Bugfix: Illogical "Avoid computing higher temperatures on no_speech" (o…

…penai#1903) * Bugfix: Illogical "Avoid computing higher temperatures on no_speech" Bugfix for openai#1279 It's "silence" when decoding has failed due to `compression_ratio_threshold` too, when further down the code it's not "silence" anymore. "Silence" should be only when decoding has failed due to `logprob_threshold`. Like described there: https://github.com/openai/whisper/blob/8bc8860694949db53c42ba47ddc23786c2e02a8b/whisper/transcribe.py#L421 And in code there: https://github.com/openai/whisper/blob/8bc8860694949db53c42ba47ddc23786c2e02a8b/whisper/transcribe.py#L243-L251 * Fix if "logprob_threshold=None" --------- Co-authored-by: Jong Wook Kim <[email protected]>
joelvaneenwyk · Dec 31, 2024 · 07e961b · 07e961b
1 parent 7c3c485
commit 07e961b
Showing 1 changed file with 2 additions and 0 deletions.
diff --git a/src/whisper/transcribe.py b/src/whisper/transcribe.py
@@ -219,6 +219,8 @@ def decode_with_fallback(segment: torch.Tensor) -> DecodingResult:
             if (
                 no_speech_threshold is not None
                 and decode_result.no_speech_prob > no_speech_threshold
+                and logprob_threshold is not None
+                and decode_result.avg_logprob < logprob_threshold
             ):
                 needs_fallback = False  # silence
             if not needs_fallback: