ExternalVendorCode/whisper.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/whisper.cpp.git synced 2025-04-26 22:09:44 +00:00
Code Issues Actions 4 Packages Projects Releases Wiki Activity

CUDA: don't convert BF16 weights to FP32 (ggml/1174)

Browse Source 
* add bf16 support

* use convert_from_bf16_cuda instead of convert_unary_cuda for f32

* revert 7ec5085

* move functionality into convert_unary with constexpr
This commit is contained in:
Sigbjørn Skjæret 2025-04-04 21:05:12 +02:00 committed by Georgi Gerganov
parent 8b92060a10
commit 06ce8f83e6
3 changed files with 52 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
ggml/src/ggml-cuda/convert.cu
View File

@ -579,7 +579,13 @@ static __global__ void convert_unary(const void * __restrict__ vx, dst_t * __res
const src_t * x = (const src_t *) vx;
y[i] = x[i];
if constexpr (std::is_same_v<src_t, nv_bfloat16>) {
y[i] = __bfloat162float(x[i]);
} else if constexpr (std::is_same_v<dst_t, nv_bfloat16> && std::is_same_v<src_t, half>) {
y[i] = (float)x[i];
} else {
y[i] = x[i];
}
}
template <typename src_t, typename dst_t>
@ -588,6 +594,17 @@ static void convert_unary_cuda(const void * __restrict__ vx, dst_t * __restrict_
convert_unary<src_t><<<num_blocks, CUDA_DEQUANTIZE_BLOCK_SIZE, 0, stream>>>(vx, y, k);
}
to_bf16_cuda_t ggml_get_to_bf16_cuda(ggml_type type) {
switch (type) {
case GGML_TYPE_F32:
return convert_unary_cuda<float>;
case GGML_TYPE_F16:
return convert_unary_cuda<half>;
default:
return nullptr;
}
}
to_fp16_cuda_t ggml_get_to_fp16_cuda(ggml_type type) {
switch (type) {
case GGML_TYPE_Q4_0:
@ -633,6 +650,8 @@ to_fp16_cuda_t ggml_get_to_fp16_cuda(ggml_type type) {
return dequantize_row_iq3_s_cuda;
case GGML_TYPE_F32:
return convert_unary_cuda<float>;
case GGML_TYPE_BF16:
return convert_unary_cuda<nv_bfloat16>;
default:
return nullptr;
}

3
ggml/src/ggml-cuda/convert.cuh
View File

@ -7,7 +7,10 @@ using to_t_cuda_t = void (*)(const void * __restrict__ x, T * __restrict__ y, in
typedef to_t_cuda_t<float> to_fp32_cuda_t;
typedef to_t_cuda_t<half> to_fp16_cuda_t;
typedef to_t_cuda_t<nv_bfloat16> to_bf16_cuda_t;
to_fp16_cuda_t ggml_get_to_fp16_cuda(ggml_type type);
to_bf16_cuda_t ggml_get_to_bf16_cuda(ggml_type type);
to_fp32_cuda_t ggml_get_to_fp32_cuda(ggml_type type);

30
ggml/src/ggml-cuda/ggml-cuda.cu
View File

@ -1194,7 +1194,35 @@ static void ggml_cuda_op_mul_mat_cublas(
const bool use_fp16 = (src0->type == GGML_TYPE_F16 || ggml_is_quantized(src0->type)) && ggml_is_contiguous(src0) && row_diff == src0->ne[1] && dst->op_params[0] == GGML_PREC_DEFAULT;
if (((GGML_CUDA_CC_IS_NVIDIA(cc) && cc >= GGML_CUDA_CC_VOLTA) || GGML_CUDA_CC_IS_AMD(cc)) && use_fp16) {
if (src0->type == GGML_TYPE_BF16 && ggml_is_contiguous(src0) && row_diff == src0->ne[1]) {
ggml_cuda_pool_alloc<nv_bfloat16> src1_as_bf16(ctx.pool(id));
if (src1->type != GGML_TYPE_BF16) {
const to_bf16_cuda_t to_bf16_cuda = ggml_get_to_bf16_cuda(src1->type);
GGML_ASSERT(to_bf16_cuda != nullptr);
size_t ne = src1_ncols*ne10;
src1_as_bf16.alloc(ne);
to_bf16_cuda(src1_ddf_i, src1_as_bf16.get(), ne, stream);
}
const nv_bfloat16 * src1_ptr = src1->type == GGML_TYPE_BF16 ? (const nv_bfloat16 *) src1_ddf_i : src1_as_bf16.get();
const nv_bfloat16 * src0_ptr = (const nv_bfloat16 *)src0_dd_i;
ggml_cuda_pool_alloc<nv_bfloat16> dst_bf16(ctx.pool(id), row_diff*src1_ncols);
const float alpha_f32 = 1.0f;
const float beta_f32 = 0.0f;
CUBLAS_CHECK(cublasSetStream(ctx.cublas_handle(id), stream));
CUBLAS_CHECK(
cublasGemmEx(ctx.cublas_handle(id), CUBLAS_OP_T, CUBLAS_OP_N,
row_diff, src1_ncols, ne10,
&alpha_f32, src0_ptr, CUDA_R_16BF, ne00,
src1_ptr, CUDA_R_16BF, ne10,
&beta_f32, dst_bf16.get(), CUDA_R_16BF, ldc,
CUBLAS_COMPUTE_32F,
CUBLAS_GEMM_DEFAULT_TENSOR_OP));
const to_fp32_cuda_t to_fp32_cuda = ggml_get_to_fp32_cuda(GGML_TYPE_BF16);
to_fp32_cuda(dst_bf16.get(), dst_dd_i, row_diff*src1_ncols, stream);
} else if (((GGML_CUDA_CC_IS_NVIDIA(cc) && cc >= GGML_CUDA_CC_VOLTA) || GGML_CUDA_CC_IS_AMD(cc)) && use_fp16) {
// convert src0 and src1 to fp16, multiply as fp16, convert dst to fp32
ggml_cuda_pool_alloc<half> src0_as_f16(ctx.pool(id));
if (src0->type != GGML_TYPE_F16) {