CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-10-27 08:21:30 +00:00
Code Issues Packages Projects Releases Wiki Activity

ggml : refactor forward_dup for cpu backend (#16062)

Browse Source 
* ggml : refactor forward_dup for cpu backend

* clean up a bit

* add quant/dequant perf test
This commit is contained in:
Xuan-Son Nguyen
2025-09-19 11:31:56 +07:00
committed by GitHub
parent 69ffd89163
commit 0dd58b6877
3 changed files with 94 additions and 855 deletions
Download Patch File Download Diff File Expand all files Collapse all files

14
ggml/src/ggml-cpu/common.h
View File

@@ -28,6 +28,14 @@ static inline float bf16_to_f32(ggml_bf16_t x) {
return GGML_BF16_TO_FP32(x);
}
static inline float i32_to_f32(int32_t x) {
return x;
}
static inline int32_t f32_to_i32(float x) {
return x;
}
static inline float f32_to_f32(float x) {
return x;
}
@@ -54,6 +62,12 @@ struct type_conversion_table<ggml_bf16_t> {
static constexpr ggml_bf16_t (*from_f32)(float) = f32_to_bf16;
};
template <>
struct type_conversion_table<int32_t> {
static constexpr float (*to_f32)(int32_t) = i32_to_f32;
static constexpr int32_t (*from_f32)(float) = f32_to_i32;
};
static std::pair<int64_t, int64_t> get_thread_range(const struct ggml_compute_params * params, const struct ggml_tensor * src0) {
const int64_t ith = params->ith;
const int64_t nth = params->nth;

927
ggml/src/ggml-cpu/ops.cpp
View File

File diff suppressed because it is too large Load Diff