Types & classes — github.com/cjpais/transcribe-rs

CodeHub MCPcopy Create free account

hub / github.com/cjpais/transcribe-rs / types & classes

Types & classes70 in github.com/cjpais/transcribe-rs

⨍ Functions 374◇ Types & classes 70

ClassAccelGuard

RAII guard that serialises access to global state and restores defaults when dropped.

src/accel.rs:305

ClassBinTokenizer

Binary tokenizer for streaming models.

src/onnx/moonshine/streaming.rs:189

ClassCanaryModel

Canary speech model backed by three ONNX sessions (preprocessor, encoder, decoder).

src/onnx/canary/mod.rs:89

ClassCanaryParams

src/onnx/canary/mod.rs:58

EnumCanaryVariant

src/onnx/canary/mod.rs:18

ClassCohereModel

src/onnx/cohere/mod.rs:47

ClassCohereParams

src/onnx/cohere/mod.rs:38

ClassCtcDecoderResult

Result of CTC greedy decoding for a single utterance.

src/decode/ctc.rs:4

ClassEnergyAdaptiveChunked

Adaptive chunked transcription using energy-based split point search. Targets a fixed chunk duration but adjusts the actual split point to land on a

src/transcriber/energy_adaptive_chunked.rs:54

ClassEnergyAdaptiveConfig

Configuration for [`EnergyAdaptiveChunked`].

src/transcriber/energy_adaptive_chunked.rs:7

Simple RMS energy-based VAD. Zero dependencies. Classifies a frame as speech if its RMS energy exceeds the configured threshold. Suitable for clean a

src/vad/mod.rs:66

ClassFailOnNthModel

Mock model that fails on the Nth call.

src/transcriber/test_helpers.rs:40

src/whisperfile.rs:136

ClassGigaAMModel

src/onnx/gigaam/mod.rs:31

ClassGigaAMParams

src/onnx/gigaam/mod.rs:26

ClassGpuDeviceInfo

src/whisper_cpp/gpu.rs:23

src/whisper_cpp/gpu.rs:14

ClassGreedyDecoder

src/decode/greedy.rs:11

src/onnx/moonshine/model.rs:277

src/features/mel.rs:14

Mock model that returns the sample count as text.

src/transcriber/test_helpers.rs:8

ClassModelCapabilities

ClassMoonshineModel

src/onnx/moonshine/model.rs:41

ClassMoonshineParams

src/onnx/moonshine/model.rs:34

ClassMoonshineStreamingParams

src/onnx/moonshine/streaming.rs:33

ClassMoonshineTokenizer

src/onnx/moonshine/model.rs:349

EnumMoonshineVariant

src/onnx/moonshine/mod.rs:11

ClassMultipartForm

Custom multipart form-data builder for HTTP requests.

src/whisperfile.rs:40

ClassOpenAIEngine

src/remote/openai.rs:57

EnumOpenAIModel

src/remote/openai.rs:128

ClassOpenAIRequestParams

src/remote/openai.rs:86

EnumOrtAccelerator

src/accel.rs:31

ClassParakeetModel

src/onnx/parakeet/mod.rs:85

ClassParakeetParams

src/onnx/parakeet/mod.rs:28

EnumQuantization

src/onnx/mod.rs:15

InterfaceRemoteTranscriptionEngine

src/remote/mod.rs:14

src/onnx/parakeet/mod.rs:73

ClassSenseVoiceMetadata

src/onnx/sense_voice/mod.rs:38

ClassSenseVoiceModel

src/onnx/sense_voice/mod.rs:52

ClassSenseVoiceParams

src/onnx/sense_voice/mod.rs:29

Silero VAD using an ONNX model with LSTM state. Classifies 30ms audio frames (480 samples at 16kHz) as speech or non-speech using the Silero VAD v4 m

src/vad/silero.rs:41

ClassSmoothedVad

Wraps any [`Vad`] with onset detection and hangover smoothing. - **Onset detection**: requires `onset_frames` consecutive speech frames before transi

src/vad/mod.rs:116

InterfaceSpeechModel

Unified interface for speech-to-text models. Each model implements this trait to provide a common transcription API. Model-specific parameters are ex

ClassStreamingConfig

src/onnx/moonshine/streaming.rs:42

ClassStreamingModel

Streaming Moonshine model with 5 ONNX sessions.

src/onnx/moonshine/streaming.rs:273

ClassStreamingState

Internal state for streaming inference.

src/onnx/moonshine/streaming.rs:112

ClassSymbolTable

Symbol table mapping token IDs to strings. Supports two file formats: - `symbol id` (split on last whitespace, used by SenseVoice) - Optionally base6

src/decode/tokens.rs:44

EnumTimestampGranularity

src/onnx/parakeet/mod.rs:19

ClassTimestampedResult

src/onnx/parakeet/mod.rs:79

src/onnx/parakeet/mod.rs:57

EnumTranscribeError

ClassTranscribeOptions

InterfaceTranscriber

A chunked transcription strategy. Implementations split audio into chunks, transcribe each chunk via a [`SpeechModel`], and merge the results. The mo

src/transcriber/mod.rs:115

ClassTranscriptionResult

ClassTranscriptionSegment

Voice activity detection. Implementations classify fixed-size audio frames as speech or non-speech. The frame size is determined by the implementatio

src/vad/mod.rs:34

ClassVadChunked

VAD-based chunked transcription. Splits audio on speech/silence boundaries detected by a [`Vad`]. Each speech region is transcribed independently, th

src/transcriber/vad_chunked.rs:51

ClassVadChunkedConfig

Configuration for [`VadChunked`].

src/transcriber/vad_chunked.rs:8

src/onnx/canary/vocab.rs:7

EnumWhisperAccelerator

src/accel.rs:177

ClassWhisperEngine

Whisper speech recognition engine.

src/whisper_cpp/mod.rs:139

ClassWhisperInferenceParams

src/whisper_cpp/mod.rs:79

ClassWhisperLoadParams

src/whisper_cpp/mod.rs:54

ClassWhisperfileEngine

Whisperfile speech recognition engine. Manages the whisperfile server lifecycle automatically.

src/whisperfile.rs:206

ClassWhisperfileInferenceParams

src/whisperfile.rs:185

ClassWhisperfileLoadParams

src/whisperfile.rs:165

ClassWhisperfileOutput

src/whisperfile.rs:96

ClassWhisperfileSegment

src/whisperfile.rs:103

src/features/mel.rs:7

src/onnx/parakeet/mod.rs:65