ikawrakow/ik_llama.cpp
1
0
Fork 0
mirror of https://github.com/ikawrakow/ik_llama.cpp.git synced 2026-05-01 03:41:53 +00:00
Code Issues Packages Projects Releases Wiki Activity

WIP

Browse Source
This commit is contained in:
Iwan Kawrakow
2025-03-11 11:36:11 +02:00
parent a845e2bfd6
commit 99d9036365
1 changed files with 19 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files

26
ggml/src/ggml-cuda/cpy.cu
View File

@@ -527,19 +527,19 @@ void ggml_cuda_cpy(ggml_backend_cuda_context & ctx, const ggml_tensor * src0, gg
if (src1->type == GGML_TYPE_F16) { if (src1->type == GGML_TYPE_F16) {
auto to_fp16 = ggml_get_to_fp16_cuda(src0->type); auto to_fp16 = ggml_get_to_fp16_cuda(src0->type);
if (to_fp16) { if (to_fp16) {
to_fp16(src0->data, (half *)src1->data, ggml_nrows(src0), src0->ne[1], main_stream); to_fp16(src0->data, (half *)src1->data, ggml_nrows(src0), src0->ne[0], main_stream);
} }
} }
else if (src1->type == GGML_TYPE_F32) { else if (src1->type == GGML_TYPE_F32) {
auto to_fp32 = ggml_get_to_fp32_cuda(src0->type); auto to_fp32 = ggml_get_to_fp32_cuda(src0->type);
if (to_fp32) { if (to_fp32) {
to_fp32(src0->data, (float *)src1->data, ggml_nrows(src0), src0->ne[1], main_stream); to_fp32(src0->data, (float *)src1->data, ggml_nrows(src0), src0->ne[0], main_stream);
} }
} }
else if (src1->type == GGML_TYPE_BF16) { else if (src1->type == GGML_TYPE_BF16) {
auto to_bf16 = ggml_get_to_bf16_cuda(src0->type); auto to_bf16 = ggml_get_to_bf16_cuda(src0->type);
if (to_bf16) { if (to_bf16) {
to_bf16(src0->data, (nv_bfloat16 *)src1->data, ggml_nrows(src0), src0->ne[1], main_stream); to_bf16(src0->data, (nv_bfloat16 *)src1->data, ggml_nrows(src0), src0->ne[0], main_stream);
} }
} }
} else { } else {
@@ -579,9 +579,21 @@ void* ggml_cuda_cpy_fn(const ggml_tensor * src0, ggml_tensor * src1) {
return (void*) cpy_f32_f16<cpy_1_f16_f16>; return (void*) cpy_f32_f16<cpy_1_f16_f16>;
} else if (src0->type == GGML_TYPE_F16 && src1->type == GGML_TYPE_F32) { } else if (src0->type == GGML_TYPE_F16 && src1->type == GGML_TYPE_F32) {
return (void*) cpy_f32_f16<cpy_1_f16_f32>; return (void*) cpy_f32_f16<cpy_1_f16_f32>;
} else { } else if (ggml_is_contiguous(src0) && ggml_are_same_shape(src0, src1)) {
fprintf(stderr, "%s: unsupported type combination (%s to %s)\n", __func__, if (src1->type == GGML_TYPE_F16) {
ggml_type_name(src0->type), ggml_type_name(src1->type)); auto to_fp16 = ggml_get_to_fp16_cuda(src0->type);
GGML_ABORT("fatal error"); if (to_fp16) return (void*)to_fp16;
}
else if (src1->type == GGML_TYPE_F32) {
auto to_fp32 = ggml_get_to_fp32_cuda(src0->type);
if (to_fp32) return (void*)to_fp32;
}
else if (src1->type == GGML_TYPE_BF16) {
auto to_bf16 = ggml_get_to_bf16_cuda(src0->type);
if (to_bf16) return (void*)to_bf16;
}
} }
fprintf(stderr, "%s: unsupported type combination (%s to %s)\n", __func__,
ggml_type_name(src0->type), ggml_type_name(src1->type));
GGML_ABORT("fatal error");
} }