2023-06-04 22:34:30 +02:00
|
|
|
// Evaluate a statically exported ggml computation graph with Metal
|
|
|
|
//
|
|
|
|
// - First, export a LLaMA graph:
|
|
|
|
//
|
|
|
|
// $ ./bin/main -m ../models/7B/ggml-model-q4_0.bin --export
|
|
|
|
//
|
|
|
|
// - Run this tool to evaluate the exported graph:
|
|
|
|
//
|
|
|
|
// $ ./bin/metal llama.ggml
|
|
|
|
//
|
|
|
|
// The purpose of this tool is mostly for debugging and demonstration purposes.
|
|
|
|
// The main limitation of exporting computation graphs is that their sizes are static which often
|
|
|
|
// can be a problem for real-world applications.
|
|
|
|
//
|
|
|
|
|
|
|
|
#include "ggml.h"
|
|
|
|
#include "ggml-metal.h"
|
|
|
|
|
|
|
|
#include <cstdio>
|
|
|
|
#include <cstring>
|
|
|
|
#include <cstdlib>
|
|
|
|
|
|
|
|
int main(int argc, char ** argv) {
|
|
|
|
ggml_time_init();
|
|
|
|
|
|
|
|
if (argc != 2) {
|
|
|
|
fprintf(stderr, "Usage: %s llama.ggml\n", argv[0]);
|
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
|
|
|
|
const char * fname_cgraph = argv[1];
|
|
|
|
|
|
|
|
// load the compute graph
|
|
|
|
struct ggml_context * ctx_data = NULL;
|
|
|
|
struct ggml_context * ctx_eval = NULL;
|
|
|
|
|
|
|
|
struct ggml_cgraph gf = ggml_graph_import(fname_cgraph, &ctx_data, &ctx_eval);
|
|
|
|
|
|
|
|
// this allocates all Metal resources and memory buffers
|
2023-07-07 18:24:01 +02:00
|
|
|
auto * ctx_metal = ggml_metal_init(1);
|
2023-06-04 22:34:30 +02:00
|
|
|
|
2023-06-18 09:52:10 +02:00
|
|
|
const size_t max_size_data = ggml_get_max_tensor_size(ctx_data);
|
|
|
|
const size_t max_size_eval = ggml_get_max_tensor_size(ctx_eval);
|
|
|
|
ggml_metal_add_buffer(ctx_metal, "data", ggml_get_mem_buffer(ctx_data), ggml_get_mem_size(ctx_data), max_size_data);
|
|
|
|
ggml_metal_add_buffer(ctx_metal, "eval", ggml_get_mem_buffer(ctx_eval), ggml_get_mem_size(ctx_eval), max_size_eval);
|
2023-06-04 22:34:30 +02:00
|
|
|
|
|
|
|
// main
|
|
|
|
{
|
|
|
|
struct ggml_tensor * input = ggml_graph_get_tensor(&gf, "embd");
|
|
|
|
*(int32_t *) input->data = 1; // BOS
|
|
|
|
|
|
|
|
ggml_metal_set_tensor(ctx_metal, input);
|
|
|
|
|
|
|
|
// warmup
|
|
|
|
ggml_metal_graph_compute(ctx_metal, &gf);
|
|
|
|
|
|
|
|
const int n_iter = 16;
|
|
|
|
|
|
|
|
const int64_t t0 = ggml_time_us();
|
|
|
|
|
|
|
|
// the actual inference happens here
|
|
|
|
for (int i = 0; i < n_iter; ++i) {
|
|
|
|
ggml_metal_graph_compute(ctx_metal, &gf);
|
|
|
|
}
|
|
|
|
|
|
|
|
const int64_t t1 = ggml_time_us();
|
|
|
|
|
|
|
|
printf("time: %.2f ms, %.2f ms/tok\n", (t1 - t0) / 1000.0, (t1 - t0) / 1000.0 / n_iter);
|
|
|
|
}
|
|
|
|
|
|
|
|
// debug output
|
|
|
|
{
|
|
|
|
struct ggml_tensor * logits = gf.nodes[gf.n_nodes - 1];
|
|
|
|
ggml_metal_get_tensor(ctx_metal, logits);
|
|
|
|
|
|
|
|
float * ptr = (float *) ggml_get_data(logits);
|
|
|
|
|
|
|
|
printf("logits: ");
|
|
|
|
for (int i = 0; i < 10; i++) {
|
|
|
|
printf("%8.4f ", ptr[i]);
|
|
|
|
}
|
|
|
|
printf("\n");
|
|
|
|
int imax = 0;
|
|
|
|
double sum = 0.0;
|
|
|
|
double vmax = -1e9;
|
|
|
|
for (int i = 0; i < 32000; i++) {
|
|
|
|
sum += (double) ptr[i];
|
|
|
|
if (ptr[i] > vmax) {
|
|
|
|
vmax = ptr[i];
|
|
|
|
imax = i;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
printf("sum: %f, imax = %d, vmax = %f\n", sum, imax, vmax);
|
|
|
|
}
|
|
|
|
|
|
|
|
ggml_metal_free(ctx_metal);
|
|
|
|
|
|
|
|
ggml_free(ctx_data);
|
|
|
|
ggml_free(ctx_eval);
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|