Skip to content

feat: add tensor loading time stat #793

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

Merged
merged 4 commits into from
Sep 7, 2025
Merged

feat: add tensor loading time stat #793

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
6a1eab5
feat: add tensor loading time stat
leejet Sep 7, 2025
35ac8ed
fix typo
leejet Sep 7, 2025
5ed25a8
add missing s
leejet Sep 7, 2025
854fa23
remove redundant model loading log
leejet Sep 7, 2025
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
50 changes: 50 additions & 0 deletions model.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1966,6 +1966,16 @@ std::vector<TensorStorage> remove_duplicates(const std::vector<TensorStorage>& v
}

bool ModelLoader::load_tensors(on_new_tensor_cb_t on_new_tensor_cb) {
int64_t process_time_ms = 0;
int64_t read_time_ms = 0;
int64_t memcpy_time_ms = 0;
int64_t copy_to_backend_time_ms = 0;
int64_t convert_time_ms = 0;

int64_t prev_time_ms = 0;
int64_t curr_time_ms = 0;
int64_t start_time = ggml_time_ms();
prev_time_ms = start_time;
std::vector<TensorStorage> processed_tensor_storages;
for (auto& tensor_storage : tensor_storages) {
// LOG_DEBUG("%s", name.c_str());
Expand All @@ -1978,6 +1988,9 @@ bool ModelLoader::load_tensors(on_new_tensor_cb_t on_new_tensor_cb) {
}
std::vector<TensorStorage> dedup = remove_duplicates(processed_tensor_storages);
processed_tensor_storages = dedup;
curr_time_ms = ggml_time_ms();
process_time_ms = curr_time_ms - prev_time_ms;
prev_time_ms = curr_time_ms;

bool success = true;
for (size_t file_index = 0; file_index < file_paths_.size(); file_index++) {
Expand Down Expand Up @@ -2019,15 +2032,27 @@ bool ModelLoader::load_tensors(on_new_tensor_cb_t on_new_tensor_cb) {
size_t entry_size = zip_entry_size(zip);
if (entry_size != n) {
read_buffer.resize(entry_size);
prev_time_ms = ggml_time_ms();
zip_entry_noallocread(zip, (void*)read_buffer.data(), entry_size);
curr_time_ms = ggml_time_ms();
read_time_ms += curr_time_ms - prev_time_ms;
prev_time_ms = curr_time_ms;
memcpy((void*)buf, (void*)(read_buffer.data() + tensor_storage.offset), n);
curr_time_ms = ggml_time_ms();
memcpy_time_ms += curr_time_ms - prev_time_ms;
} else {
prev_time_ms = ggml_time_ms();
zip_entry_noallocread(zip, (void*)buf, n);
curr_time_ms = ggml_time_ms();
read_time_ms += curr_time_ms - prev_time_ms;
}
zip_entry_close(zip);
} else {
prev_time_ms = ggml_time_ms();
file.seekg(tensor_storage.offset);
file.read(buf, n);
curr_time_ms = ggml_time_ms();
read_time_ms += curr_time_ms - prev_time_ms;
if (!file) {
LOG_ERROR("read tensor data failed: '%s'", file_path.c_str());
return false;
Expand Down Expand Up @@ -2072,6 +2097,7 @@ bool ModelLoader::load_tensors(on_new_tensor_cb_t on_new_tensor_cb) {
read_data(tensor_storage, (char*)dst_tensor->data, nbytes_to_read);
}

prev_time_ms = ggml_time_ms();
if (tensor_storage.is_bf16) {
// inplace op
bf16_to_f32_vec((uint16_t*)dst_tensor->data, (float*)dst_tensor->data, tensor_storage.nelements());
Expand All @@ -2086,10 +2112,13 @@ bool ModelLoader::load_tensors(on_new_tensor_cb_t on_new_tensor_cb) {
} else if (tensor_storage.is_i64) {
i64_to_i32_vec((int64_t*)read_buffer.data(), (int32_t*)dst_tensor->data, tensor_storage.nelements());
}
curr_time_ms = ggml_time_ms();
convert_time_ms += curr_time_ms - prev_time_ms;
} else {
read_buffer.resize(std::max(tensor_storage.nbytes(), tensor_storage.nbytes_to_read()));
read_data(tensor_storage, (char*)read_buffer.data(), nbytes_to_read);

prev_time_ms = ggml_time_ms();
if (tensor_storage.is_bf16) {
// inplace op
bf16_to_f32_vec((uint16_t*)read_buffer.data(), (float*)read_buffer.data(), tensor_storage.nelements());
Expand All @@ -2109,11 +2138,14 @@ bool ModelLoader::load_tensors(on_new_tensor_cb_t on_new_tensor_cb) {

convert_tensor((void*)read_buffer.data(), tensor_storage.type, dst_tensor->data,
dst_tensor->type, (int)tensor_storage.nelements() / (int)tensor_storage.ne[0], (int)tensor_storage.ne[0]);
curr_time_ms = ggml_time_ms();
convert_time_ms += curr_time_ms - prev_time_ms;
}
} else {
read_buffer.resize(std::max(tensor_storage.nbytes(), tensor_storage.nbytes_to_read()));
read_data(tensor_storage, (char*)read_buffer.data(), nbytes_to_read);

prev_time_ms = ggml_time_ms();
if (tensor_storage.is_bf16) {
// inplace op
bf16_to_f32_vec((uint16_t*)read_buffer.data(), (float*)read_buffer.data(), tensor_storage.nelements());
Expand All @@ -2133,14 +2165,24 @@ bool ModelLoader::load_tensors(on_new_tensor_cb_t on_new_tensor_cb) {

if (tensor_storage.type == dst_tensor->type) {
// copy to device memory
curr_time_ms = ggml_time_ms();
convert_time_ms += curr_time_ms - prev_time_ms;
prev_time_ms = curr_time_ms;
ggml_backend_tensor_set(dst_tensor, read_buffer.data(), 0, ggml_nbytes(dst_tensor));
curr_time_ms = ggml_time_ms();
copy_to_backend_time_ms += curr_time_ms - prev_time_ms;
} else {
// convert first, then copy to device memory
convert_buffer.resize(ggml_nbytes(dst_tensor));
convert_tensor((void*)read_buffer.data(), tensor_storage.type,
(void*)convert_buffer.data(), dst_tensor->type,
(int)tensor_storage.nelements() / (int)tensor_storage.ne[0], (int)tensor_storage.ne[0]);
curr_time_ms = ggml_time_ms();
convert_time_ms += curr_time_ms - prev_time_ms;
prev_time_ms = curr_time_ms;
ggml_backend_tensor_set(dst_tensor, convert_buffer.data(), 0, ggml_nbytes(dst_tensor));
curr_time_ms = ggml_time_ms();
copy_to_backend_time_ms += curr_time_ms - prev_time_ms;
}
}
++tensor_count;
Expand Down Expand Up @@ -2170,6 +2212,14 @@ bool ModelLoader::load_tensors(on_new_tensor_cb_t on_new_tensor_cb) {
break;
}
}
int64_t end_time = ggml_time_ms();
LOG_INFO("loading tensors completed, taking %.2fs (process: %.2fs, read: %.2fs, memcpy: %.2fs, convert: %.2fs, copy_to_backend: %.2fs)",
(end_time - start_time) / 1000.f,
process_time_ms / 1000.f,
read_time_ms / 1000.f,
memcpy_time_ms / 1000.f,
convert_time_ms / 1000.f,
copy_to_backend_time_ms / 1000.f);
return success;
}

Expand Down
6 changes: 0 additions & 6 deletions stable-diffusion.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -557,8 +557,6 @@ class StableDiffusionGGML {
// load weights
LOG_DEBUG("loading weights");

int64_t t0 = ggml_time_ms();

std::set<std::string> ignore_tensors;
tensors["alphas_cumprod"] = alphas_cumprod_tensor;
if (use_tiny_autoencoder) {
Expand Down Expand Up @@ -656,11 +654,7 @@ class StableDiffusionGGML {
ggml_backend_is_cpu(clip_backend) ? "RAM" : "VRAM");
}

int64_t t1 = ggml_time_ms();
LOG_INFO("loading model from '%s' completed, taking %.2fs", SAFE_STR(sd_ctx_params->model_path), (t1 - t0) * 1.0f / 1000);

// check is_using_v_parameterization_for_sd2

if (sd_version_is_sd2(version)) {
if (is_using_v_parameterization_for_sd2(ctx, sd_version_is_inpaint(version))) {
is_using_v_parameterization = true;
Expand Down
Loading