Skip to content

feat: support previews streaming in openAI API #2097

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

Open
stduhpf wants to merge 1 commit into
base: master
Choose a base branch
from
+162 −2
Open

feat: support previews streaming in openAI API #2097

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
164 changes: 162 additions & 2 deletions examples/server/routes_openai.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -23,13 +23,15 @@ static std::string extract_and_remove_sd_cpp_extra_args(std::string& text) {
static bool build_openai_generation_request(const httplib::Request& req,
ServerRuntime& runtime,
ImgGenJobRequest& request,
bool& stream,
std::string& error_message) {
if (req.body.empty()) {
error_message = "empty body";
return false;
}

json j = json::parse(req.body);
stream = j.value("stream", false);
std::string prompt = j.value("prompt", "");
int n = std::max(1, j.value("n", 1));
std::string size = j.value("size", "");
Expand Down Expand Up @@ -80,12 +82,21 @@ static bool build_openai_generation_request(const httplib::Request& req,
static bool build_openai_edit_request(const httplib::Request& req,
ServerRuntime& runtime,
ImgGenJobRequest& request,
bool& stream,
std::string& error_message) {
if (!req.is_multipart_form_data()) {
error_message = "Content-Type must be multipart/form-data";
return false;
}

// Parse stream boolean fro m form field
stream = false;
if (req.form.has_field("stream")) {
std::string stream_val = req.form.get_field("stream");
std::transform(stream_val.begin(), stream_val.end(), stream_val.begin(), ::tolower);
stream = (stream_val == "true" || stream_val == "1");
}

std::string prompt = req.form.get_field("prompt");
if (prompt.empty()) {
error_message = "prompt required";
Expand Down Expand Up @@ -247,6 +258,123 @@ static bool execute_sync_img_gen_request(ServerRuntime& runtime,
return true;
}

struct PreviewStreamContext {
httplib::DataSink* sink = nullptr;
const ImgGenJobRequest* request = nullptr;
};

static void preview_callback(int step, int frame_count, sd_image_t* frames, bool is_noisy, void* data) {
if (!data || frame_count <= 0 || !frames || !frames[0].data) {
return;
}

auto* ctx = static_cast<PreviewStreamContext*>(data);
if (!ctx->sink || !ctx->sink->is_writable()) {
return; // Client disconnected
}

auto image_bytes = encode_image_to_vector(
EncodedImageFormat::JPEG,
frames[0].data,
frames[0].width,
frames[0].height,
frames[0].channel,
"",
80);

if (image_bytes.empty()) {
return;
}

json preview_chunk;
preview_chunk["object"] = "image.preview";
preview_chunk["step"] = step;
preview_chunk["data"] = json::array({{{"b64_json", base64_encode(image_bytes)}}});

std::string sse_data = "data: " + preview_chunk.dump() + "\n\n";
ctx->sink->write(sse_data.data(), sse_data.size());
}

static void execute_streaming_img_gen(ServerRuntime& runtime,
ImgGenJobRequest& request,
httplib::DataSink& sink) {
PreviewStreamContext preview_ctx{&sink, &request};
sd_img_gen_params_t img_gen_params = request.to_sd_img_gen_params_t();
SDImageVec results;
int num_results = 0;

{
std::lock_guard<std::mutex> lock(*runtime.sd_ctx_mutex);

// TODO: support other preview methods?
sd_set_preview_callback(preview_callback, PREVIEW_PROJ, 1, true, false, &preview_ctx);

try {
sd_image_t* raw_results = nullptr;
if (!generate_image(runtime.sd_ctx, &img_gen_params, &raw_results, &num_results)) {
raw_results = nullptr;
num_results = 0;
}
results.adopt(raw_results, num_results);
} catch (...) {
sd_set_preview_callback(nullptr, PREVIEW_NONE, 0, false, false, nullptr);
throw;
}

sd_set_preview_callback(nullptr, PREVIEW_NONE, 0, false, false, nullptr);
}

if (results.empty() || !sink.is_writable()) {
if (sink.is_writable()) {
json err = {{"error", "generate_image returned no results"}};
std::string err_msg = "data: " + err.dump() + "\n\n";
sink.write(err_msg.data(), err_msg.size());
}
sink.done();
return;
}

json final_out;
final_out["created"] = static_cast<long long>(std::time(nullptr));
final_out["data"] = json::array();
final_out["output_format"] = request.output_format;

int result_count = results.count();
int images_per_batch = request.gen_params.batch_count > 0
? std::max(1, result_count / request.gen_params.batch_count)
: 1;

for (int i = 0; i < result_count; ++i) {
if (results[i].data == nullptr)
continue;

std::string params = request.gen_params.embed_image_metadata
? get_image_params(*runtime.ctx_params, request.gen_params,
request.gen_params.seed + i / images_per_batch)
: "";
auto image_bytes = encode_image_to_vector(
request.output_format == "jpeg" ? EncodedImageFormat::JPEG
: request.output_format == "webp" ? EncodedImageFormat::WEBP
: EncodedImageFormat::PNG,
results[i].data, results[i].width, results[i].height,
results[i].channel, params, request.output_compression);

if (!image_bytes.empty()) {
final_out["data"].push_back({{"b64_json", base64_encode(image_bytes)}});
}
}

if (sink.is_writable()) {
std::string final_sse = "data: " + final_out.dump() + "\n\n";
sink.write(final_sse.data(), final_sse.size());

const std::string done_msg = "data: [DONE]\n\n";
sink.write(done_msg.data(), done_msg.size());
}

sink.done();
}

void register_openai_api_endpoints(httplib::Server& svr, ServerRuntime& rt) {
ServerRuntime* runtime = &rt;

Expand All @@ -267,14 +395,29 @@ void register_openai_api_endpoints(httplib::Server& svr, ServerRuntime& rt) {

ImgGenJobRequest request;
std::string error_message;
if (!build_openai_generation_request(req, *runtime, request, error_message)) {
bool stream = false;
if (!build_openai_generation_request(req, *runtime, request, stream, error_message)) {
res.status = 400;
res.set_content(json({{"error", error_message}}).dump(), "application/json");
return;
}

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

if (stream) {
res.set_header("Cache-Control", "no-cache");
res.set_header("Connection", "keep-alive");
res.set_chunked_content_provider(
"text/event-stream",
[runtime, request](size_t /*offset*/, httplib::DataSink& sink) mutable {
execute_streaming_img_gen(*runtime, request, sink);
return false;
}
);
return;
}


SDImageVec results;
if (!execute_sync_img_gen_request(*runtime, request, results, error_message)) {
res.status = 500;
Expand Down Expand Up @@ -341,14 +484,31 @@ void register_openai_api_endpoints(httplib::Server& svr, ServerRuntime& rt) {

ImgGenJobRequest request;
std::string error_message;
if (!build_openai_edit_request(req, *runtime, request, error_message)) {
bool stream = false;
if (!build_openai_edit_request(req, *runtime, request, stream, error_message)) {
res.status = 400;
res.set_content(json({{"error", error_message}}).dump(), "application/json");
return;
}

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

if (stream) {
res.set_header("Cache-Control", "no-cache");
res.set_header("Connection", "keep-alive");

auto req_ptr = std::make_shared<ImgGenJobRequest>(std::move(request));
res.set_chunked_content_provider(
"text/event-stream",
[runtime, req_ptr](size_t /*offset*/, httplib::DataSink& sink) {
execute_streaming_img_gen(*runtime, *req_ptr, sink);
return false;
}
);
return;
}

// Synchronous path
SDImageVec results;
if (!execute_sync_img_gen_request(*runtime, request, results, error_message)) {
res.status = 500;
Expand Down
Loading