Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

refactored compute forward to not pass in the src tensors each time #729

Merged
merged 8 commits into from
Feb 21, 2024
Merged

refactored compute forward to not pass in the src tensors each time #729

Changes from 1 commit
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
1690a32
refactored compute forward to not pass in the src tensors each time
Feb 11, 2024
f44f91c
fix merge issues with flags
Feb 17, 2024
9f19beb
missed one place in the last commit to fix the is_param / flags issue
Feb 17, 2024
87303cf
minor spacing fix
Feb 17, 2024
ee13028
Merge remote-tracking branch 'upstream/master' into refactor_compute_…
Feb 17, 2024
8e622db
fixed some variable assignments so all tests locally are passing
Feb 18, 2024
672856f
Merge branch 'master' into refactor_compute_forward
siddharthvader Feb 19, 2024
e5d5e80
new change after merge fix
Feb 20, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Next Next commit
fix merge issues with flags
  • Loading branch information
siddharthvader committed Feb 17, 2024
commit f44f91c7931244775a37916ddb4e41e256a26ff2
29 changes: 20 additions & 9 deletions src/ggml.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2607,7 +2607,7 @@ static struct ggml_tensor * ggml_new_tensor_impl(
/*.nb =*/ { 0, 0, 0, 0 },
/*.op =*/ GGML_OP_NONE,
/*.op_params =*/ { 0 },
/*.is_param =*/ false,
/*.is_param =*/ 0,
/*.grad =*/ NULL,
/*.src =*/ { NULL },
/*.perf_runs =*/ 0,
Expand Down Expand Up @@ -6509,7 +6509,7 @@ struct ggml_tensor * ggml_cross_entropy_loss_back(
void ggml_set_param(
struct ggml_context * ctx,
struct ggml_tensor * tensor) {
tensor->is_param = true;
tensor->flags |= GGML_TENSOR_FLAG_PARAM;

GGML_ASSERT(tensor->grad == NULL);
tensor->grad = ggml_dup_tensor(ctx, tensor);
Expand Down Expand Up @@ -15580,7 +15580,7 @@ static struct ggml_tensor * ggml_recompute_graph_node(
return NULL;
}

if (node->is_param) {
if (node->flags & GGML_TENSOR_FLAG_PARAM) {
return node;
}

Expand Down Expand Up @@ -15614,7 +15614,7 @@ static struct ggml_tensor * ggml_recompute_graph_node(

clone->op = node->op;
clone->grad = node->grad;
clone->is_param = node->is_param;
clone->flags = node->flags;
clone->extra = node->extra;
for (int k = 0; k < GGML_MAX_DIMS; ++k) {
clone->nb[k] = node->nb[k];
Expand Down Expand Up @@ -16646,7 +16646,7 @@ void ggml_build_backward_expand(struct ggml_context * ctx, struct ggml_cgraph *
for (int i = 0; i < gf->n_nodes; i++) {
struct ggml_tensor * node = gf->nodes[i];

if (node->is_param) {
if (node->flags & GGML_TENSOR_FLAG_PARAM) {
GGML_PRINT_DEBUG("%s: found root node %p\n", __func__, (void *) node);
ggml_build_forward_expand(gb, node->grad);
}
Expand Down Expand Up @@ -18131,7 +18131,7 @@ void ggml_graph_print(const struct ggml_cgraph * cgraph) {
GGML_PRINT(" - %3d: [ %5" PRId64 ", %5" PRId64 ", %5" PRId64 "] %16s %s (%3d) cpu = %7.3f / %7.3f ms, wall = %7.3f / %7.3f ms\n",
i,
node->ne[0], node->ne[1], node->ne[2],
ggml_op_name(node->op), node->is_param ? "x" : node->grad ? "g" : " ", node->perf_runs,
ggml_op_name(node->op), node->flags & GGML_TENSOR_FLAG_PARAM ? "x" : node->grad ? "g" : " ", node->perf_runs,
(double) node->perf_cycles / (double) ggml_cycles_per_ms(),
(double) node->perf_cycles / (double) ggml_cycles_per_ms() / (double) node->perf_runs,
(double) node->perf_time_us / 1000.0,
Expand Down Expand Up @@ -18224,7 +18224,7 @@ void ggml_graph_dump_dot(const struct ggml_cgraph * gb, const struct ggml_cgraph
continue;
}

if (node->is_param) {
if (node->flags & GGML_TENSOR_FLAG_PARAM) {
snprintf(color, sizeof(color), "yellow");
} else if (node->grad) {
if (ggml_graph_find(gf, node)) {
Expand Down Expand Up @@ -18398,7 +18398,7 @@ static enum ggml_opt_result ggml_opt_adam(
int np = 0;
int64_t nx = 0;
for (int i = 0; i < gf->n_nodes; ++i) {
if (gf->nodes[i]->is_param) {
if (gf->nodes[i]->flags & GGML_TENSOR_FLAG_PARAM) {
GGML_PRINT_DEBUG("found param %d: grad->op = %d\n", np, gf->nodes[i]->grad->op);

GGML_ASSERT(np < GGML_MAX_PARAMS);
Expand Down Expand Up @@ -18761,7 +18761,7 @@ static enum ggml_opt_result ggml_opt_lbfgs(
int np = 0;
int nx = 0;
for (int i = 0; i < gf->n_nodes; ++i) {
if (gf->nodes[i]->is_param) {
if (gf->nodes[i]->flags & GGML_TENSOR_FLAG_PARAM) {
GGML_PRINT_DEBUG("found param %d: grad->op = %d\n", np, gf->nodes[i]->grad->op);

GGML_ASSERT(np < GGML_MAX_PARAMS);
Expand Down Expand Up @@ -19236,6 +19236,17 @@ enum ggml_opt_result ggml_opt_resume_g(

////////////////////////////////////////////////////////////////////////////////

void ggml_set_input(struct ggml_tensor * tensor) {
tensor->flags |= GGML_TENSOR_FLAG_INPUT;
}

void ggml_set_output(struct ggml_tensor * tensor) {
tensor->flags |= GGML_TENSOR_FLAG_OUTPUT;
}

////////////////////////////////////////////////////////////////////////////////


void ggml_quantize_init(enum ggml_type type) {
ggml_critical_section_start();

Expand Down