voice_audio_multi/include/engine/models/confucius4_tts/tokenizer_text.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

31 lines
1.1 KiB
C++

#pragma once
#include "engine/framework/tokenizers/sentencepiece.h"
#include "engine/models/confucius4_tts/assets.h"
#include "engine/models/confucius4_tts/types.h"
#include <memory>
#include <string>
#include <vector>
namespace engine::models::confucius4_tts {
class ConfuciusTextTokenizer {
public:
explicit ConfuciusTextTokenizer(std::shared_ptr<const ConfuciusAssets> assets);
std::string normalize(const std::string & text, const std::string & language) const;
std::string format_prompt(const std::string & text, const std::string & language) const;
std::vector<int32_t> encode(const std::string & text) const;
std::vector<int32_t> encode_without_special_tokens(const std::string & text) const;
std::vector<int32_t> encode_prompt(const std::string & text, const std::string & language) const;
std::vector<ConfuciusTextSegment> segment_request(const ConfuciusRequest & request) const;
private:
std::shared_ptr<const ConfuciusAssets> assets_;
std::vector<engine::tokenizers::SentencePiecePiece> pieces_;
int32_t bos_id_ = 1;
int32_t eos_id_ = 2;
};
} // namespace engine::models::confucius4_tts