ZipformerOptions
Options & Configuration · AiDotNet.Audio.SpeechRecognition
Configuration options for the Zipformer speech recognition model.
Inheritance: ModelOptions → ZipformerOptions
For Beginners
Zipformer is an improved version of the Conformer that’s both faster and more accurate. It processes speech at different “zoom levels” - some parts look at fine details (individual sounds) and other parts look at the bigger picture (whole words and phrases). This makes it one of the most efficient speech encoders available.
How It Works
Zipformer (Yao et al., 2023, Next-gen Kaldi) is a more efficient variant of the Conformer with temporal downsampling, BiasNorm instead of LayerNorm, and SwooshR/SwooshL activations. It uses a U-Net-like structure with different time resolutions at different encoder stacks, achieving better accuracy with fewer parameters than standard Conformer.
Constructors
| Constructor | Summary |
|---|---|
ZipformerOptions() |
Properties
| Property | Summary |
|---|---|
OnnxModelOptions OnnxOptions { get; set; } |
Gets or sets the ONNX runtime options. |
double DropoutRate { get; set; } |
Gets or sets the dropout rate. |
double LearningRate { get; set; } |
Gets or sets the learning rate. |
int NumMels { get; set; } |
Gets or sets the number of mel spectrogram channels. |
int SampleRate { get; set; } |
Gets or sets the audio sample rate in Hz. |
int VocabSize { get; set; } |
Gets or sets the vocabulary size. |
int[] DownsampleFactors { get; set; } |
Gets or sets the downsampling factors per stack. |
int[] EncoderDims { get; set; } |
Gets or sets the encoder dimensions at each stack (U-Net style). |
int[] NumHeadsPerStack { get; set; } |
Gets or sets the attention heads per stack. |
int[] NumLayersPerStack { get; set; } |
Gets or sets the number of layers per encoder stack. |
string Language { get; set; } |
Gets or sets the language code. |
string ModelPath { get; set; } |
Gets or sets the path to the ONNX model file. |
string Variant { get; set; } |
Gets or sets the model variant (“small”, “medium”, “large”). |
string[] Vocabulary { get; set; } |
Gets or sets the CTC vocabulary (characters or BPE tokens). |