Skip to content

feat: use image and command-line dimensions by default on sdapi #1262

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

Draft
wbruna wants to merge 2 commits into
base: master
Choose a base branch
from
+52 −25
Draft

feat: use image and command-line dimensions by default on sdapi #1262

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
3c9a8d9
fix: correct sdapi handling of cfg_scale and steps
wbruna Feb 8, 2026
f8a17ca
feat: use image and command-line dimensions by default on sdapi
wbruna Feb 8, 2026
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
77 changes: 52 additions & 25 deletions examples/server/main.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -786,8 +786,8 @@ int main(int argc, const char** argv) {
std::string negative_prompt = j.value("negative_prompt", "");
int width = j.value("width", 512);
int height = j.value("height", 512);
int steps = j.value("steps", -1);
float cfg_scale = j.value("cfg_scale", 7.f);
int steps = j.value("steps", default_gen_params.sample_params.sample_steps);
float cfg_scale = j.value("cfg_scale", default_gen_params.sample_params.guidance.txt_cfg);
int64_t seed = j.value("seed", -1);
int batch_size = j.value("batch_size", 1);
int clip_skip = j.value("clip_skip", -1);
Expand Down Expand Up @@ -883,16 +883,13 @@ int main(int argc, const char** argv) {

enum scheduler_t scheduler = str_to_scheduler(scheduler_name.c_str());

// avoid excessive resource usage

SDGenerationParams gen_params = default_gen_params;
gen_params.prompt = prompt;
gen_params.negative_prompt = negative_prompt;
gen_params.width = width;
gen_params.height = height;
gen_params.seed = seed;
gen_params.sample_params.sample_steps = steps;
gen_params.batch_count = batch_size;
SDGenerationParams gen_params = default_gen_params;
gen_params.prompt = prompt;
gen_params.negative_prompt = negative_prompt;
gen_params.seed = seed;
gen_params.sample_params.sample_steps = steps;
gen_params.batch_count = batch_size;
gen_params.sample_params.guidance.txt_cfg = cfg_scale;

if (clip_skip > 0) {
gen_params.clip_skip = clip_skip;
Expand All @@ -906,32 +903,60 @@ int main(int argc, const char** argv) {
gen_params.sample_params.scheduler = scheduler;
}

// re-read to avoid applying 512 as default before the provided
// images and/or server command-line
gen_params.width = j.value("width", -1);
gen_params.height = j.value("height", -1);

LOG_DEBUG("%s\n", gen_params.to_string().c_str());

sd_image_t init_image = {(uint32_t)gen_params.width, (uint32_t)gen_params.height, 3, nullptr};
sd_image_t control_image = {(uint32_t)gen_params.width, (uint32_t)gen_params.height, 3, nullptr};
sd_image_t mask_image = {(uint32_t)gen_params.width, (uint32_t)gen_params.height, 1, nullptr};
sd_image_t init_image = {0, 0, 3, nullptr};
sd_image_t control_image = {0, 0, 3, nullptr};
sd_image_t mask_image = {0, 0, 1, nullptr};
std::vector<uint8_t> mask_data;
std::vector<sd_image_t> pmid_images;
std::vector<sd_image_t> ref_images;

auto get_resolved_width = [&gen_params, &default_gen_params]() -> int {
if (gen_params.width > 0)
return gen_params.width;
if (default_gen_params.width > 0)
return default_gen_params.width;
return 512;
};
auto get_resolved_height = [&gen_params, &default_gen_params]() -> int {
if (gen_params.height > 0)
return gen_params.height;
if (default_gen_params.height > 0)
return default_gen_params.height;
return 512;
};

if (img2img) {
auto decode_image = [](sd_image_t& image, std::string encoded) -> bool {
auto decode_image = [&gen_params](sd_image_t& image, std::string encoded) -> bool {
// remove data URI prefix if present ("data:image/png;base64,")
auto comma_pos = encoded.find(',');
if (comma_pos != std::string::npos) {
encoded = encoded.substr(comma_pos + 1);
}
std::vector<uint8_t> img_data = base64_decode(encoded);
if (!img_data.empty()) {
int img_w = image.width;
int img_h = image.height;
int expected_width = 0;
int expected_height = 0;
if (gen_params.width_and_height_are_set()) {
expected_width = gen_params.width;
expected_height = gen_params.height;
}
int img_w = expected_width;
int img_h = expected_height;

uint8_t* raw_data = load_image_from_memory(
(const char*)img_data.data(), (int)img_data.size(),
img_w, img_h,
image.width, image.height, image.channel);
expected_width, expected_height, image.channel);
if (raw_data) {
image = {(uint32_t)img_w, (uint32_t)img_h, image.channel, raw_data};
gen_params.set_width_and_height_if_unset(image.width, image.height);
return true;
}
}
Expand All @@ -953,17 +978,19 @@ int main(int argc, const char** argv) {
}
}
} else {
mask_data = std::vector<uint8_t>(width * height, 255);
mask_image.width = width;
mask_image.height = height;
int m_width = get_resolved_width();
int m_height = get_resolved_height();
mask_data = std::vector<uint8_t>(m_width * m_height, 255);
mask_image.width = m_width;
mask_image.height = m_height;
mask_image.channel = 1;
mask_image.data = mask_data.data();
}

if (j.contains("extra_images") && j["extra_images"].is_array()) {
for (auto extra_image : j["extra_images"]) {
std::string encoded = extra_image.get<std::string>();
sd_image_t tmp_image = {(uint32_t)gen_params.width, (uint32_t)gen_params.height, 3, nullptr};
sd_image_t tmp_image = {0, 0, 3, nullptr};
if (decode_image(tmp_image, encoded)) {
ref_images.push_back(tmp_image);
}
Expand All @@ -989,8 +1016,8 @@ int main(int argc, const char** argv) {
gen_params.auto_resize_ref_image,
gen_params.increase_ref_index,
mask_image,
gen_params.width,
gen_params.height,
get_resolved_width(),
get_resolved_height(),
gen_params.sample_params,
gen_params.strength,
gen_params.seed,
Expand Down
Loading