diff options
| author | Konstantin <const@const.me> | 2023-01-16 14:52:43 +0100 |
|---|---|---|
| committer | Konstantin <const@const.me> | 2023-01-16 14:52:43 +0100 |
| commit | 8c4603c73675958efc960fbd4bb599a2909d106a (patch) | |
| tree | 714dc6fc9a1672d5fd7f89676b97e10959662abc /WhisperNet/API/SpecialTokens.cs | |
| parent | 990a8d0dbaefc996244097397259e92758b15cce (diff) | |
Source codes
Diffstat (limited to 'WhisperNet/API/SpecialTokens.cs')
| -rw-r--r-- | WhisperNet/API/SpecialTokens.cs | 23 |
1 files changed, 23 insertions, 0 deletions
diff --git a/WhisperNet/API/SpecialTokens.cs b/WhisperNet/API/SpecialTokens.cs new file mode 100644 index 0000000..d672369 --- /dev/null +++ b/WhisperNet/API/SpecialTokens.cs @@ -0,0 +1,23 @@ +namespace Whisper +{ + /// <summary>Special tokens defined in the model</summary> + public readonly struct SpecialTokens + { + /// <summary>The end of a transcription</summary> + public readonly int TranscriptionEnd; // token_eot + /// <summary>Start of a transcription</summary> + public readonly int TranscriptionStart; // token_sot + /// <summary>Represents the previous word in the transcription. It is used to help the model predict the current word based on the context of the words that came before it.</summary> + public readonly int PreviousWord; // token_prev + /// <summary>Start of a sentence</summary> + public readonly int SentenceStart; // token_solm + /// <summary>Represents the word "not" in the transcription</summary> + public readonly int Not; // token_not + /// <summary>New transcription</summary> + public readonly int TranscriptionBegin; // token_beg + /// <summary>token_translate</summary> + public readonly int TaskTranslate; + /// <summary>token_transcribe</summary> + public readonly int TaskTranscribe; + } +}
\ No newline at end of file |
