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

Function usage

examples/quantize/quantize.cpp:72–88  ·  view source on GitHub ↗

usage: ./quantize [--allow-requantize] [--leave-output-tensor] [--pure] models/llama/ggml-model.gguf [models/llama/ggml-model-quant.gguf] type [nthreads]

Source from the content-addressed store, hash-verified

70// ./quantize [--allow-requantize] [--leave-output-tensor] [--pure] models/llama/ggml-model.gguf [models/llama/ggml-model-quant.gguf] type [nthreads]
71//
72[[noreturn]]
73static void usage(const char * executable) {
74 printf("usage: %s [--help] [--allow-requantize] [--leave-output-tensor] [--pure] model-f32.gguf [model-quant.gguf] type [nthreads]\n\n", executable);
75 printf(" --allow-requantize: Allows requantizing tensors that have already been quantized. Warning: This can severely reduce quality compared to quantizing from 16bit or 32bit\n");
76 printf(" --leave-output-tensor: Will leave output.weight un(re)quantized. Increases model size but may also increase quality, especially when requantizing\n");
77 printf(" --pure: Disable k-quant mixtures and quantize all tensors to the same type\n");
78 printf("\nAllowed quantization types:\n");
79 for (auto & it : QUANT_OPTIONS) {
80 if (it.name != "COPY") {
81 printf(" %2d or ", it.ftype);
82 } else {
83 printf(" ");
84 }
85 printf("%-6s : %s\n", it.name.c_str(), it.desc.c_str());
86 }
87 exit(1);
88}
89
90int main(int argc, char ** argv) {
91 if (argc < 3) {

Callers 1

mainFunction · 0.70

Calls 2

printfFunction · 0.85
c_strMethod · 0.45

Tested by

no test coverage detected