root/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2024-12-25 02:44:36 +00:00
Code Issues Actions 9 Packages Projects Releases Wiki Activity

ggml : improve ggml_graph_dump_dot, add ggml_format_name (#1978)

Browse Source 
* Improve ggml_graph_dump_dot, add ggml_format_name

* add more automatic names to view ops

* fix name of copies
This commit is contained in:
slaren 2023-06-24 12:57:18 +02:00 committed by GitHub
parent 11da1a85cd
commit f2c754e1c3
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
2 changed files with 97 additions and 39 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

135
ggml.c
View File

@ -24,6 +24,7 @@
#include <stdio.h> #include <stdio.h>
#include <float.h> #include <float.h>
#include <limits.h> #include <limits.h>
#include <stdarg.h>
#ifdef GGML_USE_METAL #ifdef GGML_USE_METAL
#include <unistd.h> #include <unistd.h>
@ -4734,10 +4735,19 @@ struct ggml_tensor * ggml_set_name(struct ggml_tensor * tensor, const char * nam
return tensor; return tensor;
} }
struct ggml_tensor * ggml_format_name(struct ggml_tensor * tensor, const char * fmt, ...) {
va_list args;
va_start(args, fmt);
vsnprintf(tensor->name, sizeof(tensor->name), fmt, args);
va_end(args);
return tensor;
}
struct ggml_tensor * ggml_view_tensor( struct ggml_tensor * ggml_view_tensor(
struct ggml_context * ctx, struct ggml_context * ctx,
const struct ggml_tensor * src) { const struct ggml_tensor * src) {
struct ggml_tensor * result = ggml_new_tensor_impl(ctx, src->type, src->n_dims, src->ne, src->data); struct ggml_tensor * result = ggml_new_tensor_impl(ctx, src->type, src->n_dims, src->ne, src->data);
ggml_format_name(result, "%s (view)", src->name);
result->nb[0] = src->nb[0]; result->nb[0] = src->nb[0];
result->nb[1] = src->nb[1]; result->nb[1] = src->nb[1];
@ -5899,6 +5909,11 @@ struct ggml_tensor * ggml_cpy_impl(
// make a view of the destination // make a view of the destination
struct ggml_tensor * result = ggml_view_tensor(ctx, b); struct ggml_tensor * result = ggml_view_tensor(ctx, b);
if (strlen(b->name) > 0) {
ggml_format_name(result, "%s (copy of %s)", b->name, a->name);
} else {
ggml_format_name(result, "%s (copy)", a->name);
}
result->op = GGML_OP_CPY; result->op = GGML_OP_CPY;
result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL; result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL;
@ -5935,6 +5950,7 @@ struct ggml_tensor * ggml_cont_impl(
} }
struct ggml_tensor * result = inplace ? ggml_view_tensor(ctx, a) : ggml_dup_tensor(ctx, a); struct ggml_tensor * result = inplace ? ggml_view_tensor(ctx, a) : ggml_dup_tensor(ctx, a);
ggml_format_name(result, "%s (cont)", a->name);
result->op = GGML_OP_CONT; result->op = GGML_OP_CONT;
result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL; result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL;
@ -5978,6 +5994,7 @@ struct ggml_tensor * ggml_reshape(
} }
struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, b->n_dims, b->ne, a->data); struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, b->n_dims, b->ne, a->data);
ggml_format_name(result, "%s (reshaped)", a->name);
result->op = GGML_OP_RESHAPE; result->op = GGML_OP_RESHAPE;
result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL; result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL;
@ -6002,6 +6019,7 @@ struct ggml_tensor * ggml_reshape_1d(
const int64_t ne[1] = { ne0 }; const int64_t ne[1] = { ne0 };
struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 1, ne, a->data); struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 1, ne, a->data);
ggml_format_name(result, "%s (reshaped)", a->name);
result->op = GGML_OP_RESHAPE; result->op = GGML_OP_RESHAPE;
result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL; result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL;
@ -6027,6 +6045,7 @@ struct ggml_tensor * ggml_reshape_2d(
const int64_t ne[2] = { ne0, ne1 }; const int64_t ne[2] = { ne0, ne1 };
struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 2, ne, a->data); struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 2, ne, a->data);
ggml_format_name(result, "%s (reshaped)", a->name);
result->op = GGML_OP_RESHAPE; result->op = GGML_OP_RESHAPE;
result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL; result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL;
@ -6053,6 +6072,7 @@ struct ggml_tensor * ggml_reshape_3d(
const int64_t ne[3] = { ne0, ne1, ne2 }; const int64_t ne[3] = { ne0, ne1, ne2 };
struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 3, ne, a->data); struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 3, ne, a->data);
ggml_format_name(result, "%s (reshaped)", a->name);
result->op = GGML_OP_RESHAPE; result->op = GGML_OP_RESHAPE;
result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL; result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL;
@ -6081,6 +6101,7 @@ struct ggml_tensor * ggml_reshape_4d(
const int64_t ne[4] = { ne0, ne1, ne2, ne3 }; const int64_t ne[4] = { ne0, ne1, ne2, ne3 };
struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 4, ne, a->data); struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 4, ne, a->data);
ggml_format_name(result, "%s (reshaped)", a->name);
result->op = GGML_OP_RESHAPE; result->op = GGML_OP_RESHAPE;
result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL; result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL;
@ -6105,10 +6126,12 @@ struct ggml_tensor * ggml_view_1d(
} }
struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 1, &ne0, (char *) a->data + offset); struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 1, &ne0, (char *) a->data + offset);
ggml_format_name(result, "%s (view)", a->name);
ggml_scratch_save(ctx); ggml_scratch_save(ctx);
struct ggml_tensor * offs = ggml_new_tensor_1d(ctx, GGML_TYPE_I32, 2); struct ggml_tensor * offs = ggml_new_tensor_1d(ctx, GGML_TYPE_I32, 2);
ggml_set_name(offs, "offset");
memcpy(offs->data, &offset, 2*sizeof(int32_t)); memcpy(offs->data, &offset, 2*sizeof(int32_t));
ggml_scratch_load(ctx); ggml_scratch_load(ctx);
@ -6141,10 +6164,12 @@ struct ggml_tensor * ggml_view_2d(
const int64_t ne[GGML_MAX_DIMS] = { ne0, ne1, 1, 1 }; const int64_t ne[GGML_MAX_DIMS] = { ne0, ne1, 1, 1 };
struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 2, ne, (char *) a->data + offset); struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 2, ne, (char *) a->data + offset);
ggml_format_name(result, "%s (view)", a->name);
ggml_scratch_save(ctx); ggml_scratch_save(ctx);
struct ggml_tensor * offs = ggml_new_tensor_1d(ctx, GGML_TYPE_I32, 2); struct ggml_tensor * offs = ggml_new_tensor_1d(ctx, GGML_TYPE_I32, 2);
ggml_set_name(offs, "offset");
memcpy(offs->data, &offset, 2*sizeof(int32_t)); memcpy(offs->data, &offset, 2*sizeof(int32_t));
ggml_scratch_load(ctx); ggml_scratch_load(ctx);
@ -6183,10 +6208,12 @@ struct ggml_tensor * ggml_view_3d(
const int64_t ne[GGML_MAX_DIMS] = { ne0, ne1, ne2, 1 }; const int64_t ne[GGML_MAX_DIMS] = { ne0, ne1, ne2, 1 };
struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 3, ne, (char *) a->data + offset); struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 3, ne, (char *) a->data + offset);
ggml_format_name(result, "%s (view)", a->name);
ggml_scratch_save(ctx); ggml_scratch_save(ctx);
struct ggml_tensor * offs = ggml_new_tensor_1d(ctx, GGML_TYPE_I32, 2); struct ggml_tensor * offs = ggml_new_tensor_1d(ctx, GGML_TYPE_I32, 2);
ggml_set_name(offs, "offset");
memcpy(offs->data, &offset, 2*sizeof(int32_t)); memcpy(offs->data, &offset, 2*sizeof(int32_t));
ggml_scratch_load(ctx); ggml_scratch_load(ctx);
@ -6227,10 +6254,12 @@ struct ggml_tensor * ggml_view_4d(
const int64_t ne[GGML_MAX_DIMS] = { ne0, ne1, ne2, ne3 }; const int64_t ne[GGML_MAX_DIMS] = { ne0, ne1, ne2, ne3 };
struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 4, ne, (char *) a->data + offset); struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 4, ne, (char *) a->data + offset);
ggml_format_name(result, "%s (view)", a->name);
ggml_scratch_save(ctx); ggml_scratch_save(ctx);
struct ggml_tensor * offs = ggml_new_tensor_1d(ctx, GGML_TYPE_I32, 2); struct ggml_tensor * offs = ggml_new_tensor_1d(ctx, GGML_TYPE_I32, 2);
ggml_set_name(offs, "offset");
memcpy(offs->data, &offset, 2*sizeof(int32_t)); memcpy(offs->data, &offset, 2*sizeof(int32_t));
ggml_scratch_load(ctx); ggml_scratch_load(ctx);
@ -6276,6 +6305,7 @@ struct ggml_tensor * ggml_permute(
} }
struct ggml_tensor * result = ggml_view_tensor(ctx, a); struct ggml_tensor * result = ggml_view_tensor(ctx, a);
ggml_format_name(result, "%s (permuted)", a->name);
int ne[GGML_MAX_DIMS]; int ne[GGML_MAX_DIMS];
int nb[GGML_MAX_DIMS]; int nb[GGML_MAX_DIMS];
@ -6335,6 +6365,7 @@ struct ggml_tensor * ggml_transpose(
} }
struct ggml_tensor * result = ggml_view_tensor(ctx, a); struct ggml_tensor * result = ggml_view_tensor(ctx, a);
ggml_format_name(result, "%s (transposed)", a->name);
result->ne[0] = a->ne[1]; result->ne[0] = a->ne[1];
result->ne[1] = a->ne[0]; result->ne[1] = a->ne[0];
@ -16004,7 +16035,7 @@ static void ggml_visit_parents(struct ggml_cgraph * cgraph, struct ggml_tensor *
GGML_ASSERT(cgraph->n_leafs < GGML_MAX_NODES); GGML_ASSERT(cgraph->n_leafs < GGML_MAX_NODES);
if (strlen(node->name) == 0) { if (strlen(node->name) == 0) {
snprintf(node->name, sizeof(node->name), "leaf_%d", cgraph->n_leafs); ggml_format_name(node, "leaf_%d", cgraph->n_leafs);
} }
cgraph->leafs[cgraph->n_leafs] = node; cgraph->leafs[cgraph->n_leafs] = node;
@ -16013,7 +16044,7 @@ static void ggml_visit_parents(struct ggml_cgraph * cgraph, struct ggml_tensor *
GGML_ASSERT(cgraph->n_nodes < GGML_MAX_NODES); GGML_ASSERT(cgraph->n_nodes < GGML_MAX_NODES);
if (strlen(node->name) == 0) { if (strlen(node->name) == 0) {
snprintf(node->name, sizeof(node->name), "node_%d", cgraph->n_nodes); ggml_format_name(node, "node_%d", cgraph->n_nodes);
} }
cgraph->nodes[cgraph->n_nodes] = node; cgraph->nodes[cgraph->n_nodes] = node;
@ -17397,6 +17428,26 @@ static struct ggml_tensor * ggml_graph_get_parent(const struct ggml_cgraph * cgr
return NULL; return NULL;
} }
static void ggml_graph_dump_dot_node_edge(FILE * fp, const struct ggml_cgraph * gb, struct ggml_tensor * node, struct ggml_tensor * parent, const char * label) {
struct ggml_tensor * gparent = ggml_graph_get_parent(gb, node);
struct ggml_tensor * gparent0 = ggml_graph_get_parent(gb, parent);
fprintf(fp, " \"%p\":%s -> \"%p\":%s [ arrowhead = %s; style = %s; label = \"%s\"; ]\n",
gparent0 ? (void *) gparent0 : (void *) parent,
gparent0 ? "g" : "x",
gparent ? (void *) gparent : (void *) node,
gparent ? "g" : "x",
gparent ? "empty" : "vee",
gparent ? "dashed" : "solid",
label);
}
static void ggml_graph_dump_dot_leaf_edge(FILE * fp, struct ggml_tensor * node, struct ggml_tensor * parent, const char * label) {
fprintf(fp, " \"%p\":%s -> \"%p\":%s [ label = \"%s\"; ]\n",
(void *) parent, "x",
(void *) node, "x",
label);
}
void ggml_graph_dump_dot(const struct ggml_cgraph * gb, const struct ggml_cgraph * gf, const char * filename) { void ggml_graph_dump_dot(const struct ggml_cgraph * gb, const struct ggml_cgraph * gf, const char * filename) {
char color[16]; char color[16];
@ -17432,7 +17483,9 @@ void ggml_graph_dump_dot(const struct ggml_cgraph * gb, const struct ggml_cgraph
(void *) node, color); (void *) node, color);
if (strlen(node->name) > 0) { if (strlen(node->name) > 0) {
fprintf(fp, "%s |", node->name); fprintf(fp, "%s (%s)|", node->name, ggml_type_name(node->type));
} else {
fprintf(fp, "(%s)|", ggml_type_name(node->type));
} }
if (node->n_dims == 2) { if (node->n_dims == 2) {
@ -17441,7 +17494,6 @@ void ggml_graph_dump_dot(const struct ggml_cgraph * gb, const struct ggml_cgraph
fprintf(fp, "%d [%" PRId64 ", %" PRId64 ", %" PRId64 "] | <x>%s", i, node->ne[0], node->ne[1], node->ne[2], GGML_OP_SYMBOL[node->op]); fprintf(fp, "%d [%" PRId64 ", %" PRId64 ", %" PRId64 "] | <x>%s", i, node->ne[0], node->ne[1], node->ne[2], GGML_OP_SYMBOL[node->op]);
} }
if (node->grad) { if (node->grad) {
fprintf(fp, " | <g>%s\"; ]\n", GGML_OP_SYMBOL[node->grad->op]); fprintf(fp, " | <g>%s\"; ]\n", GGML_OP_SYMBOL[node->grad->op]);
} else { } else {
@ -17460,18 +17512,29 @@ void ggml_graph_dump_dot(const struct ggml_cgraph * gb, const struct ggml_cgraph
(void *) node, color); (void *) node, color);
if (strlen(node->name) > 0) { if (strlen(node->name) > 0) {
fprintf(fp, "%s | ", node->name); fprintf(fp, "%s (%s)|", node->name, ggml_type_name(node->type));
} else {
fprintf(fp, "(%s)|", ggml_type_name(node->type));
} }
if (ggml_nelements(node) == 1) {
if (node->type == GGML_TYPE_I8 || node->type == GGML_TYPE_I16 || node->type == GGML_TYPE_I32) { fprintf(fp, "CONST %d [%" PRId64 ", %" PRId64 "]", i, node->ne[0], node->ne[1]);
fprintf(fp, "%d", ggml_get_i32_1d(node, 0)); if (ggml_nelements(node) < 5) {
fprintf(fp, " | (");
for (int j = 0; j < ggml_nelements(node); j++) {
if (node->type == GGML_TYPE_I8 || node->type == GGML_TYPE_I16 || node->type == GGML_TYPE_I32) {
fprintf(fp, "%d", ggml_get_i32_1d(node, j));
}
else if (node->type == GGML_TYPE_F32 || node->type == GGML_TYPE_F16) {
fprintf(fp, "%.1e", (double)ggml_get_f32_1d(node, j));
}
else {
fprintf(fp, "#");
}
if (j < ggml_nelements(node) - 1) {
fprintf(fp, ", ");
}
} }
else { fprintf(fp, ")");
fprintf(fp, "%.1e", (double)ggml_get_f32_1d(node, 0));
}
}
else {
fprintf(fp, "CONST %d [%" PRId64 ", %" PRId64 "]", i, node->ne[0], node->ne[1]);
} }
fprintf(fp, "\"; ]\n"); fprintf(fp, "\"; ]\n");
} }
@ -17479,30 +17542,20 @@ void ggml_graph_dump_dot(const struct ggml_cgraph * gb, const struct ggml_cgraph
for (int i = 0; i < gb->n_nodes; i++) { for (int i = 0; i < gb->n_nodes; i++) {
struct ggml_tensor * node = gb->nodes[i]; struct ggml_tensor * node = gb->nodes[i];
struct ggml_tensor * parent = ggml_graph_get_parent(gb, node);
if (node->src0) { if (node->src0) {
struct ggml_tensor * parent0 = ggml_graph_get_parent(gb, node->src0); ggml_graph_dump_dot_node_edge(fp, gb, node, node->src0, "x");
fprintf(fp, " \"%p\":%s -> \"%p\":%s [ arrowhead = %s; style = %s; label = \"x\"; ]\n",
parent0 ? (void *) parent0 : (void *) node->src0,
parent0 ? "g" : "x",
parent ? (void *) parent : (void *) node,
parent ? "g" : "x",
parent ? "empty" : "vee",
parent ? "dashed" : "solid");
} }
if (node->src1) { if (node->src1) {
struct ggml_tensor * parent1 = ggml_graph_get_parent(gb, node->src1); ggml_graph_dump_dot_node_edge(fp, gb, node, node->src1, "y");
}
fprintf(fp, " \"%p\":%s -> \"%p\":%s [ arrowhead = %s; style = %s; label = \"y\"; ]\n", for (int j = 0; j < GGML_MAX_OPT; j++) {
parent1 ? (void *) parent1 : (void *) node->src1, if (node->opt[j]) {
parent1 ? "g" : "x", char label[16];
parent ? (void *) parent : (void *) node, snprintf(label, sizeof(label), "opt %d", j);
parent ? "g" : "x", ggml_graph_dump_dot_node_edge(fp, gb, node, node->opt[j], label);
parent ? "empty" : "vee", }
parent ? "dashed" : "solid");
} }
} }
@ -17510,15 +17563,19 @@ void ggml_graph_dump_dot(const struct ggml_cgraph * gb, const struct ggml_cgraph
struct ggml_tensor * node = gb->leafs[i]; struct ggml_tensor * node = gb->leafs[i];
if (node->src0) { if (node->src0) {
fprintf(fp, " \"%p\":%s -> \"%p\":%s [ label = \"x\"; ]\n", ggml_graph_dump_dot_leaf_edge(fp, node, node->src0, "x");
(void *) node->src0, "x",
(void *) node, "x");
} }
if (node->src1) { if (node->src1) {
fprintf(fp, " \"%p\":%s -> \"%p\":%s [ label = \"y\"; ]\n", ggml_graph_dump_dot_leaf_edge(fp, node, node->src1, "y");
(void *) node->src1, "x", }
(void *) node, "x");
for (int j = 0; j < GGML_MAX_OPT; j++) {
if (node->opt[j]) {
char label[16];
snprintf(label, sizeof(label), "opt %d", j);
ggml_graph_dump_dot_leaf_edge(fp, node, node->opt[j], label);
}
} }
} }

1
ggml.h
View File

@ -563,6 +563,7 @@ extern "C" {
GGML_API const char * ggml_get_name(const struct ggml_tensor * tensor); GGML_API const char * ggml_get_name(const struct ggml_tensor * tensor);
GGML_API struct ggml_tensor * ggml_set_name(struct ggml_tensor * tensor, const char * name); GGML_API struct ggml_tensor * ggml_set_name(struct ggml_tensor * tensor, const char * name);
GGML_API struct ggml_tensor * ggml_format_name(struct ggml_tensor * tensor, const char * fmt, ...);
// //
// operations on tensors with backpropagation // operations on tensors with backpropagation