MCPcopy Create free account
hub / github.com/ROCm/AMDMIGraphX / quant_tracer

Function quant_tracer

src/quantization.cpp:55–61  ·  view source on GitHub ↗

Source from the content-addressed store, hash-verified

53MIGRAPHX_DECLARE_ENV_VAR(MIGRAPHX_TRACE_QUANTIZATION)
54
55static tracer quant_tracer()
56{
57 if(enabled(MIGRAPHX_TRACE_QUANTIZATION{}))
58 return tracer{std::cout};
59
60 return tracer{};
61};
62
63// This function is to convert any instructions specified in the input
64// from double or float to float16 by inserting a convert operator.

Callers 4

quantize_fp16Function · 0.85
quantize_bf16Function · 0.85
quantize_8bitsFunction · 0.85
quantize_int4_weightsFunction · 0.85

Calls 1

enabledFunction · 0.70

Tested by

no test coverage detected