voice_audio_multi/include/engine/models/midashenglm_gen/assets.h
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
Initial import: audio.cpp with Orange Pi documentation
2026-11-04 19:05:22 +01:00

41 lines
1.2 KiB
C++

#pragma once
#include "engine/framework/assets/resource_bundle.h"
#include "engine/framework/assets/tensor_source.h"
#include <filesystem>
#include <memory>
#include <string>
namespace engine::models::midashenglm_gen {
struct MiDashengLmGenConfig {
int64_t vocab_size = 151697;
int64_t hidden_size = 2048;
int64_t audio_embedding_size = 768;
int64_t target_embedding_size = 768;
int64_t sequence_length = 500;
int64_t patch_size = 5;
int64_t block_size = 3200;
int64_t flow_hidden_size = 2048;
int64_t flow_depth = 16;
int64_t flow_heads = 8;
int64_t flow_intermediate_size = 8192;
int64_t decoder_layers = 8;
int64_t decoder_intermediate_size = 1536;
int64_t istft_n_fft = 1280;
int64_t istft_hop = 320;
int64_t sample_rate = 16000;
};
struct MiDashengLmGenAssets {
std::filesystem::path model_root;
MiDashengLmGenConfig config;
engine::assets::ResourceBundle resources;
std::shared_ptr<const engine::assets::TensorSource> weights;
};
std::shared_ptr<const MiDashengLmGenAssets> load_midashenglm_gen_assets(
const std::filesystem::path & model_path);
} // namespace engine::models::midashenglm_gen