echodict/llama.cpp
version https://git-lfs.github.com/spec/v1 oid sha256:cfc44b7ba25614df70e6b65e3341cae0310163bd32fd31a6b928a542df433faf size 30786
0479
1// Unit tests for quantization specific functions - quantize, dequantize and dot product2 3#include "ggml.h"4#include "ggml-cpu.h"5 6#undef NDEBUG7#include <assert.h>8#include <math.h>9#include <stdio.h>10#include <string>11#include <vector>12 13#if defined(_MSC_VER)14#pragma warning(disable: 4244 4267) // possible loss of data15#endif16 17constexpr float MAX_QUANTIZATION_REFERENCE_ERROR = 0.0001f;18constexpr float MAX_QUANTIZATION_TOTAL_ERROR = 0.002f;19constexpr float MAX_QUANTIZATION_TOTAL_ERROR_BINARY = 0.025f;20constexpr float MAX_QUANTIZATION_TOTAL_ERROR_TERNARY = 0.01f;21constexpr float MAX_QUANTIZATION_TOTAL_ERROR_2BITS = 0.0075f;22constexpr float MAX_QUANTIZATION_TOTAL_ERROR_3BITS = 0.0040f;23constexpr float MAX_QUANTIZATION_TOTAL_ERROR_3BITS_XXS = 0.0050f;24constexpr float MAX_QUANTIZATION_TOTAL_ERROR_FP4 = 0.0030f;25constexpr float MAX_DOT_PRODUCT_ERROR = 0.02f;26constexpr float MAX_DOT_PRODUCT_ERROR_LOWBIT = 0.04f;27constexpr float MAX_DOT_PRODUCT_ERROR_FP4 = 0.03f;28constexpr float MAX_DOT_PRODUCT_ERROR_BINARY = 0.40f;29constexpr float MAX_DOT_PRODUCT_ERROR_TERNARY = 0.15f;30 31static const char* RESULT_STR[] = {"ok", "FAILED"};32 33 34// Generate synthetic data35static void generate_data(float offset, size_t n, float * dst) {36 for (size_t i = 0; i < n; i++) {37 dst[i] = 0.1 + 2*cosf(i + offset);38 }39}40 41// Calculate RMSE between two float arrays42static float array_rmse(const float * a1, const float * a2, size_t n) {43 double sum = 0;44 for (size_t i = 0; i < n; i++) {45 double diff = a1[i] - a2[i];46 sum += diff * diff;47 }48 return sqrtf(sum) / n;49}50 51// Total quantization error on test data52static float total_quantization_error(const ggml_type_traits * qfns, const ggml_type_traits_cpu * qfns_cpu, size_t test_size, const float * test_data) {53 std::vector<uint8_t> tmp_q(2*test_size);54 std::vector<float> tmp_out(test_size);55 56 qfns_cpu->from_float(test_data, tmp_q.data(), test_size);57 qfns->to_float(tmp_q.data(), tmp_out.data(), test_size);58 return array_rmse(test_data, tmp_out.data(), test_size);59}60 61// Total quantization error on test data62static float reference_quantization_error(const ggml_type_traits * qfns, const ggml_type_traits_cpu * qfns_cpu, size_t test_size, const float * test_data) {63 std::vector<uint8_t> tmp_q(2*test_size);64 std::vector<float> tmp_out(test_size);65 std::vector<float> tmp_out_ref(test_size);66 67 // FIXME: why is done twice?68 qfns_cpu->from_float(test_data, tmp_q.data(), test_size);69 qfns->to_float(tmp_q.data(), tmp_out.data(), test_size);70 71 qfns->from_float_ref(test_data, tmp_q.data(), test_size);72 qfns->to_float(tmp_q.data(), tmp_out_ref.data(), test_size);73 74 return array_rmse(tmp_out.data(), tmp_out_ref.data(), test_size);75}76 77static float dot_product(const float * a1, const float * a2, size_t test_size) {78 double sum = 0;79 for (size_t i = 0; i < test_size; i++) {80 sum += a1[i] * a2[i];81 }82 return sum;83}84 85// Total dot product error86static float dot_product_error(const ggml_type_traits * qfns, const ggml_type_traits_cpu * qfns_cpu, size_t test_size, const float * test_data1, const float * test_data2) {87 GGML_UNUSED(qfns);88 89 std::vector<uint8_t> tmp_q1(2*test_size);90 std::vector<uint8_t> tmp_q2(2*test_size);91 92 const auto * vdot = ggml_get_type_traits_cpu(qfns_cpu->vec_dot_type);93 94 qfns_cpu->from_float(test_data1, tmp_q1.data(), test_size);95 vdot->from_float(test_data2, tmp_q2.data(), test_size);96 97 float result = INFINITY;98 qfns_cpu->vec_dot(test_size, &result, 0, tmp_q1.data(), 0, tmp_q2.data(), 0, 1);99 100 const float dot_ref = dot_product(test_data1, test_data2, test_size);101 102 return fabsf(result - dot_ref) / test_size;103}104 105int main(int argc, char * argv[]) {106 bool verbose = false;107 const size_t test_size = 32 * 128;108 109 std::string arg;110 for (int i = 1; i < argc; i++) {111 arg = argv[i];112 113 if (arg == "-v") {114 verbose = true;115 } else {116 fprintf(stderr, "error: unknown argument: %s\n", arg.c_str());117 return 1;118 }119 }120 121 std::vector<float> test_data(test_size);122 std::vector<float> test_data2(test_size);123 124 generate_data(0.0, test_data.size(), test_data.data());125 generate_data(1.0, test_data2.size(), test_data2.data());126 127 ggml_cpu_init();128 129 int num_failed = 0;130 bool failed = false;131 132 for (int i = 0; i < GGML_TYPE_COUNT; i++) {133 ggml_type type = (ggml_type) i;134 const auto * qfns = ggml_get_type_traits(type);135 const auto * qfns_cpu = ggml_get_type_traits_cpu(type);136 137 // deprecated - skip138 if (qfns->blck_size == 0) {139 continue;140 }141 142 const ggml_type ei = (ggml_type)i;143 144 printf("Testing %s\n", ggml_type_name((ggml_type) i));145 ggml_quantize_init(ei);146 147 if (qfns_cpu->from_float && qfns->to_float) {148 const float total_error = total_quantization_error(qfns, qfns_cpu, test_size, test_data.data());149 const float max_quantization_error =150 type == GGML_TYPE_Q1_0 ? MAX_QUANTIZATION_TOTAL_ERROR_BINARY :151 type == GGML_TYPE_TQ1_0 ? MAX_QUANTIZATION_TOTAL_ERROR_TERNARY :152 type == GGML_TYPE_TQ2_0 ? MAX_QUANTIZATION_TOTAL_ERROR_TERNARY :153 type == GGML_TYPE_Q2_K ? MAX_QUANTIZATION_TOTAL_ERROR_2BITS :154 type == GGML_TYPE_IQ2_S ? MAX_QUANTIZATION_TOTAL_ERROR_2BITS :155 type == GGML_TYPE_Q3_K ? MAX_QUANTIZATION_TOTAL_ERROR_3BITS :156 type == GGML_TYPE_IQ3_S ? MAX_QUANTIZATION_TOTAL_ERROR_3BITS :157 type == GGML_TYPE_IQ3_XXS ? MAX_QUANTIZATION_TOTAL_ERROR_3BITS_XXS :158 type == GGML_TYPE_NVFP4 ? MAX_QUANTIZATION_TOTAL_ERROR_FP4 : MAX_QUANTIZATION_TOTAL_ERROR;159 failed = !(total_error < max_quantization_error);160 num_failed += failed;161 if (failed || verbose) {162 printf("%5s absolute quantization error: %s (%f)\n", ggml_type_name(type), RESULT_STR[failed], total_error);163 }164 165 const float reference_error = reference_quantization_error(qfns, qfns_cpu, test_size, test_data.data());166 failed = !(reference_error < MAX_QUANTIZATION_REFERENCE_ERROR);167 num_failed += failed;168 if (failed || verbose) {169 printf("%5s reference implementation error: %s (%f)\n", ggml_type_name(type), RESULT_STR[failed], reference_error);170 }171 172 const float vec_dot_error = dot_product_error(qfns, qfns_cpu, test_size, test_data.data(), test_data2.data());173 const float max_allowed_error = type == GGML_TYPE_Q2_K || type == GGML_TYPE_IQ2_XS || type == GGML_TYPE_IQ2_XXS ||174 type == GGML_TYPE_IQ3_XXS || type == GGML_TYPE_IQ3_S || type == GGML_TYPE_IQ2_S175 ? MAX_DOT_PRODUCT_ERROR_LOWBIT176 : type == GGML_TYPE_Q1_0177 ? MAX_DOT_PRODUCT_ERROR_BINARY178 : type == GGML_TYPE_TQ1_0 || type == GGML_TYPE_TQ2_0179 ? MAX_DOT_PRODUCT_ERROR_TERNARY180 : type == GGML_TYPE_NVFP4181 ? MAX_DOT_PRODUCT_ERROR_FP4182 : MAX_DOT_PRODUCT_ERROR;183 failed = !(vec_dot_error < max_allowed_error);184 num_failed += failed;185 if (failed || verbose) {186 printf("%5s dot product error: %s (%f)\n", ggml_type_name(type), RESULT_STR[failed], vec_dot_error);187 }188 }189 }190 191 if (num_failed || verbose) {192 printf("%d tests failed\n", num_failed);193 }194 195 return num_failed > 0;196}197 