vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2025-09-11 01:24:36 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Merge branch 'upstream' into concedo_experimental

Browse source 
# Conflicts:
#	ggml.c
#	scripts/compare-llama-bench.py
#	tests/test-backend-ops.cpp
This commit is contained in:
Concedo 2024-05-08 18:19:28 +08:00
commit 165a56088b
13 changed files with 1291 additions and 87 deletions
Download patch file Download diff file Expand all files Collapse all files

20
llama.cpp
View file

@ -3206,6 +3206,7 @@ struct llama_model_loader {
switch (type_max) {
case GGML_TYPE_F32: ftype = LLAMA_FTYPE_ALL_F32; break;
case GGML_TYPE_F16: ftype = LLAMA_FTYPE_MOSTLY_F16; break;
case GGML_TYPE_BF16: ftype = LLAMA_FTYPE_MOSTLY_BF16; break;
case GGML_TYPE_Q4_0: ftype = LLAMA_FTYPE_MOSTLY_Q4_0; break;
case GGML_TYPE_Q4_1: ftype = LLAMA_FTYPE_MOSTLY_Q4_1; break;
case GGML_TYPE_Q5_0: ftype = LLAMA_FTYPE_MOSTLY_Q5_0; break;
@ -3710,6 +3711,7 @@ static std::string llama_model_ftype_name(llama_ftype ftype) {
switch (ftype) {
case LLAMA_FTYPE_ALL_F32: return "all F32";
case LLAMA_FTYPE_MOSTLY_F16: return "F16";
case LLAMA_FTYPE_MOSTLY_BF16: return "BF16";
case LLAMA_FTYPE_MOSTLY_Q4_0: return "Q4_0";
case LLAMA_FTYPE_MOSTLY_Q4_1: return "Q4_1";
case LLAMA_FTYPE_MOSTLY_Q4_1_SOME_F16:
@ -6199,6 +6201,7 @@ static int llama_model_load(const std::string & fname, llama_model & model, llam
|| !(
model.ftype == LLAMA_FTYPE_ALL_F32 ||
model.ftype == LLAMA_FTYPE_MOSTLY_F16 ||
model.ftype == LLAMA_FTYPE_MOSTLY_BF16 ||
model.ftype == LLAMA_FTYPE_MOSTLY_Q4_0 ||
model.ftype == LLAMA_FTYPE_MOSTLY_Q4_1
)
@ -14473,13 +14476,16 @@ static void llama_tensor_dequantize_internal(
if (qtype.to_float == NULL) {
throw std::runtime_error(format("type %s unsupported for integer quantization: no dequantization available", ggml_type_name(tensor->type)));
}
} else if (tensor->type != GGML_TYPE_F16) {
} else if (tensor->type != GGML_TYPE_F16 &&
tensor->type != GGML_TYPE_BF16) {
throw std::runtime_error(format("cannot dequantize/convert tensor type %s", ggml_type_name(tensor->type)));
}
if (nthread < 2) {
if (tensor->type == GGML_TYPE_F16) {
ggml_fp16_to_fp32_row((ggml_fp16_t *)tensor->data, f32_output, nelements);
} else if (tensor->type == GGML_TYPE_BF16) {
ggml_bf16_to_fp32_row((ggml_bf16_t *)tensor->data, f32_output, nelements);
} else if (ggml_is_quantized(tensor->type)) {
qtype.to_float(tensor->data, f32_output, nelements);
} else {
@ -14488,7 +14494,14 @@ static void llama_tensor_dequantize_internal(
return;
}
size_t block_size = tensor->type == GGML_TYPE_F16 ? 1 : (size_t)ggml_blck_size(tensor->type);
size_t block_size;
if (tensor->type == GGML_TYPE_F16 ||
tensor->type == GGML_TYPE_BF16) {
block_size = 1;
} else {
block_size = (size_t)ggml_blck_size(tensor->type);
}
size_t block_size_bytes = ggml_type_size(tensor->type);
GGML_ASSERT(nelements % block_size == 0);
@ -14507,6 +14520,8 @@ static void llama_tensor_dequantize_internal(
auto compute = [qtype] (ggml_type typ, uint8_t * inbuf, float * outbuf, int nels) {
if (typ == GGML_TYPE_F16) {
ggml_fp16_to_fp32_row((ggml_fp16_t *)inbuf, outbuf, nels);
} else if (typ == GGML_TYPE_BF16) {
ggml_bf16_to_fp32_row((ggml_bf16_t *)inbuf, outbuf, nels);
} else {
qtype.to_float(inbuf, outbuf, nels);
}
@ -14867,6 +14882,7 @@ static void llama_model_quantize_internal(const std::string & fname_inp, const s
case LLAMA_FTYPE_MOSTLY_Q5_1: default_type = GGML_TYPE_Q5_1; break;
case LLAMA_FTYPE_MOSTLY_Q8_0: default_type = GGML_TYPE_Q8_0; break;
case LLAMA_FTYPE_MOSTLY_F16: default_type = GGML_TYPE_F16; break;
case LLAMA_FTYPE_MOSTLY_BF16: default_type = GGML_TYPE_BF16; break;
case LLAMA_FTYPE_ALL_F32: default_type = GGML_TYPE_F32; break;
// K-quants