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

ggml : add Q5_0 and Q5_1 quantization (#1187)

Browse source 
* ggml : add Q5_0 quantization (cuBLAS only)

* ggml : fix Q5_0 qh -> uint32_t

* ggml : fix q5_0 histogram stats

* ggml : q5_0 scalar dot product

* ggml : q5_0 ARM NEON dot

* ggml : q5_0 more efficient ARM NEON using uint64_t masks

* ggml : rename Q5_0 -> Q5_1

* ggml : adding Q5_0 mode

* quantize : add Q5_0 and Q5_1 to map

* ggml : AVX2 optimizations for Q5_0, Q5_1 (#1195)

---------

Co-authored-by: Stephan Walter <stephan@walter.name>
This commit is contained in:
Georgi Gerganov 2023-04-26 23:14:13 +03:00 • committed by GitHub
parent 87a6f846d3
commit 574406dc7e
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
8 changed files with 711 additions and 30 deletions
Download patch file Download diff file Expand all files Collapse all files

8
llama.cpp
View file

@ -484,6 +484,8 @@ struct llama_file_loader {
case GGML_TYPE_Q4_1:
case GGML_TYPE_Q4_2:
case GGML_TYPE_Q4_3:
case GGML_TYPE_Q5_0:
case GGML_TYPE_Q5_1:
case GGML_TYPE_Q8_0:
break;
default: {
@ -559,6 +561,8 @@ struct llama_file_saver {
case GGML_TYPE_Q4_1:
case GGML_TYPE_Q4_2:
case GGML_TYPE_Q4_3:
case GGML_TYPE_Q5_0:
case GGML_TYPE_Q5_1:
case GGML_TYPE_Q8_0:
break;
default: LLAMA_ASSERT(false);
@ -850,6 +854,8 @@ static const char *llama_ftype_name(enum llama_ftype ftype) {
return "mostly Q4_1, some F16";
case LLAMA_FTYPE_MOSTLY_Q4_2: return "mostly Q4_2";
case LLAMA_FTYPE_MOSTLY_Q4_3: return "mostly Q4_3";
case LLAMA_FTYPE_MOSTLY_Q5_0: return "mostly Q5_0";
case LLAMA_FTYPE_MOSTLY_Q5_1: return "mostly Q5_1";
case LLAMA_FTYPE_MOSTLY_Q8_0: return "mostly Q8_0";
default: return "unknown, may not work";
}
@ -1588,6 +1594,8 @@ static void llama_model_quantize_internal(const std::string & fname_inp, const s
case LLAMA_FTYPE_MOSTLY_Q4_1: quantized_type = GGML_TYPE_Q4_1; break;
case LLAMA_FTYPE_MOSTLY_Q4_2: quantized_type = GGML_TYPE_Q4_2; break;
case LLAMA_FTYPE_MOSTLY_Q4_3: quantized_type = GGML_TYPE_Q4_3; break;
case LLAMA_FTYPE_MOSTLY_Q5_0: quantized_type = GGML_TYPE_Q5_0; break;
case LLAMA_FTYPE_MOSTLY_Q5_1: quantized_type = GGML_TYPE_Q5_1; break;
case LLAMA_FTYPE_MOSTLY_Q8_0: quantized_type = GGML_TYPE_Q8_0; break;
default: throw format("invalid output file type %d\n", ftype);
};