doubaospeech

package module
v0.0.0-...-2a34a3e Latest Latest
Warning

This package is not in the latest version of its module.

 Go to latest Published: May 12, 2026 License: MIT Imports: 28 Imported by: 0

README

doubao-speech-go

CI Pass Code Scan

Go SDK for Doubao/Volc speech APIs.

Features

  • ASR V2 SAUC WebSocket streaming
  • TTS V2 HTTP stream
  • TTS V2 WebSocket (unidirectional/bidirectional flows)
  • Realtime session API
  • Voice clone upload + polling workflow

Requirements

  • Go 1.26+
  • Git LFS (for embedded audio fixture)
git lfs pull

Install

go get github.com/GizClaw/doubao-speech-go

Quick Start

Run the ASR V2 example:

DOUBAO_APP_ID=<your_app_id> \
DOUBAO_API_KEY=<your_api_key> \
go run ./examples/asr_v2_sauc_ws

Run the Voice Clone example:

DOUBAO_APP_ID=<your_app_id> \
DOUBAO_TOKEN=<your_token> \
go run ./examples/voice_clone -speaker-id <speaker_id> -audio /path/to/sample.wav

Development

Format:

gofmt -w *.go internal/auth/*.go internal/protocol/*.go internal/transport/*.go internal/util/*.go examples/asr_v2_sauc_ws/*.go examples/realtime/*.go examples/tts_v2/http_stream/*.go examples/tts_v2/websocket/*.go examples/voice_clone/*.go

Build / test / vet:

go build ./...
go test ./...
go vet ./...

Run one test:

go test . -run TestVoiceCloneUploadAndWaitSuccess -count=1 -v

Documentation

  • docs/asr_v2_sauc_ws.md
  • docs/tts_v2_http_stream.md
  • docs/tts_v2_websocket.md
  • docs/voice_clone.md

License

This project is licensed under the MIT License. See LICENSE.

Documentation

Index

Constants

View Source 
const (
	AppKeyRealtime = "PlgvMymc7f3tQnJ6"
	AppKeyPodcast  = "aGjiRDfUWi"
)

V2/V3 fixed app keys (official constants, not user credentials).

View Source 
const (
	ResourceTTSV1        = "seed-tts-1.0"
	ResourceTTSV1Concurr = "seed-tts-1.0-concurr"
	ResourceTTSV2        = "seed-tts-2.0"
	ResourceTTSV2Concurr = "seed-tts-2.0-concurr"
	ResourceVoiceCloneV1 = "seed-icl-1.0"
	ResourceVoiceCloneV2 = "seed-icl-2.0"

	ResourceASRStream   = "volc.bigasr.sauc.duration"
	ResourceASRStreamV2 = "volc.seedasr.sauc.duration"
	ResourceASRFile     = "volc.bigasr.auc.duration"

	ResourceRealtime    = "volc.speech.dialog"
	ResourcePodcast     = "volc.service_type.10050"
	ResourceTranslation = "volc.megatts.simt"
)

V2/V3 Resource IDs.

View Source 
const (
	CodeSuccess     = 3000
	CodeParamError  = 3001
	CodeAuthError   = 3002
	CodeRateLimit   = 3003
	CodeQuotaExceed = 3004
	CodeServerError = 3005
	CodeASRSuccess  = 1000
)

Variables

This section is empty.

Functions

This section is empty.

Types

type ASRChunk 

type ASRChunk = ASRV2Result

type ASRServiceV2 

type ASRServiceV2 struct {
	// contains filtered or unexported fields
}

ASRServiceV2 provides SAUC WebSocket streaming recognition.

func (*ASRServiceV2) OpenStreamSession 

func (s *ASRServiceV2) OpenStreamSession(ctx context.Context, cfg *ASRV2Config) (*ASRV2Session, error)

OpenStreamSession opens a SAUC V2 WebSocket session.

type ASRV2Config 

type ASRV2Config struct {
	Format     AudioFormat `json:"format" yaml:"format"`
	SampleRate SampleRate  `json:"sample_rate" yaml:"sample_rate"`
	Channel    int         `json:"channel,omitempty" yaml:"channel,omitempty"`
	Channels   int         `json:"channels,omitempty" yaml:"channels,omitempty"` // Backward-compatible alias field.
	Bits       int         `json:"bits,omitempty" yaml:"bits,omitempty"`
	Language   Language    `json:"language,omitempty" yaml:"language,omitempty"`

	EnableITN         bool     `json:"enable_itn,omitempty" yaml:"enable_itn,omitempty"`
	EnablePunc        bool     `json:"enable_punc,omitempty" yaml:"enable_punc,omitempty"`
	EnableDiarization bool     `json:"enable_diarization,omitempty" yaml:"enable_diarization,omitempty"`
	SpeakerNum        int      `json:"speaker_num,omitempty" yaml:"speaker_num,omitempty"`
	Hotwords          []string `json:"hotwords,omitempty" yaml:"hotwords,omitempty"`
	ResultType        string   `json:"result_type,omitempty" yaml:"result_type,omitempty"` // single/full

	ResourceID string `json:"resource_id,omitempty" yaml:"resource_id,omitempty"`
}

ASRV2Config is SAUC V2 streaming session config.

type ASRV2Result 

type ASRV2Result struct {
	Text       string           `json:"text"`
	Utterances []ASRV2Utterance `json:"utterances,omitempty"`
	IsFinal    bool             `json:"is_final"`
	Duration   int              `json:"duration,omitempty"`
	ReqID      string           `json:"reqid,omitempty"`
	TraceID    string           `json:"trace_id,omitempty"`
	LogID      string           `json:"log_id,omitempty"`
	ConnectID  string           `json:"connect_id,omitempty"`
}

ASRV2Result is one parsed server response.

type ASRV2Session 

type ASRV2Session struct {
	// contains filtered or unexported fields
}

ASRV2Session represents one streaming recognition session.

func (*ASRV2Session) Close 

func (s *ASRV2Session) Close() error

Close closes the session.

func (*ASRV2Session) Recv 

func (s *ASRV2Session) Recv() iter.Seq2[*ASRV2Result, error]

Recv yields recognition results as a stream.

func (*ASRV2Session) SendAudio 

func (s *ASRV2Session) SendAudio(ctx context.Context, audio []byte, isLast bool) error

SendAudio sends one audio chunk.

isLast=true marks the last frame (flags=2).

type ASRV2Utterance 

type ASRV2Utterance struct {
	Text       string      `json:"text"`
	StartTime  int         `json:"start_time"`
	EndTime    int         `json:"end_time"`
	Definite   bool        `json:"definite"`
	SpeakerID  string      `json:"speaker_id,omitempty"`
	Words      []ASRV2Word `json:"words,omitempty"`
	Confidence float64     `json:"confidence,omitempty"`
}

ASRV2Utterance contains utterance-level info.

type ASRV2Word 

type ASRV2Word struct {
	Text      string  `json:"text"`
	StartTime int     `json:"start_time"`
	EndTime   int     `json:"end_time"`
	Conf      float64 `json:"conf,omitempty"`
}

ASRV2Word contains word-level timing info.

type AudioFormat 

type AudioFormat string

AudioFormat represents audio encoding format. 

const (
	FormatPCM AudioFormat = "pcm"
	FormatWAV AudioFormat = "wav"
	FormatMP3 AudioFormat = "mp3"
	FormatOGG AudioFormat = "ogg_opus"
	FormatAAC AudioFormat = "aac"
	FormatM4A AudioFormat = "m4a"
)

type Client 

type Client struct {
	// ASR V2 streaming recognition.
	ASR   *ASRServiceV2
	ASRV2 *ASRServiceV2

	// Voice cloning.
	VoiceClone *VoiceCloneService

	// Realtime dialogue.
	Realtime *RealtimeService

	// TTS V2 WebSocket synthesis.
	TTS   *TTSServiceV2
	TTSV2 *TTSServiceV2
	// contains filtered or unexported fields
}

Client is the SDK entry point.

In this migration stage, ASR V2, TTS V2 WS, Voice Clone, and Realtime are implemented.

func NewClient 

func NewClient(appID string, opts ...Option) *Client

NewClient creates an SDK client.

type Error 

type Error struct {
	Code       int    `json:"code"`
	Message    string `json:"message"`
	TraceID    string `json:"trace_id,omitempty"`
	LogID      string `json:"log_id,omitempty"`
	ConnectID  string `json:"connect_id,omitempty"`
	HTTPStatus int    `json:"-"`
	ReqID      string `json:"reqid,omitempty"`
}

Error is the unified error model.

func AsError 

func AsError(err error) (*Error, bool)

AsError converts an error to *Error.

func (*Error) Error 

func (e *Error) Error() string

func (*Error) IsAuthError 

func (e *Error) IsAuthError() bool

func (*Error) IsInvalidParam 

func (e *Error) IsInvalidParam() bool

func (*Error) IsQuotaExceeded 

func (e *Error) IsQuotaExceeded() bool

func (*Error) IsRateLimit 

func (e *Error) IsRateLimit() bool

func (*Error) IsServerError 

func (e *Error) IsServerError() bool

func (*Error) Retryable 

func (e *Error) Retryable() bool

type Language 

type Language string

Language represents recognition language. 

const (
	LanguageZhCN Language = "zh-CN"
	LanguageEnUS Language = "en-US"
	LanguageJaJP Language = "ja-JP"
	LanguageKoKR Language = "ko-KR"
)

type MultipartFile 

type MultipartFile struct {
	FieldName   string
	FileName    string
	ContentType string
	Data        []byte
}

MultipartFile is one file part in multipart/form-data payload.

type Option 

type Option func(*clientConfig)

Option configures Client.

func WithAPIKey 

func WithAPIKey(apiKey string) Option

WithAPIKey sets x-api-key.

func WithBaseURL 

func WithBaseURL(url string) Option

WithBaseURL sets the HTTP base URL.

func WithBearerToken 

func WithBearerToken(token string) Option

WithBearerToken sets Bearer token. V1 header format is `Authorization: Bearer;{token}` (historical convention).

func WithCluster 

func WithCluster(cluster string) Option

WithCluster sets the V1 cluster (kept for backward compatibility).

func WithHTTPClient 

func WithHTTPClient(client *http.Client) Option

WithHTTPClient sets a custom HTTP client.

func WithHTTPTransport 

func WithHTTPTransport(doer transport.HTTPDoer) Option

WithHTTPTransport sets a custom HTTP transport doer.

func WithRealtimeAPIKey 

func WithRealtimeAPIKey(accessKey, appKey string) Option

WithRealtimeAPIKey is a compatibility alias.

func WithResourceID 

func WithResourceID(resourceID string) Option

WithResourceID sets the default resource_id.

func WithTimeout 

func WithTimeout(timeout time.Duration) Option

WithTimeout sets request timeout.

func WithUserID 

func WithUserID(userID string) Option

WithUserID sets user.uid.

func WithV2APIKey 

func WithV2APIKey(accessKey, appKey string) Option

WithV2APIKey sets V2/V3 authentication.

func WithWebSocketURL 

func WithWebSocketURL(url string) Option

WithWebSocketURL sets the WebSocket base URL.

type RealtimeASRConfig 

type RealtimeASRConfig struct {
	Language Language       `json:"language,omitempty" yaml:"language,omitempty"`
	Extra    map[string]any `json:"extra,omitempty" yaml:"extra,omitempty"`
}

RealtimeASRConfig configures ASR behavior.

type RealtimeAudioConfig 

type RealtimeAudioConfig struct {
	Channel    int         `json:"channel" yaml:"channel"`
	Format     AudioFormat `json:"format" yaml:"format"`
	SampleRate SampleRate  `json:"sample_rate" yaml:"sample_rate"`
	Bits       int         `json:"bits,omitempty" yaml:"bits,omitempty"`
}

RealtimeAudioConfig describes audio IO parameters.

type RealtimeConfig 

type RealtimeConfig struct {
	ASR    RealtimeASRConfig    `json:"asr" yaml:"asr"`
	TTS    RealtimeTTSConfig    `json:"tts" yaml:"tts"`
	Dialog RealtimeDialogConfig `json:"dialog" yaml:"dialog"`

	Prompt  RealtimePromptConfig          `json:"prompt" yaml:"prompt,omitempty"`
	Props   RealtimeGenerationProps       `json:"props" yaml:"props,omitempty"`
	History []RealtimeConversationMessage `json:"history,omitempty" yaml:"history,omitempty"`

	ResourceID string `json:"resource_id,omitempty" yaml:"resource_id,omitempty"`

	// Local runtime controls (not sent to server).
	EventBuffer         int           `json:"-" yaml:"-"`
	BackpressureTimeout time.Duration `json:"-" yaml:"-"`
}

RealtimeConfig represents one realtime session config.

func DefaultRealtimeConfig 

func DefaultRealtimeConfig() RealtimeConfig

DefaultRealtimeConfig returns a baseline realtime config.

type RealtimeConnection 

type RealtimeConnection struct {
	// contains filtered or unexported fields
}

RealtimeConnection represents an established realtime websocket connection.

func (*RealtimeConnection) Close 

func (c *RealtimeConnection) Close() error

Close closes websocket connection.

func (*RealtimeConnection) StartSession 

func (c *RealtimeConnection) StartSession(ctx context.Context, cfg *RealtimeConfig) (*RealtimeSession, error)

StartSession starts one realtime session on current connection.

type RealtimeConversationMessage 

type RealtimeConversationMessage struct {
	Role    string `json:"role" yaml:"role"`
	Content string `json:"content" yaml:"content"`
}

RealtimeConversationMessage is one dialog history entry.

type RealtimeDialogConfig 

type RealtimeDialogConfig struct {
	BotName           string         `json:"bot_name,omitempty" yaml:"bot_name,omitempty"`
	SystemRole        string         `json:"system_role,omitempty" yaml:"system_role,omitempty"`
	SpeakingStyle     string         `json:"speaking_style,omitempty" yaml:"speaking_style,omitempty"`
	CharacterManifest string         `json:"character_manifest,omitempty" yaml:"character_manifest,omitempty"`
	Extra             map[string]any `json:"extra,omitempty" yaml:"extra,omitempty"`
}

RealtimeDialogConfig configures dialogue behavior.

type RealtimeEvent 

type RealtimeEvent struct {
	Type      RealtimeEventType `json:"type"`
	SessionID string            `json:"session_id,omitempty"`
	ConnectID string            `json:"connect_id,omitempty"`
	Sequence  int32             `json:"sequence,omitempty"`

	Text    string `json:"text,omitempty"`
	Audio   []byte `json:"audio,omitempty"`
	Payload []byte `json:"payload,omitempty"`

	Error   *Error `json:"error,omitempty"`
	IsFinal bool   `json:"is_final,omitempty"`

	ReqID   string `json:"reqid,omitempty"`
	TraceID string `json:"trace_id,omitempty"`
	LogID   string `json:"log_id,omitempty"`
}

RealtimeEvent is one parsed server event.

type RealtimeEventType 

type RealtimeEventType int32

RealtimeEventType represents realtime websocket event ID. 

const (
	// Connection events.
	EventConnectionStarted RealtimeEventType = 50
	EventConnectionFailed  RealtimeEventType = 51
	EventConnectionEnded   RealtimeEventType = 52

	// Session events.
	EventSessionStarted  RealtimeEventType = 150
	EventSessionFinished RealtimeEventType = 152
	EventSessionFailed   RealtimeEventType = 153
	EventUsageResponse   RealtimeEventType = 154

	// ASR events.
	EventASRInfo     RealtimeEventType = 450
	EventASRResponse RealtimeEventType = 451
	EventASREnded    RealtimeEventType = 459

	// TTS events.
	EventTTSStarted    RealtimeEventType = 350
	EventTTSSegmentEnd RealtimeEventType = 351
	EventTTSAudioData  RealtimeEventType = 352
	EventTTSFinished   RealtimeEventType = 359

	// Chat events.
	EventChatResponse RealtimeEventType = 550
	EventChatEnded    RealtimeEventType = 559
)

type RealtimeGenerationProps 

type RealtimeGenerationProps struct {
	Temperature      float64        `json:"temperature,omitempty" yaml:"temperature,omitempty"`
	TopP             float64        `json:"top_p,omitempty" yaml:"top_p,omitempty"`
	MaxTokens        int            `json:"max_tokens,omitempty" yaml:"max_tokens,omitempty"`
	PresencePenalty  float64        `json:"presence_penalty,omitempty" yaml:"presence_penalty,omitempty"`
	FrequencyPenalty float64        `json:"frequency_penalty,omitempty" yaml:"frequency_penalty,omitempty"`
	Extra            map[string]any `json:"extra,omitempty" yaml:"extra,omitempty"`
}

RealtimeGenerationProps controls generation params.

type RealtimePromptConfig 

type RealtimePromptConfig struct {
	System    string            `json:"system,omitempty" yaml:"system,omitempty"`
	Variables map[string]string `json:"variables,omitempty" yaml:"variables,omitempty"`
}

RealtimePromptConfig controls prompt and prompt variables.

type RealtimeService 

type RealtimeService struct {
	// contains filtered or unexported fields
}

RealtimeService provides real-time dialogue operations.

func (*RealtimeService) Connect 

func (s *RealtimeService) Connect(ctx context.Context, cfg *RealtimeConfig) (*RealtimeSession, error)

Connect is a convenience method for Dial + StartSession.

func (*RealtimeService) Dial 

func (s *RealtimeService) Dial(ctx context.Context) (*RealtimeConnection, error)

Dial opens a realtime websocket connection and completes StartConnection handshake.

func (*RealtimeService) OpenSession 

func (s *RealtimeService) OpenSession(ctx context.Context, cfg *RealtimeConfig) (*RealtimeSession, error)

OpenSession is a compatibility alias for Connect.

type RealtimeSession 

type RealtimeSession struct {
	// contains filtered or unexported fields
}

RealtimeSession represents one realtime dialogue session.

func (*RealtimeSession) Close 

func (s *RealtimeSession) Close() error

Close closes current session. It is idempotent.

func (*RealtimeSession) Interrupt 

func (s *RealtimeSession) Interrupt(ctx context.Context) error

Interrupt interrupts current generation (event=102).

func (*RealtimeSession) Recv 

func (s *RealtimeSession) Recv() iter.Seq2[*RealtimeEvent, error]

Recv returns a streaming iterator. Concurrent Recv is not supported.

func (*RealtimeSession) RecvEvent 

func (s *RealtimeSession) RecvEvent(ctx context.Context) (*RealtimeEvent, error)

RecvEvent receives one event. Concurrent Recv/RecvEvent is not supported.

func (*RealtimeSession) ReplaceHistory 

func (s *RealtimeSession) ReplaceHistory(index int, message RealtimeConversationMessage) error

ReplaceHistory replaces one item in local history by index.

func (*RealtimeSession) SayHello 

func (s *RealtimeSession) SayHello(ctx context.Context, content string) error

SayHello sends SayHello event (event=300).

func (*RealtimeSession) SendAudio 

func (s *RealtimeSession) SendAudio(ctx context.Context, audio []byte) error

SendAudio sends one audio chunk (event=200).

func (*RealtimeSession) SendTTSText 

func (s *RealtimeSession) SendTTSText(ctx context.Context, text string) error

SendTTSText sends incremental TTS text (event=500).

func (*RealtimeSession) SendText 

func (s *RealtimeSession) SendText(ctx context.Context, text string) error

SendText sends user text (event=501).

func (*RealtimeSession) SendUserMessage 

func (s *RealtimeSession) SendUserMessage(ctx context.Context, text string) error

SendUserMessage sends one user text with current history/prompt/props snapshot.

func (*RealtimeSession) SessionID 

func (s *RealtimeSession) SessionID() string

SessionID returns current session ID.

func (*RealtimeSession) UpdateHistory 

func (s *RealtimeSession) UpdateHistory(history []RealtimeConversationMessage)

UpdateHistory replaces the whole local history snapshot used by future turns.

func (*RealtimeSession) UpdatePrompt 

func (s *RealtimeSession) UpdatePrompt(prompt RealtimePromptConfig)

UpdatePrompt replaces current prompt config used by future turns.

func (*RealtimeSession) UpdateProps 

func (s *RealtimeSession) UpdateProps(props RealtimeGenerationProps)

UpdateProps replaces current generation props used by future turns.

type RealtimeTTSConfig 

type RealtimeTTSConfig struct {
	Speaker     string              `json:"speaker" yaml:"speaker"`
	AudioConfig RealtimeAudioConfig `json:"audio_config" yaml:"audio_config"`
	Extra       map[string]any      `json:"extra,omitempty" yaml:"extra,omitempty"`
}

RealtimeTTSConfig configures TTS behavior.

type SampleRate 

type SampleRate int

SampleRate represents audio sample rate. 

const (
	SampleRate8000  SampleRate = 8000
	SampleRate16000 SampleRate = 16000
	SampleRate22050 SampleRate = 22050
	SampleRate24000 SampleRate = 24000
	SampleRate32000 SampleRate = 32000
	SampleRate44100 SampleRate = 44100
	SampleRate48000 SampleRate = 48000
)

type StreamASRConfig 

type StreamASRConfig = ASRV2Config

Backward-compatible aliases mapped to V2 types.

type TTSServiceV2 

type TTSServiceV2 struct {
	// contains filtered or unexported fields
}

TTSServiceV2 provides TTS V2 WebSocket streaming synthesis.

func (*TTSServiceV2) OpenStreamSession 

func (s *TTSServiceV2) OpenStreamSession(ctx context.Context, cfg *TTSV2WSConfig) (*TTSV2WSSession, error)

OpenStreamSession opens a TTS V2 bidirectional WebSocket stream session.

func (*TTSServiceV2) Stream 

func (s *TTSServiceV2) Stream(ctx context.Context, req *TTSV2Request) iter.Seq2[*TTSV2Chunk, error]

Stream synthesizes speech with TTS V2 HTTP streaming endpoint.

type TTSV2Chunk 

type TTSV2Chunk struct {
	Audio   []byte `json:"-"`
	IsLast  bool   `json:"is_last"`
	ReqID   string `json:"reqid,omitempty"`
	TraceID string `json:"trace_id,omitempty"`
	LogID   string `json:"log_id,omitempty"`
	Code    int    `json:"code"`
	Message string `json:"message,omitempty"`
}

TTSV2Chunk is one stream chunk from TTS V2 HTTP streaming API.

type TTSV2MixSpeaker 

type TTSV2MixSpeaker struct {
	Speakers []TTSV2MixSpeakerSource `json:"speakers,omitempty" yaml:"speakers,omitempty"`
}

TTSV2MixSpeaker represents mixed-speaker parameters.

type TTSV2MixSpeakerSource 

type TTSV2MixSpeakerSource struct {
	SourceSpeaker string  `json:"source_speaker" yaml:"source_speaker"`
	MixFactor     float64 `json:"mix_factor" yaml:"mix_factor"`
}

TTSV2MixSpeakerSource is one source speaker in a mixed-speaker request.

type TTSV2Request 

type TTSV2Request struct {
	Text    string `json:"text" yaml:"text"`
	Speaker string `json:"speaker" yaml:"speaker"`

	Format     AudioFormat `json:"format,omitempty" yaml:"format,omitempty"`
	SampleRate SampleRate  `json:"sample_rate,omitempty" yaml:"sample_rate,omitempty"`
	BitRate    int         `json:"bit_rate,omitempty" yaml:"bit_rate,omitempty"`

	SpeechRate int `json:"speech_rate,omitempty" yaml:"speech_rate,omitempty"`
	PitchRate  int `json:"pitch_rate,omitempty" yaml:"pitch_rate,omitempty"`
	VolumeRate int `json:"volume_rate,omitempty" yaml:"volume_rate,omitempty"`

	Emotion  string `json:"emotion,omitempty" yaml:"emotion,omitempty"`
	Language string `json:"language,omitempty" yaml:"language,omitempty"`

	ResourceID string           `json:"resource_id,omitempty" yaml:"resource_id,omitempty"`
	MixSpeaker *TTSV2MixSpeaker `json:"mix_speaker,omitempty" yaml:"mix_speaker,omitempty"`
}

TTSV2Request represents a TTS V2 stream request.

type TTSV2WSChunk 

type TTSV2WSChunk struct {
	Audio     []byte `json:"-"`
	IsFinal   bool   `json:"is_final"`
	Event     int32  `json:"event"`
	ReqID     string `json:"reqid,omitempty"`
	TraceID   string `json:"trace_id,omitempty"`
	LogID     string `json:"log_id,omitempty"`
	ConnectID string `json:"connect_id,omitempty"`
}

TTSV2WSChunk is one downstream chunk from TTS V2 WebSocket stream.

type TTSV2WSConfig 

type TTSV2WSConfig struct {
	Speaker    string      `json:"speaker" yaml:"speaker"`
	Format     AudioFormat `json:"format,omitempty" yaml:"format,omitempty"`
	SampleRate SampleRate  `json:"sample_rate,omitempty" yaml:"sample_rate,omitempty"`

	// ResourceID defaults to seed-tts-2.0 when empty.
	ResourceID string `json:"resource_id,omitempty" yaml:"resource_id,omitempty"`
}

TTSV2WSConfig is bidirectional TTS V2 WebSocket session config.

type TTSV2WSSession 

type TTSV2WSSession struct {
	// contains filtered or unexported fields
}

TTSV2WSSession represents one bidirectional TTS V2 WebSocket session.

func (*TTSV2WSSession) CancelSession 

func (s *TTSV2WSSession) CancelSession(ctx context.Context) error

CancelSession cancels the current session.

func (*TTSV2WSSession) Close 

func (s *TTSV2WSSession) Close() error

Close closes the session.

func (*TTSV2WSSession) Recv 

func (s *TTSV2WSSession) Recv() iter.Seq2[*TTSV2WSChunk, error]

Recv yields TTS output chunks.

func (*TTSV2WSSession) SendText 

func (s *TTSV2WSSession) SendText(ctx context.Context, text string, isLast bool) error

SendText sends one text piece. When isLast=true, it also sends FinishSession.

func (*TTSV2WSSession) StartNextSession 

func (s *TTSV2WSSession) StartNextSession(ctx context.Context) error

StartNextSession starts a new session on the same WebSocket connection.

type Task 

type Task[T any] struct {
	ID string
	// contains filtered or unexported fields
}

Task represents an asynchronous task.

func NewTask 

func NewTask[T any](id string, poller TaskPoller[T], interval time.Duration) *Task[T]

NewTask creates an asynchronous task handle.

func (*Task[T]) SetFailureMapper 

func (t *Task[T]) SetFailureMapper(mapper TaskFailureMapper[T]) *Task[T]

SetFailureMapper customizes terminal task failure error mapping.

func (*Task[T]) SetStatusMapper 

func (t *Task[T]) SetStatusMapper(mapper TaskStatusMapper) *Task[T]

SetStatusMapper customizes task status normalization.

func (*Task[T]) Wait 

func (t *Task[T]) Wait(ctx context.Context) (*T, error)

Wait blocks until task completion.

type TaskFailureMapper 

type TaskFailureMapper[T any] func(status TaskStatus, result *T) error

TaskFailureMapper converts terminal failure status to a concrete error.

type TaskPoller 

type TaskPoller[T any] func(ctx context.Context, taskID string) (status TaskStatus, result *T, err error)

TaskPoller polls current task status and optional result.

type TaskStatus 

type TaskStatus string

TaskStatus is async task status. 

const (
	TaskStatusPending    TaskStatus = "pending"
	TaskStatusProcessing TaskStatus = "processing"
	TaskStatusSuccess    TaskStatus = "success"
	TaskStatusFailed     TaskStatus = "failed"
	TaskStatusCancelled  TaskStatus = "cancelled"
)

type TaskStatusMapper 

type TaskStatusMapper func(status TaskStatus) TaskStatus

TaskStatusMapper normalizes raw task status values.

type Utterance 

type Utterance = ASRV2Utterance

type VoiceCloneRequest 

type VoiceCloneRequest struct {
	// VoiceID is a custom voice identifier.
	VoiceID string `json:"voice_id,omitempty" yaml:"voice_id,omitempty"`
	// SpeakerID is an alias of VoiceID for compatibility with official docs.
	SpeakerID string `json:"speaker_id,omitempty" yaml:"speaker_id,omitempty"`

	Audio []byte `json:"-" yaml:"-"`

	AudioFileName    string `json:"audio_file_name,omitempty" yaml:"audio_file_name,omitempty"`
	AudioContentType string `json:"audio_content_type,omitempty" yaml:"audio_content_type,omitempty"`
	AudioFormat      string `json:"audio_format,omitempty" yaml:"audio_format,omitempty"`

	Text      string `json:"text,omitempty" yaml:"text,omitempty"`
	Language  int    `json:"language,omitempty" yaml:"language,omitempty"`
	ModelType int    `json:"model_type,omitempty" yaml:"model_type,omitempty"`
	Source    int    `json:"source,omitempty" yaml:"source,omitempty"`

	ResourceID   string        `json:"resource_id,omitempty" yaml:"resource_id,omitempty"`
	PollInterval time.Duration `json:"-" yaml:"-"`
}

VoiceCloneRequest is the request payload for voice clone upload task.

type VoiceCloneService 

type VoiceCloneService struct {
	// contains filtered or unexported fields
}

VoiceCloneService provides voice clone training and status operations.

func (*VoiceCloneService) Activate 

func (s *VoiceCloneService) Activate(ctx context.Context, voiceID string) error

Activate formalizes a trained cloned voice.

func (*VoiceCloneService) GetStatus 

func (s *VoiceCloneService) GetStatus(ctx context.Context, speakerOrVoiceID string) (*VoiceCloneStatus, error)

GetStatus queries current voice clone task status.

func (*VoiceCloneService) Submit 

func (s *VoiceCloneService) Submit(ctx context.Context, req *VoiceCloneRequest) (*Task[VoiceCloneStatus], error)

Submit uploads training audio and returns a task handle for polling.

func (*VoiceCloneService) Upload 

func (s *VoiceCloneService) Upload(ctx context.Context, req *VoiceCloneRequest) (*Task[VoiceCloneStatus], error)

Upload uploads training audio and returns a task handle for polling.

type VoiceCloneStatus 

type VoiceCloneStatus struct {
	TaskID    string `json:"task_id,omitempty"`
	SpeakerID string `json:"speaker_id,omitempty"`
	VoiceID   string `json:"voice_id,omitempty"`

	Status        TaskStatus `json:"status"`
	RawStatus     string     `json:"raw_status,omitempty"`
	RawStatusCode int        `json:"raw_status_code,omitempty"`

	StatusCode    int    `json:"status_code,omitempty"`
	StatusMessage string `json:"status_message,omitempty"`

	Version    string `json:"version,omitempty"`
	DemoAudio  string `json:"demo_audio,omitempty"`
	CreateTime int64  `json:"create_time,omitempty"`

	ReqID   string `json:"reqid,omitempty"`
	TraceID string `json:"trace_id,omitempty"`
	LogID   string `json:"log_id,omitempty"`
}

VoiceCloneStatus is one status snapshot of clone training task.

type Word 

type Word = ASRV2Word

Source Files

View all Source files

Directories

Path Synopsis
examples
asr_v2_sauc_ws command
realtime command
tts_v2/http_stream command
tts_v2/websocket command
voice_clone command
internal
auth
protocol
transport
util

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL
go.dev uses cookies from Google to deliver and enhance the quality of its services and to analyze traffic. Learn more.