Fabio
c36510c21e
Some checks failed
Linux Build / Linux cpu (push) Has been cancelled
Linux Build / Linux cpu-clang (push) Has been cancelled
Linux Build / Linux vulkan (push) Has been cancelled
mac-build / macOS CPU (push) Has been cancelled
Nix Build / Nix (cpu) (push) Has been cancelled
Nix Build / Nix (vulkan) (push) Has been cancelled
windows-build / Windows CPU (push) Has been cancelled
386 lines
15 KiB
C++
386 lines
15 KiB
C++
#include "engine/framework/audio/wav_reader.h"
|
|
#include "engine/framework/audio/wav_writer.h"
|
|
#include "engine/framework/debug/trace.h"
|
|
#include "engine/framework/io/json.h"
|
|
#include "engine/framework/runtime/registry.h"
|
|
#include "engine/framework/runtime/session.h"
|
|
#include "engine/models/voxcpm2/session.h"
|
|
|
|
#include <algorithm>
|
|
#include <chrono>
|
|
#include <cmath>
|
|
#include <filesystem>
|
|
#include <iostream>
|
|
#include <stdexcept>
|
|
#include <string>
|
|
#include <utility>
|
|
#include <vector>
|
|
|
|
namespace {
|
|
|
|
std::string arg_value(int argc, char **argv, const std::string &name,
|
|
const std::string &fallback) {
|
|
for (int i = 1; i + 1 < argc; ++i) {
|
|
if (argv[i] == name) {
|
|
return argv[i + 1];
|
|
}
|
|
}
|
|
return fallback;
|
|
}
|
|
|
|
int int_arg(int argc, char **argv, const std::string &name, int fallback) {
|
|
return std::stoi(arg_value(argc, argv, name, std::to_string(fallback)));
|
|
}
|
|
|
|
std::vector<std::pair<std::string, std::string>>
|
|
parse_session_options(int argc, char **argv) {
|
|
std::vector<std::pair<std::string, std::string>> out;
|
|
for (int i = 1; i + 1 < argc; ++i) {
|
|
if (std::string(argv[i]) != "--session-option") {
|
|
continue;
|
|
}
|
|
const std::string option = argv[i + 1];
|
|
const size_t eq = option.find('=');
|
|
if (eq == std::string::npos || eq == 0) {
|
|
throw std::runtime_error("invalid VoxCPM2 --session-option: " + option);
|
|
}
|
|
out.emplace_back(option.substr(0, eq), option.substr(eq + 1));
|
|
}
|
|
return out;
|
|
}
|
|
|
|
engine::core::BackendType parse_backend(const std::string &value) {
|
|
if (value == "cpu") {
|
|
return engine::core::BackendType::Cpu;
|
|
}
|
|
if (value == "cuda") {
|
|
return engine::core::BackendType::Cuda;
|
|
}
|
|
if (value == "vulkan") {
|
|
return engine::core::BackendType::Vulkan;
|
|
}
|
|
if (value == "best") {
|
|
return engine::core::BackendType::BestAvailable;
|
|
}
|
|
throw std::runtime_error("unsupported VoxCPM2 warmbench backend: " + value);
|
|
}
|
|
|
|
std::filesystem::path resolve_path(const std::string &value) {
|
|
std::filesystem::path path(value);
|
|
if (path.is_absolute()) {
|
|
return path;
|
|
}
|
|
return std::filesystem::current_path() / path;
|
|
}
|
|
|
|
engine::runtime::AudioBuffer read_audio(const std::string &value) {
|
|
const auto wav = engine::audio::read_wav_f32(resolve_path(value));
|
|
return engine::runtime::AudioBuffer{wav.sample_rate, wav.channels,
|
|
wav.samples};
|
|
}
|
|
|
|
std::string optional_string(const engine::io::json::Value &object,
|
|
const std::string &key) {
|
|
const auto *value = object.find(key);
|
|
return value == nullptr || value->is_null() ? std::string{}
|
|
: value->as_string();
|
|
}
|
|
|
|
std::string required_string(const engine::io::json::Value &object,
|
|
const std::string &key) {
|
|
const auto out = optional_string(object, key);
|
|
if (out.empty()) {
|
|
throw std::runtime_error("VoxCPM2 warmbench request missing field: " + key);
|
|
}
|
|
return out;
|
|
}
|
|
|
|
std::string option_text(const engine::io::json::Value &value) {
|
|
if (value.is_bool()) {
|
|
return value.as_bool() ? "true" : "false";
|
|
}
|
|
if (value.is_number()) {
|
|
return engine::io::json::stringify_number(value.as_number());
|
|
}
|
|
return value.as_string();
|
|
}
|
|
|
|
void set_optional_option(engine::runtime::TaskRequest &request,
|
|
const engine::io::json::Value &object,
|
|
const std::string &source,
|
|
const std::string &target) {
|
|
const auto *value = object.find(source);
|
|
if (value != nullptr && !value->is_null()) {
|
|
request.options[target] = option_text(*value);
|
|
}
|
|
}
|
|
|
|
void reject_enabled_bool(const engine::io::json::Value &object,
|
|
const std::string &key,
|
|
const std::string &message) {
|
|
const auto *value = object.find(key);
|
|
if (value != nullptr && value->is_bool() && value->as_bool()) {
|
|
throw std::runtime_error(message);
|
|
}
|
|
}
|
|
|
|
engine::runtime::TaskRequest make_request(const engine::io::json::Value &object,
|
|
const std::string &noise_file) {
|
|
engine::runtime::TaskRequest request;
|
|
request.text_input = engine::runtime::Transcript{required_string(object, "text"), ""};
|
|
if (const auto prompt_wav = optional_string(object, "prompt_wav_path");
|
|
!prompt_wav.empty()) {
|
|
request.audio_input = read_audio(prompt_wav);
|
|
}
|
|
if (const auto reference_wav = optional_string(object, "reference_wav_path");
|
|
!reference_wav.empty()) {
|
|
request.voice = engine::runtime::VoiceCondition{};
|
|
request.voice->speaker = engine::runtime::VoiceReference{};
|
|
request.voice->speaker->audio = read_audio(reference_wav);
|
|
}
|
|
reject_enabled_bool(object, "normalize",
|
|
"VoxCPM2 C++ warmbench does not implement normalize");
|
|
set_optional_option(request, object, "prompt_text", "prompt_text");
|
|
set_optional_option(request, object, "cfg_value", "guidance_scale");
|
|
set_optional_option(request, object, "inference_timesteps",
|
|
"num_inference_steps");
|
|
set_optional_option(request, object, "min_len", "min_tokens");
|
|
set_optional_option(request, object, "max_len", "max_tokens");
|
|
set_optional_option(request, object, "retry_badcase", "retry_badcase");
|
|
set_optional_option(request, object, "retry_badcase_max_times",
|
|
"retry_badcase_max_times");
|
|
set_optional_option(request, object, "retry_badcase_ratio_threshold",
|
|
"retry_badcase_ratio_threshold");
|
|
set_optional_option(request, object, "seed", "seed");
|
|
set_optional_option(request, object, "denoise", "denoise");
|
|
set_optional_option(request, object, "load_denoiser", "load_denoiser");
|
|
if (!noise_file.empty()) {
|
|
request.options["voxcpm2.cfm_noise_file"] = noise_file;
|
|
}
|
|
return request;
|
|
}
|
|
|
|
std::vector<engine::runtime::TaskRequest>
|
|
parse_requests(const std::string &request_sequence_json,
|
|
const std::string &noise_file) {
|
|
const auto root = engine::io::json::parse(request_sequence_json);
|
|
std::vector<engine::runtime::TaskRequest> requests;
|
|
for (const auto &item : root.as_array()) {
|
|
requests.push_back(make_request(item, noise_file));
|
|
}
|
|
return requests;
|
|
}
|
|
|
|
engine::io::json::Value number(double value) {
|
|
return engine::io::json::Value::make_number(value);
|
|
}
|
|
|
|
engine::io::json::Value string(std::string value) {
|
|
return engine::io::json::Value::make_string(std::move(value));
|
|
}
|
|
|
|
engine::io::json::Value
|
|
audio_summary_json(const engine::runtime::AudioBuffer &audio) {
|
|
if (audio.samples.empty()) {
|
|
throw std::runtime_error("VoxCPM2 warmbench received empty audio output");
|
|
}
|
|
double sum = 0.0;
|
|
double abs_sum = 0.0;
|
|
double sq_sum = 0.0;
|
|
float min_value = audio.samples.front();
|
|
float max_value = audio.samples.front();
|
|
for (const float sample : audio.samples) {
|
|
sum += static_cast<double>(sample);
|
|
abs_sum += std::abs(static_cast<double>(sample));
|
|
sq_sum += static_cast<double>(sample) * static_cast<double>(sample);
|
|
min_value = std::min(min_value, sample);
|
|
max_value = std::max(max_value, sample);
|
|
}
|
|
const auto channels = std::max(1, audio.channels);
|
|
const auto frames =
|
|
static_cast<double>(audio.samples.size() / static_cast<size_t>(channels));
|
|
const auto count = static_cast<double>(audio.samples.size());
|
|
return engine::io::json::Value::make_object({
|
|
{"sample_rate", number(static_cast<double>(audio.sample_rate))},
|
|
{"channels", number(static_cast<double>(audio.channels))},
|
|
{"samples", number(count)},
|
|
{"frames", number(frames)},
|
|
{"duration_sec",
|
|
number(audio.sample_rate > 0 ? frames / audio.sample_rate : 0.0)},
|
|
{"sum", number(sum)},
|
|
{"mean_abs", number(abs_sum / count)},
|
|
{"rms", number(std::sqrt(sq_sum / count))},
|
|
{"min", number(min_value)},
|
|
{"max", number(max_value)},
|
|
});
|
|
}
|
|
|
|
engine::io::json::Value step_json(const engine::runtime::TaskResult &result,
|
|
int request_index, double wall_ms,
|
|
const std::filesystem::path &audio_path,
|
|
int64_t chunk_count) {
|
|
if (!result.audio_output.has_value()) {
|
|
throw std::runtime_error("VoxCPM2 warmbench expected audio output");
|
|
}
|
|
engine::io::json::Value::Object stem{
|
|
{"name", string("audio")},
|
|
{"summary", audio_summary_json(*result.audio_output)},
|
|
};
|
|
if (!audio_path.empty()) {
|
|
stem.emplace("audio", string(audio_path.string()));
|
|
}
|
|
return engine::io::json::Value::make_object({
|
|
{"request_index", number(static_cast<double>(request_index))},
|
|
{"stems", engine::io::json::Value::make_array(
|
|
{engine::io::json::Value::make_object(std::move(stem))})},
|
|
{"metrics", engine::io::json::Value::make_object(
|
|
{{"wall_ms", number(wall_ms)},
|
|
{"chunk_count", number(static_cast<double>(chunk_count))}})},
|
|
});
|
|
}
|
|
|
|
} // namespace
|
|
|
|
int main(int argc, char **argv) {
|
|
try {
|
|
const std::filesystem::path model_path =
|
|
arg_value(argc, argv, "--model", "models/VoxCPM2");
|
|
const std::string backend_name = arg_value(argc, argv, "--backend", "cuda");
|
|
const int device = int_arg(argc, argv, "--device", 0);
|
|
const int threads = int_arg(argc, argv, "--threads", 8);
|
|
const int warmup = int_arg(argc, argv, "--warmup", 0);
|
|
const int iterations = int_arg(argc, argv, "--iterations", 1);
|
|
const std::string run_mode = arg_value(argc, argv, "--run-mode", "offline");
|
|
const std::string request_sequence_json =
|
|
arg_value(argc, argv, "--request-sequence-json", "");
|
|
const std::filesystem::path output_dir =
|
|
arg_value(argc, argv, "--output-dir", "");
|
|
const std::filesystem::path timing_path =
|
|
arg_value(argc, argv, "--timing-file",
|
|
"/tmp/voxcpm2_warm_bench_timing.log");
|
|
const std::string noise_file = arg_value(argc, argv, "--noise-file", "");
|
|
if (request_sequence_json.empty()) {
|
|
throw std::runtime_error("VoxCPM2 warmbench requires --request-sequence-json");
|
|
}
|
|
engine::debug::configure_logging(
|
|
engine::debug::LoggingConfig{true, timing_path.string()});
|
|
|
|
const auto session_option_overrides = parse_session_options(argc, argv);
|
|
engine::runtime::ModelLoadRequest load_request;
|
|
load_request.model_path = model_path;
|
|
load_request.family_hint = "voxcpm2";
|
|
auto registry = engine::runtime::make_default_registry();
|
|
auto model = registry.load(load_request);
|
|
|
|
engine::runtime::SessionOptions options;
|
|
options.backend.type = parse_backend(backend_name);
|
|
options.backend.device = device;
|
|
options.backend.threads = threads;
|
|
options.options.insert_or_assign("voxcpm2.weight_type", "f32");
|
|
for (const auto &[key, value] : session_option_overrides) {
|
|
if (key == "voxcpm2.weight_type" && value != "f32") {
|
|
throw std::runtime_error(
|
|
"VoxCPM2 warmbench forces voxcpm2.weight_type=f32 for parity");
|
|
}
|
|
options.options.insert_or_assign(key, value);
|
|
}
|
|
|
|
const bool streaming = run_mode == "streaming";
|
|
if (!streaming && run_mode != "offline") {
|
|
throw std::runtime_error("unsupported VoxCPM2 warmbench run mode: " +
|
|
run_mode);
|
|
}
|
|
auto session_base = model->create_task_session(
|
|
{engine::runtime::VoiceTaskKind::Tts,
|
|
streaming ? engine::runtime::RunMode::Streaming
|
|
: engine::runtime::RunMode::Offline},
|
|
options);
|
|
auto *offline_session =
|
|
dynamic_cast<engine::runtime::IOfflineVoiceTaskSession *>(
|
|
session_base.get());
|
|
auto *streaming_session =
|
|
dynamic_cast<engine::runtime::IStreamingVoiceTaskSession *>(
|
|
session_base.get());
|
|
if (streaming) {
|
|
if (streaming_session == nullptr) {
|
|
throw std::runtime_error(
|
|
"loaded VoxCPM2 session is not streaming-capable");
|
|
}
|
|
} else if (offline_session == nullptr) {
|
|
throw std::runtime_error("loaded VoxCPM2 session is not offline-capable");
|
|
}
|
|
|
|
const auto requests = parse_requests(request_sequence_json, noise_file);
|
|
if (requests.empty()) {
|
|
throw std::runtime_error("VoxCPM2 warmbench request sequence is empty");
|
|
}
|
|
session_base->prepare(engine::runtime::build_preparation_request(requests.front()));
|
|
for (int i = 0; i < warmup; ++i) {
|
|
if (streaming) {
|
|
streaming_session->start_stream(requests.front());
|
|
while (streaming_session->next_stream_event().has_value()) {
|
|
}
|
|
(void)streaming_session->finish_stream();
|
|
} else {
|
|
(void)offline_session->run(requests.front());
|
|
}
|
|
}
|
|
|
|
if (!output_dir.empty()) {
|
|
std::filesystem::create_directories(output_dir);
|
|
}
|
|
|
|
engine::io::json::Value::Array steps;
|
|
steps.reserve(requests.size());
|
|
for (size_t request_index = 0; request_index < requests.size();
|
|
++request_index) {
|
|
engine::runtime::TaskResult last_result;
|
|
double total_ms = 0.0;
|
|
for (int iteration = 0; iteration < std::max(1, iterations); ++iteration) {
|
|
const auto started = std::chrono::steady_clock::now();
|
|
if (streaming) {
|
|
streaming_session->start_stream(requests[request_index]);
|
|
while (streaming_session->next_stream_event().has_value()) {
|
|
}
|
|
last_result = streaming_session->finish_stream();
|
|
} else {
|
|
last_result = offline_session->run(requests[request_index]);
|
|
}
|
|
const auto ended = std::chrono::steady_clock::now();
|
|
total_ms +=
|
|
std::chrono::duration<double, std::milli>(ended - started).count();
|
|
}
|
|
if (!last_result.audio_output.has_value()) {
|
|
throw std::runtime_error("VoxCPM2 warmbench expected audio output");
|
|
}
|
|
const double wall_ms =
|
|
total_ms / static_cast<double>(std::max(1, iterations));
|
|
std::filesystem::path audio_path;
|
|
if (!output_dir.empty()) {
|
|
audio_path = output_dir / ("audio_" + std::to_string(request_index) + ".wav");
|
|
engine::audio::write_pcm16_wav(
|
|
audio_path, last_result.audio_output->sample_rate,
|
|
last_result.audio_output->channels, last_result.audio_output->samples);
|
|
}
|
|
std::cout << "voxcpm2.wall_ms=" << wall_ms << "\n";
|
|
steps.push_back(
|
|
step_json(last_result, static_cast<int>(request_index), wall_ms,
|
|
audio_path,
|
|
static_cast<int64_t>(
|
|
last_result.named_audio_outputs.size())));
|
|
}
|
|
|
|
const auto summary = engine::io::json::Value::make_object({
|
|
{"family", string("voxcpm2")},
|
|
{"backend", string(backend_name)},
|
|
{"mode", string(run_mode)},
|
|
{"sequence_steps", engine::io::json::Value::make_array(std::move(steps))},
|
|
});
|
|
std::cout << "summary_json=" << engine::io::json::stringify(summary) << "\n";
|
|
engine::debug::reset_logging();
|
|
return 0;
|
|
} catch (const std::exception &ex) {
|
|
std::cerr << "voxcpm2_warm_bench failed: " << ex.what() << "\n";
|
|
return 1;
|
|
}
|
|
}
|