github/faster-whisper
1
1
Fork 0
mirror of https://github.com/SYSTRAN/faster-whisper.git synced 2026-01-08 13:14:00 -05:00
Code Issues Packages Projects Releases Wiki Activity

Fix neg_threshold (#1191)

Browse Source
This commit is contained in:
Purfview
2024-12-29 11:38:58 +00:00
committed by GitHub
parent 95164297ff
commit 11fd8ab301
1 changed files with 4 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
faster_whisper/vad.py
View File

@@ -35,7 +35,7 @@ class VadOptions:
"""
threshold: float = 0.5
neg_threshold: float = threshold - 0.15
neg_threshold: float = None
min_speech_duration_ms: int = 0
max_speech_duration_s: float = float("inf")
min_silence_duration_ms: int = 2000
@@ -63,6 +63,7 @@ def get_speech_timestamps(
vad_options = VadOptions(**kwargs)
threshold = vad_options.threshold
neg_threshold = vad_options.neg_threshold
min_speech_duration_ms = vad_options.min_speech_duration_ms
max_speech_duration_s = vad_options.max_speech_duration_s
min_silence_duration_ms = vad_options.min_silence_duration_ms
@@ -90,7 +91,8 @@ def get_speech_timestamps(
triggered = False
speeches = []
current_speech = {}
neg_threshold = vad_options.neg_threshold
if neg_threshold is None:
neg_threshold = max(threshold - 0.15, 0.01)
# to save potential segment end (and tolerate some silence)
temp_end = 0