MMQ Kernel if it's available (standard + k-quants) # Use MMQ Kernel if it's available (standard + k-quants)
elif qweight_type in MMQ_QUANT_TYPES: elif qweight_type in MMQ_QUANT_TYPES:
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment