ggml: add ggml_pad example

fundamentals/ggml/src/pad.c

@@ -0,0 +1,63 @@
+#include <stdio.h>
+
+#include "ggml.h"
+#include "ggml-cpu.h"
+#include "ggml-alloc.h"
+#include "ggml-backend.h"
+
+int main(int argc, char **argv) {
+  printf("GGML pad example\n");
+
+  struct ggml_init_params params = {
+    .mem_size   = 16*1024*1024,
+    .mem_buffer = NULL,
+  };
+  struct ggml_context* ctx = ggml_init(params);
+
+  struct ggml_tensor* one = ggml_new_tensor_1d(ctx, GGML_TYPE_F32, 3);
+  ggml_set_f32_nd(one, 0, 0, 0, 0, 100);
+  ggml_set_f32_nd(one, 1, 0, 0, 0, 200);
+  ggml_set_f32_nd(one, 2, 0, 0, 0, 300);
+  // So we start out with the following tensor:
+  //    [100, 200, 300]
+  printf("Only has %ld elements\n", ggml_nelements(one));
+  printf("\n");
+
+  struct ggml_tensor* result = ggml_pad(ctx, one, 2, 0, 1, 0);
+  // And padding will create a new tensor:
+  // z_0
+  //    [100, 200, 300, 0, 0]
+  // z_1
+  //    [  0,   0,   0, 0, 0]
+  ggml_set_name(result, "result");
+
+  struct ggml_cgraph* c_graph = ggml_new_graph(ctx);
+  ggml_build_forward_expand(c_graph, result);
+  enum ggml_status st = ggml_graph_compute_with_ctx(ctx, c_graph, 1);
+  if (st != GGML_STATUS_SUCCESS) {
+    printf("could not compute graph\n");
+    return 1;
+  }
+  printf("result: dims: %d\n", ggml_n_dims(result));
+  printf("result: ne[0]: %ld\n", result->ne[0]);
+  printf("result: ne[1]: %ld\n", result->ne[1]);
+  printf("result: ne[2]: %ld\n", result->ne[2]);
+  printf("result: ne[3]: %ld\n", result->ne[3]);
+  printf("result: elements: %ld\n", ggml_nelements(result));
+
+  printf("result: z_0 0: %f\n", ggml_get_f32_nd(result, 0, 0, 0, 0));
+  printf("result: z_0 1: %f\n", ggml_get_f32_nd(result, 1, 0, 0, 0));
+  printf("result: z_0 2: %f\n", ggml_get_f32_nd(result, 2, 0, 0, 0));
+  printf("result: z_0 3: %f\n", ggml_get_f32_nd(result, 3, 0, 0, 0));
+  printf("result: z_0 4: %f\n", ggml_get_f32_nd(result, 4, 0, 0, 0));
+  printf("\n");
+
+  printf("result: z_1 0: %f\n", ggml_get_f32_nd(result, 0, 0, 1, 0));
+  printf("result: z_1 1: %f\n", ggml_get_f32_nd(result, 1, 0, 1, 0));
+  printf("result: z_1 2: %f\n", ggml_get_f32_nd(result, 2, 0, 1, 0));
+  printf("result: z_1 3: %f\n", ggml_get_f32_nd(result, 3, 0, 1, 0));
+  printf("result: z_1 4: %f\n", ggml_get_f32_nd(result, 4, 0, 1, 0));
+
+  ggml_free(ctx);
+  return 0;
+}