ver4a/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

Support pure float16 add/sub/mul/div operations in the CUDA (and CPU) backend (ggml/1121)

Browse source 
* Support float16-to-float16 add/sub/mul/div operations in the CUDA backend

* Add fp16 support for add/sub/mul/div on the CPU backend

* Add test cases for fp16 add/sub/mul/div
This commit is contained in:
cmdr2 2025-02-25 18:06:34 +05:30 • committed by Georgi Gerganov
parent aede2074f6
commit f54a4ba11e
3 changed files with 237 additions and 43 deletions
Download patch file Download diff file Expand all files Collapse all files

6
ggml/src/ggml-cuda/binbcast.cu
View file

@ -294,11 +294,13 @@ static void ggml_cuda_op_bin_bcast(
const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst,
const void * src0_dd, const void * src1_dd, void * dst_dd, cudaStream_t stream) {
GGML_ASSERT(src1->type == GGML_TYPE_F32);
GGML_ASSERT(src1->type == GGML_TYPE_F32 || src1->type == GGML_TYPE_F16);
if (src0->type == GGML_TYPE_F32 && dst->type == GGML_TYPE_F32) {
op()(src0, src1, dst, (const float *)src0_dd, (const float *)src1_dd, (float *)dst_dd, stream);
} else if (src0->type == GGML_TYPE_F16 && dst->type == GGML_TYPE_F16) {
} else if (src0->type == GGML_TYPE_F16 && src1->type == GGML_TYPE_F16 && dst->type == GGML_TYPE_F16) {
op()(src0, src1, dst, (const half *) src0_dd, (const half *)src1_dd, (half *) dst_dd, stream);
} else if (src0->type == GGML_TYPE_F16 && src1->type == GGML_TYPE_F32 && dst->type == GGML_TYPE_F16) {
op()(src0, src1, dst, (const half *) src0_dd, (const float *)src1_dd, (half *) dst_dd, stream);
} else if (src0->type == GGML_TYPE_F16 && dst->type == GGML_TYPE_F32) {
op()(src0, src1, dst, (const half *) src0_dd, (const float *)src1_dd, (float *)dst_dd, stream);