MCPcopy Create free account
hub / github.com/Tiiny-AI/PowerInfer / llama_model_quantize_default_params

Function llama_model_quantize_default_params

smallthinker/src/llama-quant.cpp:936–953  ·  view source on GitHub ↗

Source from the content-addressed store, hash-verified

934//
935
936llama_model_quantize_params llama_model_quantize_default_params() {
937 llama_model_quantize_params result = {
938 /*.nthread =*/ 0,
939 /*.ftype =*/ LLAMA_FTYPE_MOSTLY_Q5_1,
940 /*.output_tensor_type =*/ GGML_TYPE_COUNT,
941 /*.token_embedding_type =*/ GGML_TYPE_COUNT,
942 /*.allow_requantize =*/ false,
943 /*.quantize_output_tensor =*/ true,
944 /*.only_copy =*/ false,
945 /*.pure =*/ false,
946 /*.keep_split =*/ false,
947 /*.imatrix =*/ nullptr,
948 /*.kv_overrides =*/ nullptr,
949 /*.tensor_type =*/ nullptr,
950 };
951
952 return result;
953}
954
955uint32_t llama_model_quantize(
956 const char * fname_inp,

Callers 1

mainFunction · 0.50

Calls

no outgoing calls

Tested by

no test coverage detected